Hi Alexander, I have thought about that, but mostly for functions like collet closer, or tailstock engage, which don't have an indication on the screen.
@@RotarySMP You just want to hang onto your 50-Euro-Interface card... I think all the 9 buttons you've shown like CCW/OFF/CW, Flood/OFF/Mist etc. would benefit from it. Additional bonus with LED buttons vs. mechanical switches for example is that a program can override it, but you still get the visual feedback. If anything, running a program that goes back and forth between modes is going to be a nice light show.
@@graealex I think that MESA 7i73 has some outputs, so it could probably drive button LED's. But I was going to machine the buttons to shape, and actuate standard industrial contact blocks. That is what I did on the Maho (except I 3D printed the buttons).
@@RotarySMP Most button contact block systems have light blocks that can be stacked in with the switch contacts. And for the button machining itself, even if you want to make the button bodies themselves from an opaque material, you can still drill a series of holes on the bottom of the text engraving and then fill them and the engraving with epoxy resin.
Put the jog increment selector on the same panel as the X and Z encoders. Also make it a push-button to set Manual mode so that jogging actually works.
@@RotarySMP my Okuma has them both on the panel and on the pendant. If the pendant is set to anything other than 0, it overrides the controls on the panel, and lights up an indicator to let you know the pendant is in control. It's nice to have it in both places.
@@jrucker2004 Some of those nice to have features would probably be quite easy to do in LinuxCNC for an expert, but for a guy like me get kind of overwhelming.
Unfortunately, most of them are designed with laptop like motherboards in terms of io and power management, which is the kiss of latency death for decent Linuxcnc performance.
For your keyboard, you might consider a drawer, there are metal drawers especially for hiding your keyboard in machine control panels even including a slight angle to make it more comfortable to type on. Apparently they are a bit pricy at 200€+, but maybe you can make an affordable custom piece from all the sheet metal.
I know a thing or two about UX design, and the mockup you showed makes a lot of sense. One thing another commenter mentioned that I wouldn't definitely do is illuminated buttons: If you put in some effort, you can go for a "dark cockpit" style control, where only buttons that have a function at the moment light up (I know it's not 100% the same, but I like the concept). Also, consider breaking up that 2x3 matrix a bit more, so that it's clear that those 3 keys are a group - and if you do more groups that feed into the same thing, keep them in horizontal groups as well. On the pendant, I'd add a spindle clamp button, and spindle controls - maybe also tailstock controls.
Vincent... we have been friends for so long, and now you go and sell me out to the LED lit button mafia :(. I am so disappointed! Good inputs though :) Actually I once when down that rabbit hole of the dark cockpit, Boeing style integral lighting. There are some amazing sim builders who model all that stuff. Why would you put the collet clamp and tailstock down on the pendant? Those would seem like set up tasks.
@@RotarySMP I never said you had to use LED buttons, I'm sure you can find some nice buttons with incandescent bulbs in the scrap bin at work. As for setup buttons - yeah, that was the intention. I'm not 100% sure if you would use them, but to me, they seem like potentially useful additions. If in doubt, just look if some other MPGs have buttons like this, those manufacturers seem to know what they're doing.
@@vincentguttmann2231 I thought a bit more about it and you are right, that the collet clamping will probably have me holding some small part at the spindle, and it could be quite a stretch to reach up to the control panel to clamp it. They whole volume of comments have rather overwhelmed my plans and ideas a bit. I need to thick this all through more. I was going to make all the buttons on the Maho.
Love that the algorithm led me here. I've never designed a control, but I've used a lot and developed some strong opinions on them. One thing I've thought of, and wished controller designers would do, is make a spaghetti diagram of the workflow on a controller. In other words, use a string to follow the sequence of button pushes for any particular function, like loading a program, starting from the middle of a program, setting tools, setting work offsets, etc. Achieving the shortest possible string would probably give you the best user experience. But this is just an unproven idea.
Nice idea, but isnt that something for the training manual? I am not sure how I could execute this, as my interface is modal, so buttons have different function in different modes.
@@RotarySMP maybe spaghetti diagram isn't the right term. The purpose is to reduce the amount of hand movements and button pushes. For example, to load a program from a USB drive on my fanuc control, I have to go to edit mode, program hard key, folder soft key, change device, USB mem, cursor down to the program, select soft key, copy, plus, device change, CNC mem, select, paste, cursor down to the program, plus, main program. And more if I want to run that program and have the proper screen displayed. It's tedious. The hass control is a little better, but my favorite was the Heidenhain MillPlus 3.4 which has the added benefit of customizing your layout (which I never used, the default was good enough).
I'm an HCI person and I highly recommend having tactile/non-binary buttons for things. Especially, if you want to reach them "blindly". Having the keyboard in front is better for your wrists. Also I'd highly recommend a modern game controller with haptic feedback (not just rumble) for jogging/tool changing - just go and try out the PS5 controller on the PS5 demo game and you'll know what I'm talking about. It is mind blowing :)
Maho control box looks awesome on Schaublin, I'd use that. Also more than enough space to fit all LinuxCNC hardware inside, and it's above the lathe, so no further space needed. Electronic handwheel somewhere with good access, always sucks to use stationary handwheel on the other end of CNC lathe, when dialing something in.
For USB ports, maybe use the round style, that saves you from having to file out a rectangle in the plate. There are even variants available with two 3.0 in a single housing for around 15 EUR. There's also an Austrian company that sells specialized types, I recently ordered a few from them.
@@DreitTheDarkDragon Good ideas, although I didn't file out the rectangular slots on the Maho panel, as it was laser cut. I think I'll get the front sheets laser cut again, as I want them to be stainless, and I hate working with the stuff. :)
@@RotarySMP Btw the company from Austria I was referring to is called "RIBU Elektronik". The dual-USB panel mount is Amazon ASIN B086Y944DX for less than 15 EUR and even has a splash-proof cover, so you don't get metal shavings or coolant in there. And of course Conrad and the likes have a lot more stuff.
Dutch! It all makes sense now 😄 Thank you for the shared wisdom! I'm moving this summer hopefully into a place with a proper garage and room for these sort of setups.
I like it. For the rail attached pendant: I've seen a demo for a force feedback HMI for a lathe a long time ago. That way one can use a CNC as a manual lathe. Even if you don't do that yet, it would be a nice upgrade in the future.
@@RotarySMP I don't have experience with it, but it seems very capable. Theoretically, you have to put a tiny servo motor in the wheel. The wheel servo's encoder signal fed to the axis servo as target position, and the A-servo's torque is fed to the W-servo as torque target. If LinuxCNC has access to the servo drives' data, it is possible. Tuning the system can be an issue. Delays, feedback loop oscillations, SW filters, etc. I'm pretty sure somebody tried this in the community.
I run an old Emco Concept mill 150 at the moment. It has two chunky rotary coded knobs for the mode/jog increment selection and also one dedicated to the feed override. They’re beautifully tactile sturdy and very easy to find without looking. If I’m running a new program my hand is basically always on the feed knob and if anything looks like it might be going pear shaped I just crank it to zero. I don’t even use the pause or stop button because this gets me out of 99% of situations. With other machines it’s always much more of a panic to find the tiny buttons or switch the pendant into the right mode. I would highly recommend looking into something like this or at least doing yourself a favour and increasing your knob size.
Thanks Nick, possibly the first time I have been critiqued online for my knob size :) I have larger feed and jog rate overide knobs on the Maho, and you are right it is very fast to dial it back to zero,
No to the joystick! Yes to 2 encoder wheels. Imagine trying to touch off on a part with a joystick 😬 - I like the sliding pendant, and use something similar at work. I would add the e-stop on the top instead of parallel to the machine - a long with a button for coolant and buttons for the collet, and any other buttons you'd fancy having there.
Thanks for your feedback Mark. It is good to hear from people with experience on machines like this. I already picked up that I will need more buttons on that front sliding "pendant".
For my money I’d rather have an air blast button than a mist coolant. I’ve never really found a use for mist in a lathe with a full cage. The air blast though is used often.
Hi Mark just back from an enjoyable Easter dinner with my family, hope everyone has had an enjoyable weekend as well. In other comments you have mentioned that you usually program on the console. given that there are 3 basic activities on any CNC , 1) program inputting,editing and selection. 2) machine setup or manual operation, 3) fully automatic operation, It is important to avoid any egonomic decisions that would make any of these operations needlessly interfere with the other one such decision would be keyboard operation and estop or other machine control switch fouling each other. As you have stated , You are inclined to mount the handwheel and other controls onto the apron control ,in which case anything associated with setup /manual operation could be mounted there . During this time the screen becomes a big DRO. tool offsetting is about the only reason that you would touch the keyboard and monitor. During auto mode about the only things happening are cycle start/stop, feedrate override, spindle override , and coolant override, oh and hoping that the program is correct . these functions and the e-stop could be grouped together in a third VERY accessable spot .
Hi Steve, that is a good approach to it Thanks. We also had family over for lunch, so there was a food induced coma in the afternoon. Got nothing done :/
about the Jog wheels - I really liked the pendant idea in front of the machine, as during jogging you will probably often be in a situation where you want to watch closely how far you jog, so the position in front of the machine is super useful there. one more thought - you could put the jogging encoder wheels not next to each other but on two sides (front and right) of the pendant - oriented as their respective axes are. this way it would almost feel like you're operating the cross slide of a traditional lathe
That is the way Schaublin did it on the 125CCN with Fanuc, the X encoder is on the side For me that might be weird, as the Boly has the cross slide hand wheel in the LH, rather than the RH.
I got a iiyama ProLite T2435MSC-B2 24" Touchscreen-Monitor works very good. Has a glass screen so better protected against metal chips. Get two encoders for X and Z for slow step and one joystick for fast stepping. The most problems are forgetting it is on fast stepping.
I would suggest you go with the biggest screen you can physically fit in the cabinet. Almost everybody bemoans the fact that they cut costs by going to a smaller screen whereas you rarely hear people moaning that their screen is to big. To cater for buttons down the side they can be outside the existing cabinet (a side car arrangement). Just make a facia panel that covers the full width. (spend the coin and get a stainless sheet laser cut and etched) As for the keyboard either makeup a drawer with ball roller slides that slides under the lower control panel. Roller slide rails are quite cheap online, particularly for the size you would require. Alternately a hinged panel that has the keyboard on the under side (when closed). Opening the panel exposes the control panel and swings the keyboard up the right way, also lowering the keyboard to a better height. When closed the keyboard would cover the control panel (keyboard would be upside down). You could also make it so the E-Stop is pressed when the panel is closed.
Thanks Kim, good ideas. Since Gmoccapy is a fix resolution, I dont feel the need for a huge monitor. The 17" on the Maho seems somewhat too big. I like the slide out keyboard though.
Two thoughts: keep your E stop well clear of things and on the left. I reckon go separate manual jog wheels for Z and X on a lathe, but go different sizes and shapes and even positions so that they approximate cross slide and apron handwheels.
I did think about making up a box with real wheels driving the encoders, but couldnt find reasonably priced 100PPR hollow shaft encoders, and it would take me forever to do that project. I will stagger the jog wheels though.
On my lathe the hand wheels are the same size and both on the same level. It didn't take long to get used to them. I found a high count encoder (500PPR or more) makes it feel more like a manual handwheel. If you use encoders you have to have some sort of brake/friction mechanism otherwise the wheel will turn due to vibration or even gravity if it's not perfectly balanced.
I like your videos and the way you solve problems and explain at the same time. Can i give you a sugestion? Put the control of the collet open/close next to XZ jog, also one increment selector for each axel next to them also. Gut gemacht 👍🏻
Thanks for the tip. I realised afterwards that I dont need collet clamping on the panel. It is a footswitch I already set up. I am not sure if LinuxCNC supports separate jog increments per axis, or if I would want that. It seems more predicable if both are the same.
There is an electric brake integrated into the main spindle motor. I just haven't got around to wiring it up, and integrating into the control system yet. I figured I'd use the mechanical break rather than electrical.
@@RotarySMP Electric brake doesn't have any wear, and you don't need any special wiring. Just put in a resistor, and tell the VFD to use it. It basically shorts the motor to the resistor, creating magnetic drag. You can always combine it with the mechanical brake, which has it's own benefits. For example, servo drives with mechanical normally-closed brakes don't need any homing - although probably not relevant for your spindle, maybe even counterproductive.
@@graealex I dont really have space for the breaking resister, as some moron chopped down that control cabinet so small that it is full up :/ I guess I could put one in the next bay, I'll keep that as plan B if the mechanical brake thing doesn't work for me.
Nice work so far. I have a lathe at work I use that has jog encoders mounted to the carriage, and it also has a jog joystick mounted between the encoders. The joystick is set to a faster jog rate so far rapid moves it's quite nice. I use the joy stick nearly every time I use the lathe, as well as the jog encoders. I think it's valuable to have both options!
For things that can be G-code controlled (on my Holbrook this is only the coolant, but in your case it could include the collet, if you were to fit a bar puller) I have three-way switches for on, auto, off. It's nice to be able to force the coolant on or off if the G-code omitted it (or has it on when it should be off)
@@andypughtube Nice. I even have some three way switches which use the standard contact blocks I design around. I can see overriding coolant being really handy. I have often made a mess because the CAM set it automatically.
I just continue to be awed by your ability to put all this together. Very impressive. And I know there are people out there even more versed in this which just makes me feel inadequate! Nice work. And the occasional aircraft tidbits are most welcome.
Thanks for the kind feedback Glenn. I also feel like I have little idea, as there are so many smart people out there. I really appreciate when they take their time to share their knowledge.
I like the overall plan. Reusing the factory box and cutting it down is a great idea but it seems like it'll still be on the big side. Maybe a flip up keyboard for when not in use? Separate encoders are a given instead of the switch. I'd be constantly moving the wrong axis. FYI that Bridgeport series 2 I mentioned before is now almost free but finding a place for that 15K pound 7500KG beast and the cost of getting it here... Still no film to drop off in Naxos yet but I'm trying to come up with something else for you. Will be there on the 8th next month.
Thanks for the video update, no sparks and swarf is totally forgivable. I wish I could offer some useful insights regarding the control panel layout but I'll be reading the other responses because I have no idea. Hopefully there are plenty of good suggestions inbound. - Heather
In my opinion the keyboard definitely needs to be in front (closer to the user) of the button block. I would recommend a little palm rest if you plan or expect to spend any amount of time typing. At work we have one machine where the keyboard is arranged between the screen and the buttons, that is really annoying to use because you have to hover your whole arm over the keyboard while typing to not press any of the buttons or stab your wrist on the key-switch.
Yes, having no way to actually rest your arms/palms without pressing buttons or manipulating dials when using the keyboard is cumbersome. Keyboard always needs to be the nearest thing. And the membrane keyboards don't really suffer from accidental key presses either.
@@RotarySMP Drawer is probably the best idea. You could even go for a real keyboard then, instead of the membrane one. It's just going to be a lot of work.
I have the solution for cleaning the tag plate, just remove it, build a vaporblasting cabinet, buy a big compressor and then vaporblast it. And as a bonus you can now vaporblast parts. I built mine so it also can be used for dryblasting.
I use a joystick for X and Z jogging and love it. The speed of jogging is proportional to the stick deviation. Would hate going back to wheels. Ciao, Marco.
@@RotarySMP Made a video about it 8 years ago, here it is... th-cam.com/video/ixOsdKvbKlk/w-d-xo.html&ab_channel=EZshop I also use it with my lathe, obviously the buttons are mapped differently in that use case.
They are a great system. If you speak german, check out Peter, (Talla83) channel on youtube. If you dont speak german, switch on subtitles and check out Peters channel anyway, as he has the best MESA knoeledge of anyone, and shares it generously. Also PWC, a designer from Mesa is active on the LinuxCNC forum, so you can get support direct from the source their.
I've worked a while on a Weiler C30 lathe, it had a joystick which could latch in X+/X- or Z+/Z- direction while turning, which made it feel kinda similar to a manual lathe with feed engaged. It also had two jog wheels, one for each axis. Maybe worth looking into.
I liked the trouble shooting. I have no suggestions regards the control layout, however making a cabinet could be an interesting project and will let you set the angle you want for the screen and controls. Perhaps have it on a adjustable mounting?
Hi over the ditch. I would use the Maho control unit. I would move all the control buttons onto the slop use the larger screen with no buttons next to it. It makes for a cleaner unit. Place the key board at the bottom.
@@RotarySMP Also if touchscreen fails, you can still use physical keys :) At work we used one robot pendant with wireless mouse because screen died and we just had keep machine working to finish rest of batch. It was working, but it really wasn't comfortable. There actually were physical keys to start, stop and restart cycle + to move all axis. Problem is that to move axis you have to switch to different screen on touch panel and then back to run in automatic mode. That was one thing which wasn't present on physical key and only reason we had to use mouse at all. Also to change program, but that batch had to be running for two more weeks.
@@RotarySMP Yeah, I guess sometimes that happens everywhere. Funny thing is that I suggested to buy replacement pendant for that robot some months before it happened. We even got price offer but somehow nobody ordered it.
As a hobbyist, I understand the frustration of not knowing which way to turn the encoder. I've considered using a combination joystick and encoder as a solution. The joystick would only determine the direction, and then the encoder would move in that direction, regardless of which way you turn the encoder. However, it could build bad habits on other machines. Just thinking....
For the screen, I would go with the larger screen and build small extensions for the soft keys. The increased screen size is going to be more valuable.
@@RotarySMP that’s what I thought too when I built my CNC router. By the time the UI was completed though Icons were too small and the gcode viewer was effectively useless. I was thinking more about your eyesight though. My near vision plummeted between 45 and 50 so maybe I’m a bit biased then.
Hi. The only 2 things that I can think of adding is some more control over the spindle speed (since you have that complicated variator + vfd + gearbox setup, it might be nice to have finer control over each part) and a switch for a light (maybe adding one of those gooseneck lights to the crosslide, so that you can more easily see what's happening on the cutting edge). Other than that, looking great! :)
Thanks for reminding me. I was also thinking that it would be nice to at least have a switch from hi-auto-low on the gearbox, as it is a bear to turn the chuck by hand when in back gear. I got two nice goose neck lights with the machine. You remind me that I need to see if I can still mount one on the original mount pad on the back of the machine, with it closer to the wall.
The two controls that I find very useful are the jog speed and jog turtle mode. I have an analog pot hooked up to the 7i73 to adjust jog speed. The turtle button effectively gives you a low range with finer adjust. I would put this on your dual encoder panel on the lathe apron. You probably want the jog step increment controls on that panel as well.
Yes the layout for the Schablin have the same as the Maho (Heidenhain control 320,530 etc) and your keyboard far from the screen between, you have to put the machine buttons , and it is NOODstop not nudestop the only Nude we have is a street in the city Utrcht tells the Dutch Monopoly game 😉👍👍
Hi Mark , regarding Handwheels and Jog control , Fanuc, Meldas, and Seimens, all arrange the handwheel as you have done on the minilathe Jog control is done with simple momentary contact joystick, panel push button ,or membrane panel button . jog speed is controlled either with the 0-200% feed rate override or a second jog speed selector with specific mm/rev settings, just like a manual lathe and probably the prefered approach for a toolroom style lathe. Joysticks were discontinued a long time ago,but I suspect it had more to do with ecomomics and not really being needed in a production environment. Perhaps ,being as Joysticks and pushbuttons are not very expensive a removable plate with either a joystick or 4 buttons could be done to find out which you prefer the electrical connections would be identical. FYI japanese builders have always used omron a16 series buttons ,they are very nice ,and as accessable as potato chips in canada. alco and fuji also make identical switches. Regarding sense of feel,that just doesn't exist in cnc, electronic leadscrew or even powerfeed on a 100 year old lathe . What I have seen with industrial cutting tools is that the feed and speed information that they provide is absolutely true and reliable , if you have done your setup correctly , it will work. This along with being vastly more stable, is why jog instead of handwheel. try aligning something with the jog control instead of the handwheel and you will see what I mean. what you do feel is a sense of doom if something is going wrong😱
I would prefer to have the additional controls and knobs on the right side of the keyboard, also i would put the stop feed hold and single line step buttons on the right side as they are what i use most of the time when testing my gcode! Great job as always!
@@RotarySMP I had spelling wrong Hurco. Made in US. Large control panel. I heard about airplane collection in Nelson NZ. Lot of ww2 and 50’s Vintage. He kept it private and hidden from public.
I would lean very hard towards putting all "manual control" style switches/controls on the carriage (don't even bother with the front rail)... Wheels, e stop, jog increment, even spindle control. I don't have a CNC but I drive my lathe entirely by DRO, and having everything within easy reach of the work area is critical for me. I know that with CNC it's a bit different, but for checking, zeroing, setup, etc, I'd want as much control near the workpiece as I could get. Redundancy doesn't hurt so hopefully you can optimize your controller but also duplicate the necessary functions onto the carriage!
@@RotarySMP I can't wait to see! I'm considering buying a small CNC router to get my feet wet in CNC, because i'd like to do a conversion on my own lathe - even the soft keys you're implementing are an eye opener for me! I'll probably get a wireless pendant to start (I'll likely go Centroid instead of LinuxCNC) and using it enough on the router should hopefully give me the experience needed to design a control panel like the ones you've done!
@@hogtownpens LinuxCNC can seem daunting with it's learning curve, but the guys on the forum are really helpful and knowledgeable. I would recommend it to anyone.
While you’re running programs I’m concerned you’ll rest your wrists on the keyboard, so perhaps put it on a drawer under the control unit? It would also tidy up the unit significantly.
Is there enough space to put the screen up higher, and fit the E-Stop button and other controls below, on the vertical panel, rather than on the sloped base? If needed, cut down the sloped console to the lower the keyboard, or use a drawer, as others have suggested. That would also allow reducing the depth of the sloped console, so you aren't reaching out too far, to get to the buttons and touchscreen. The sloped console then only needs to be deep enough for the keyboard, with a little clearance vs. the buttons above, and perhaps something of a palmrest. Definitely like the sliding pendant idea!
Always, always go with the keyboard layout you're used to (usually qwerty), you wont realize how much a non-standard key layout will slow you down and make you crazy.
Yeah true. I have used the german QUERTZ my whole working life, and english keyboards drive me nuts. I can never find the special symbols. At least I dont need them for CNC.
From experience ... you probably want the jog increment control right next to the jog wheels themselves. Going back and forth between the two will get annoying otherwise.
Just use the cabinet as-is. Use the monitor you already have-- remember you said you wanted a bigger screen... As for the softkeys, just add buttons around the monitor that "hang out" on the outside of the cabinet!
@@RotarySMP Yes, I heard of savages that prefer light mode, but I've never seen one in the wild. What other barbaric traditions do you frown upon? Using analog 0-10V signals to drive a VFD? Admit it!
I think for tool room work a jog button and jog lock button is really useful for each axis. It gives an equivalent function as engaging the power feed on a manual machine. Some tool room machines will link the X and Z axis to machine tapers manually without a program using jog buttons, but I don't think that's very useful.
Gmoccpy locks out jogging when you are not on the jogging page, so if you bump a jog encoder while on the MDI or program page the input will be ignored. The Schaublin 125CCN with the Fanuc control can also link axis and drive along a taper with a single jog button, but I will be implementing Andy Pughs Lathe Makros, so cutting a taper will be easy enough in MDI.
Möglicherweise wäre die angeschrägte Oberseite für alle Hardware Knöpfe gut und nach vorne einen Schubladen Auszug um die Tastatur zu verstauen. Die Schaublin Kiste wäre groß genug um so etwas zu machen.
I think either a pair of 8 jog buttons or a joy stick would be grate for quick and dirty turning „manual style“ and the 45 deg movements for easy Chamfers. A joystick with guided movements is ideal I would guess, so you don’t have to worry about engaging the wrong feed. On the Deckel FP‘s it’s sometimes a bit flimsy 😂 on some Euchner 8 way joysticks there is a push button on top, maybe could be nice for feed override.
@@RotarySMP I mean the manual Deckels 👍🏻 on my Fp1 you can engage the Z and the X feed at the same time (actually not to sure if there was a part preventing this, which was removed)
I don't know a thing about CNC-machines, so I can't halp you with this. But I enjoyed the video and did hit the "like" button. I hope that helps somehow :)
Have you tried a sharpened wooden skewer in a cordless drill with lapping paste to clean the nameplate? I’m developing a twitch when I see clips of it. Love the content as always. Cheers man.
Hi Mark I assume that you have seen the early version of the 125 ccn where the handwheels, jog controls etc are mounted on the apron? same as you were talking about in the video.
I guarantee you crash machines less with a dedicated feed hold button! I use my feed hold button all the time and never hit the e-stop on our Haas. Even in an "emergency" I hit feed hold and reset.
I type this as i watch the video so if you mention this before the end forgive me. With the feedrate override switch instead of a pot, go a twist switch with resistors for adjusting the value ie 0% 1% 5% 10% 25% 50% 75% 90% 95% 100% 105% and so on. This is how the big machines do it and its so hard to switch between the two plus you learn the spot and its amazing
@@RotarySMP I think the benefit he is describing is that a twist-switch would have a proper mechanical stop, while an encoder is just relative. So with a twist switch, you could actually put some proper markings on the panel, and be able to visually confirm the current setting. You could use a binary twist switch, and use only a few input lines to encode many states. 3 lines = 8 states, 4 lines = 16 states, 5 lines = 32 states, etc.
The advantage of a rotary switch instead of an encoder is that you can set it to a specific position by feel. That way you can keep your eyes on the machine while you have your hand on the feed override. Being able to quickly dial it down and then up again is a nice safety thing, to prevent crashes, for new programs. The important thing is to be able to do out by feel.
a very importan choice making a useful control panel, i think make a new one is better that modifying the old one also a nice thing to use is buttons whit led so you have more feedback
It is a nice idea, but were do you get your input for force required from? Servo current is more speed and inertia dependent than cutting force dependent.
@@RotarySMP I am not sure all the cool things one might use it for, but they let you do things like program the acceleration of the knob, so e.g. if you turn the knob "harder" (because you programmed resistance to increase past the 3 o'clock position) the jog goes faster. Or programmable haptic "detents." Or input from a current sensor on the spindle motor for operations where you risk stalling the spindle. There are a lot of possibilities and BLDC motors as haptic input controls are a relatively new idea that is relatively underexplored. Lot of space for creative innovation.
What about some lighting up buttons for mode selecting so you can out of the corner of your eyes quickly see what mode it in ( mdi. Check ( where everything is running at 25% and g-code can be changed well running. Like on a citizen lathe that I use to work on and loved that feature) and auto ) that's my idea for the controller And a light on off button on the side if you also put some lights on the machine
Since this will never be a production machine, I doubt I would ever need the high speed side of it. I am trying to practice dialing back the feed rate and living with the hand on the feed hold with untested programs.
@@RotarySMP yeah but that's not as good for the tool life. since insert tools are ment/designed for a serten feed pr. rpm and temp at the cutting edge (830 C in what's called the flow zone. where the insert is still at full hard/stiffness but the material gets softer and gets pushed away/seperated from the main stock and forms the chips so at least make sure to keep the feed pr. rpm constant and was just personal recomendation from my experience. and what i had put on if it was me cnc but it's your cnc mate and the most important point is that your happy with how the machine turns out
@@flikflak24 I appreciate your inputs on this. As you can see I am not a machinist, so my ideas are pretty half baked. Some of the best ideas presented in this thread (and there are a lot), may be beyond my ablity to implement or code in LinuxCNC.
@@RotarySMP that's also fair. i am however glad to finally see the idear i came up with inteams of air ( if form of mist ) that i came with some time ago finnaly come into excistance by the way are you planing on useing coolant or just straight oil ? the oil is only slightly worse at cooling ( in my experience ) but never needs to be changed at all. it lubs the tools way better so they last way longer ( we whend from about 800 parts pr conrer to 3500 parts pr corner of the insert ( same feed.speed and deapth of cut )) and only needs topping up from the oil that splash out of the machine/part/ships. nothing eraperates away over time ( at least not compared to coolent ) and it also constantly oil the ways and chuck. so need to oil them by hand or any chance of rust
I would seriously consider DUAL jog wheels, vertically positioned with the lower one for lathe Z and the upper one for lathe X. That would be fast and intiutive because it would be like the 2 handwheels on a manual lathe. In some cases you could even drive it like a manual lathe using the two jogwheels, like for simple facing tasks without needing to run a program...
That is what I showed n the video right? Except I put them next to each other. Probably more intuitive if I place then diagonally. X FWD Higher, and Z lower back, as that would be the same as my Boley.
@@RotarySMP ... whoops sorry to confess I only watched the first half of your vid then was called away to do other stuff. I'm a dufuss! 😁 I like Tom S suggestion the best, put the 2 dials ON the machine itself, and possibly a couple other jog-critical controls like a feed rate switch and reversing switches. All in a place where your hands instinctively find the controls so you dont have to take your eyes off the cutter during setup or simple turning tasks.
@@wizrom3046 That is kind of where I was going with the two encoders down in a "pendant" running on the rail at the front of the base. I hear a lot of negative about putting dual encoders on a lathe. I would have thought that to be intuitive?
@@RotarySMP ... I cant see why they think it's a negative. Ultimately its your baby so if you want your baby to have two heads then that is the right amount! 😎
For me... nothing gives personality to a tool / panel like having physical buttons and switches. Personally I enjoy having light indicators in / above the buttons, but that also introduces a point of failure and maintenance. Are the trade-offs for the buttons worth it? A few encoders would also be nice, however just like valves and switches you really want there to be muscle memory for the controls. So having a single function for each control or atleast for each mode is best. Also note... the four corners, and then the edges are the most valuable locations because those can easily be found without having to look over at the controls. Good luck, hope you find a mix you can enjoy... there is always rev2 if you don't like it. =D
Would the estop be better on the front (vertical) panel?seems quite far back. I’ve seen some jog dongles have jog wheels on different faces e.g so they are at 90 degrees to each other does that make it more intuitive? Interesting video 😀
I can only give you a feedback on what I'm currently designing... My aim it to convert a small chinese lathe to a semi-automatic one with electronic gearbox, removed compound and some predefined macros for feeding and threading. All configuration data is handled by a stationary panel with keyboard, small encoders etc, but the operational controls are on a small mobile box that can be hand-held or attached to the carriage (as mechanical ones). I've found that I could live with a single wheel coupled to 2 (X1-X100) or 3 (X1-X10-X100) "speed selectors" pushbuttons and 2 mode selectors, one for jog, the other for manual move/macro definition, both of them providing Neutral/Z move/X move/+X+Z diagonal/+X-Z diagonal. This way I can both handle 30° manual jog for threading (emulated compound) or angled XZ macro for tapers... I really don't feel like using 4_joysticks unless you can lock at times one axis - which defeats the purpose of a joystick... Nice work anyway from Maho to Schaubling - I wish I could handle such big projects...
Thanks for the feedback Jean. Just question, once you get that far along the retrofit path, it would be almost no additional work to just use LinixCNC and have a full CNC. Why the half step from manual to CNC?
@@RotarySMP I agree, that may seem strange, but I don't feel that much interested in a full-cnc lathe. That's completely different for milling/routing. In parallel I'm reviving my first home-made cnc-router (late 1990, 1200mm x 700mm work space, ball screws) for which I will most probably move from grbl+custom python to linuxcnc (I'm particularly interested by your experience with Mesa cards ...). And if I'm still alive, do the same with my small lathe... To be fair, what worries me somewhat is how (well) linuxcnc can cope with backlash on conventional acme screws (of unimpressive precision)... Thanks for answering !
@@jeanbarbier9448 The mesa hardware is incredible. I think over three installations, I only managed to damage one board, on the mini-lathe, with sheer ignorance, and even that was pretty easily repairable. The Mini lathe has horrible backlash. LinuxCNC can be programmed to take up the backlash on reversing an axis. It does it well as you can, but it is never a good solution.
Love your videos mate. This is just pure information, sorry if it comes off like I'm being a dick: I probably do 500 setups a year as a machinist on a variety of CNC machines. To me, it looks like you're making all the same mistakes again. 1) Screen too small. You're going to be actually needing this screen when you're stood at the spindle. Think about how big your disance to go numbers are going to be when you're stood there. 2) You MUST have a dedicated cycle start and cycle pause button. If you want proof, just look at your previous video where the knurling tool crashed. You couldn't stop it. You've got to be able to have your index finger on cycle start, middle finger on cycle pause, and use them while looking at your workpiece. I can't impress how vital this is. My suggestion would be to have a setup like Guildermeiter NEF 520. A display up top, with soft keys on the side and bottom+keyboard(for programming and setup)and then a seperate box at hand level with jog handles, cycle start and stop, feed override, spindle override, coolant on/off and E stop.) Honestly I have no idea how you don't crash all your machines all the time without the ability to switch rapid speed and cycle start/pause without looking at the control. It must be terrifying. I'm more than happy to run through some ideas with you if you like :) What I've circled in red here is what I would call minimum viable, I would take some design advice from these two pictures imgur.com/a/4T2C7my
Hi Samuel, I really appreciate your feedback. To 1/ remember this is a pretty small machine. It only has 400mm of Z travel, so I think in practice there will be less than one step between headstock position and controller pos. But it is a good point. I'll check out what it looks like once the new screen come in. To 2/ I have had a feeling that I need what you say. On the MAHO I have learned to run programs with my fingers on those physical soft keys of the function especially feed hold. Although the use of the feed override knob cranking quickly to zero is also useful. I havent done much turning on the Schaublin yet, as it is really terrible to run with the keyboard as you have identified. That was only ever a temp thing. Thanks, I'll check out that link and get back to you.
Any chance you'd share your wiring diagram from Qelectrotech? Im building a CNC mill at the moment and trying to get my head around best practices (or any practices!) for laying things out cleanly and effectively. Cheers!
Hi James, I've upload it to my LinuxCNC Forum thread. forum.linuxcnc.org/26-turning/41498-schaublin-125-cnc-retrofit?start=490#268658 You need to rename from TXT to QET, as that forum only accepted certain file types. Dont learn best practices in wiring schematics from me. Mine suck ! You will see that this is very much work in progress.
I am very impressed about your progress, so that I can follow you with actions to my maschine under restauration. You are asking for ideas for the human interface. Isn*'t it a question how how to use which technology. Mouse driven or key driven or a rotary wheel with a selection button? I find your way quiet interessting. Keys allow direct access to parameter and one can set different levels of paramters in relation to the frequency of using the specific parameter. This gives you the opportunity to direct access a parameter that needs to be change. Automotive electronic made a similar journey over the last years. Single technology parameter setting by a wheel and a single key or the combination with distingt paramater setting for specific use. It is easier to learn a single technology, but the combination offers much room for improvement and single user specific layout. Keys can be places in horizontal an vertical spheres, wheels need more attention especially how the movment will take place when using right or left hand. As a role of sum I would position those parameters which will be need to get different modes of operation by just switching one or some parameters. Examples may be reversing of rotation as toggle or form feed setting by numeric key as second input etc. regards Klaus
I think the industrial machines are normally used very differently to mine. I do mainly one offs, never series production, and often code right on the machine. Typing on a vertical keyboard is not comfortable at all.
Joystick (actiual analog joystick from gamepad, not 4 buttons you show in video) is much much better than jog wheel. For Z/tailstock you can add second joystick locked to only one direction. But that's for fast travel, for actual manual machining it's better to mimick actual lathe wheels (i mean different size and positioned diagonally). With control box that high it's probably better to make new case and put encoders/buttons on the front and leave only keyboard on slated top.
This is what I plan for my small 3018. As a quick fix before building this, I will use an Arduino Pro Micro connected via USB, to simulate keyboard. Ajusting the XY position with mouse or real keyboard is boring on Candle.
@@nalinux I am getting a few recommedations to move most of the control buttons down to the apron area. I considered building the apron with real lathe wheels, but couldn't find 100PPR hollow shaft encoders for a decent price, and the project would take me ages.
@@RotarySMP I don't know what 100PPR are. I'm just an amateur with a 3018 and a minilathe :) My plan is using stupid regular analog joystick from a recycled drone remote control. Even new, there are cheap. I think I will have to modify and recompile Candle to have variable speed, but it may not be too difficult.
@@UnreasonableSteve Thanks for the info. I've often used encoders, but never payed attention to this, since in my usual application PPR is not a concern. It seems mine are 20PPR.
Looks like that’s going to be a nice fancy control panel. I don’t think I can add anything, maybe just if you give yourself the ability to make changes for future improvements…
I understand the desire to use what you have on hand, but laser cutting a new case will eliminate significant compromise and deconstruction allowing you to focus on assembly and wiring. Deconstruction will have costs in time, cutting disks, electricity, paint sanding, hole-making.
Have a look at how ThisOldTony mounted MPG wheels on his Maho. He can operate those wheels in the dark and never make a mistake which one is being turned.
Is it an option to lose the jog wheel, and install a joy stick to move both axes? That seem pretty intuitive to me. And another thing: your videos are never too long. No worries there.
@@RotarySMP Joystick on the panel, jog wheels on the pendant? Point is, if you are really in need of closely dialing in something, you probably are going to lean over the machine, especially near the spindle/chuck, anyway.
I'm using the joystick for rapids, kind sucks to traverse the whole machine with a jog mpg. also lets you get rid of the feed rate control. rapid over with the joystick, fine move with mpg.
That's a valid point in favor of joysticks. You can just hold a joystick in one position, while a wheel needs rotation as long as you want the machine to move.
Another option is a self-centering knob that controls the direction by which direction you turn it and the speed by how far you turn it. You get the functionality of a joy stick, but the feel of a knob. You would use one for X and one for Z.
@@jamiestotz2516 My shuttle contour jog wheel which I use for video editing has one of them. I dont use it. I tried setting it up for timeline zoom, but it never felt intuitive to me.
Well my experience is quite the contray. Never been happy with joystick jog on my router, as fast moves imply overshoot of target. Now (for my lathe) I stick to the weel with (preferably) 2 push-buttons, 1 for "normal" '0.01mm/detent => 1mm/turn' speed, and 1 for "fast" "1mm/detent => 10cm/turn" speed. That'is more intuitive IMO and when you release the push-button there is no risk of unexpected moves..
And as a treat, use buttons that have an LED built-in. That way you get visual feedback directly on the button.
Hi Alexander,
I have thought about that, but mostly for functions like collet closer, or tailstock engage, which don't have an indication on the screen.
@@RotarySMP You just want to hang onto your 50-Euro-Interface card...
I think all the 9 buttons you've shown like CCW/OFF/CW, Flood/OFF/Mist etc. would benefit from it. Additional bonus with LED buttons vs. mechanical switches for example is that a program can override it, but you still get the visual feedback. If anything, running a program that goes back and forth between modes is going to be a nice light show.
@@graealex I think that MESA 7i73 has some outputs, so it could probably drive button LED's. But I was going to machine the buttons to shape, and actuate standard industrial contact blocks. That is what I did on the Maho (except I 3D printed the buttons).
@@RotarySMP Industrial contact blocks usually have the option of including an LED in the center.
@@RotarySMP Most button contact block systems have light blocks that can be stacked in with the switch contacts. And for the button machining itself, even if you want to make the button bodies themselves from an opaque material, you can still drill a series of holes on the bottom of the text engraving and then fill them and the engraving with epoxy resin.
I have no clue about layout of such a thing, just leaving that comment for the almighty algorithm. Happy Easter Rotary and family
Thanks a lot.
Put the jog increment selector on the same panel as the X and Z encoders. Also make it a push-button to set Manual mode so that jogging actually works.
Hi Andy, Good points. I didn't put that button on the Maho brick pendant and keep having to walk back around the machine to change modes.
@@RotarySMP my Okuma has them both on the panel and on the pendant. If the pendant is set to anything other than 0, it overrides the controls on the panel, and lights up an indicator to let you know the pendant is in control. It's nice to have it in both places.
@@jrucker2004 Some of those nice to have features would probably be quite easy to do in LinuxCNC for an expert, but for a guy like me get kind of overwhelming.
Reusing all in one PC's is a great way to get nice touch screens.
Good idea.
Unfortunately, most of them are designed with laptop like motherboards in terms of io and power management, which is the kiss of latency death for decent Linuxcnc performance.
For your keyboard, you might consider a drawer, there are metal drawers especially for hiding your keyboard in machine control panels even including a slight angle to make it more comfortable to type on. Apparently they are a bit pricy at 200€+, but maybe you can make an affordable custom piece from all the sheet metal.
Thanks Pierre, there have been a couple of votes for that. It is a good idea.
I know a thing or two about UX design, and the mockup you showed makes a lot of sense.
One thing another commenter mentioned that I wouldn't definitely do is illuminated buttons: If you put in some effort, you can go for a "dark cockpit" style control, where only buttons that have a function at the moment light up (I know it's not 100% the same, but I like the concept).
Also, consider breaking up that 2x3 matrix a bit more, so that it's clear that those 3 keys are a group - and if you do more groups that feed into the same thing, keep them in horizontal groups as well.
On the pendant, I'd add a spindle clamp button, and spindle controls - maybe also tailstock controls.
Vincent... we have been friends for so long, and now you go and sell me out to the LED lit button mafia :(.
I am so disappointed!
Good inputs though :)
Actually I once when down that rabbit hole of the dark cockpit, Boeing style integral lighting. There are some amazing sim builders who model all that stuff.
Why would you put the collet clamp and tailstock down on the pendant? Those would seem like set up tasks.
@@RotarySMP I never said you had to use LED buttons, I'm sure you can find some nice buttons with incandescent bulbs in the scrap bin at work.
As for setup buttons - yeah, that was the intention. I'm not 100% sure if you would use them, but to me, they seem like potentially useful additions.
If in doubt, just look if some other MPGs have buttons like this, those manufacturers seem to know what they're doing.
@@vincentguttmann2231 I thought a bit more about it and you are right, that the collet clamping will probably have me holding some small part at the spindle, and it could be quite a stretch to reach up to the control panel to clamp it. They whole volume of comments have rather overwhelmed my plans and ideas a bit. I need to thick this all through more.
I was going to make all the buttons on the Maho.
Love that the algorithm led me here.
I've never designed a control, but I've used a lot and developed some strong opinions on them.
One thing I've thought of, and wished controller designers would do, is make a spaghetti diagram of the workflow on a controller. In other words, use a string to follow the sequence of button pushes for any particular function, like loading a program, starting from the middle of a program, setting tools, setting work offsets, etc. Achieving the shortest possible string would probably give you the best user experience.
But this is just an unproven idea.
Nice idea, but isnt that something for the training manual? I am not sure how I could execute this, as my interface is modal, so buttons have different function in different modes.
@@RotarySMP maybe spaghetti diagram isn't the right term. The purpose is to reduce the amount of hand movements and button pushes. For example, to load a program from a USB drive on my fanuc control, I have to go to edit mode, program hard key, folder soft key, change device, USB mem, cursor down to the program, select soft key, copy, plus, device change, CNC mem, select, paste, cursor down to the program, plus, main program. And more if I want to run that program and have the proper screen displayed. It's tedious. The hass control is a little better, but my favorite was the Heidenhain MillPlus 3.4 which has the added benefit of customizing your layout (which I never used, the default was good enough).
I'm an HCI person and I highly recommend having tactile/non-binary buttons for things. Especially, if you want to reach them "blindly". Having the keyboard in front is better for your wrists. Also I'd highly recommend a modern game controller with haptic feedback (not just rumble) for jogging/tool changing - just go and try out the PS5 controller on the PS5 demo game and you'll know what I'm talking about. It is mind blowing :)
Hi Christoph, thanks for the inputs. I need to go check one out.
Maho control box looks awesome on Schaublin, I'd use that.
Also more than enough space to fit all LinuxCNC hardware inside, and it's above the lathe, so no further space needed.
Electronic handwheel somewhere with good access, always sucks to use stationary handwheel on the other end of CNC lathe, when dialing something in.
Thanks. I want the handwheels to be movable / clampable along that front rail.
For USB ports, maybe use the round style, that saves you from having to file out a rectangle in the plate. There are even variants available with two 3.0 in a single housing for around 15 EUR.
There's also an Austrian company that sells specialized types, I recently ordered a few from them.
Good point, there are many interesting things to fit into 22 mm hole. USB, ethernet, I think I saw even mechanical timers O.O
@@DreitTheDarkDragon Good ideas, although I didn't file out the rectangular slots on the Maho panel, as it was laser cut. I think I'll get the front sheets laser cut again, as I want them to be stainless, and I hate working with the stuff. :)
@@RotarySMP Then I'd still go for the round panel-mount stuff, as it looks a lot cleaner.
@@RotarySMP Btw the company from Austria I was referring to is called "RIBU Elektronik". The dual-USB panel mount is Amazon ASIN B086Y944DX for less than 15 EUR and even has a splash-proof cover, so you don't get metal shavings or coolant in there. And of course Conrad and the likes have a lot more stuff.
@@graealex It's funny, I find the look of the slots lasered directly into the front panel to look cleaner to my eye. Different tastes there. .)
Dutch! It all makes sense now 😄 Thank you for the shared wisdom! I'm moving this summer hopefully into a place with a proper garage and room for these sort of setups.
Hi Jim, :)
You must be looking forward to that. I felt really constrained living in an apartment.
@@RotarySMP I am, and I look forward to following in your footsteps with the mini lathe and all that.
@@jimsvideos7201 I would recommend something with a more rigid bed (which means about ever lathe ever :)
Exciting times to set up a work shop!
I like it.
For the rail attached pendant: I've seen a demo for a force feedback HMI for a lathe a long time ago. That way one can use a CNC as a manual lathe. Even if you don't do that yet, it would be a nice upgrade in the future.
I wonder if LinuxCNC even supports that?
@@RotarySMP I don't have experience with it, but it seems very capable.
Theoretically, you have to put a tiny servo motor in the wheel. The wheel servo's encoder signal fed to the axis servo as target position, and the A-servo's torque is fed to the W-servo as torque target. If LinuxCNC has access to the servo drives' data, it is possible. Tuning the system can be an issue. Delays, feedback loop oscillations, SW filters, etc.
I'm pretty sure somebody tried this in the community.
I run an old Emco Concept mill 150 at the moment. It has two chunky rotary coded knobs for the mode/jog increment selection and also one dedicated to the feed override. They’re beautifully tactile sturdy and very easy to find without looking.
If I’m running a new program my hand is basically always on the feed knob and if anything looks like it might be going pear shaped I just crank it to zero. I don’t even use the pause or stop button because this gets me out of 99% of situations. With other machines it’s always much more of a panic to find the tiny buttons or switch the pendant into the right mode.
I would highly recommend looking into something like this or at least doing yourself a favour and increasing your knob size.
Thanks Nick, possibly the first time I have been critiqued online for my knob size :)
I have larger feed and jog rate overide knobs on the Maho, and you are right it is very fast to dial it back to zero,
@@RotarySMP 🤣
Yay, she now does the Hokey Cokey _and_ the schematic is up to date😊
It already was :/ I had just rewired it wrong because I'd labelled it wrong, and was too lazy to relabel it the first time.
No to the joystick! Yes to 2 encoder wheels. Imagine trying to touch off on a part with a joystick 😬 - I like the sliding pendant, and use something similar at work. I would add the e-stop on the top instead of parallel to the machine - a long with a button for coolant and buttons for the collet, and any other buttons you'd fancy having there.
Thanks for your feedback Mark. It is good to hear from people with experience on machines like this. I already picked up that I will need more buttons on that front sliding "pendant".
For my money I’d rather have an air blast button than a mist coolant. I’ve never really found a use for mist in a lathe with a full cage. The air blast though is used often.
Good idea. Especially as this machine already has air.
Hi Mark just back from an enjoyable Easter dinner with my family, hope everyone has had an enjoyable weekend as well.
In other comments you have mentioned that you usually program on the console.
given that there are 3 basic activities on any CNC , 1) program inputting,editing and selection. 2) machine setup or manual operation, 3) fully automatic operation,
It is important to avoid any egonomic decisions that would make any of these operations needlessly interfere with the other
one such decision would be keyboard operation and estop or other machine control switch fouling each other.
As you have stated , You are inclined to mount the handwheel and other controls onto the apron control ,in which case anything associated with setup /manual operation could
be mounted there . During this time the screen becomes a big DRO. tool offsetting is about the only reason that you would touch the keyboard and monitor.
During auto mode about the only things happening are cycle start/stop, feedrate override, spindle override , and coolant override, oh and hoping that the program is correct .
these functions and the e-stop could be grouped together in a third VERY accessable spot .
Hi Steve, that is a good approach to it Thanks.
We also had family over for lunch, so there was a food induced coma in the afternoon. Got nothing done :/
about the Jog wheels - I really liked the pendant idea in front of the machine, as during jogging you will probably often be in a situation where you want to watch closely how far you jog, so the position in front of the machine is super useful there.
one more thought - you could put the jogging encoder wheels not next to each other but on two sides (front and right) of the pendant - oriented as their respective axes are. this way it would almost feel like you're operating the cross slide of a traditional lathe
That is the way Schaublin did it on the 125CCN with Fanuc, the X encoder is on the side For me that might be weird, as the Boly has the cross slide hand wheel in the LH, rather than the RH.
I got a iiyama ProLite T2435MSC-B2 24" Touchscreen-Monitor works very good. Has a glass screen so better protected against metal chips. Get two encoders for X and Z for slow step and one joystick for fast stepping. The most problems are forgetting it is on fast stepping.
Thanks for the feedback on that.
I would suggest you go with the biggest screen you can physically fit in the cabinet. Almost everybody bemoans the fact that they cut costs by going to a smaller screen whereas you rarely hear people moaning that their screen is to big.
To cater for buttons down the side they can be outside the existing cabinet (a side car arrangement). Just make a facia panel that covers the full width. (spend the coin and get a stainless sheet laser cut and etched)
As for the keyboard either makeup a drawer with ball roller slides that slides under the lower control panel. Roller slide rails are quite cheap online, particularly for the size you would require.
Alternately a hinged panel that has the keyboard on the under side (when closed).
Opening the panel exposes the control panel and swings the keyboard up the right way, also lowering the keyboard to a better height.
When closed the keyboard would cover the control panel (keyboard would be upside down). You could also make it so the E-Stop is pressed when the panel is closed.
Thanks Kim, good ideas. Since Gmoccapy is a fix resolution, I dont feel the need for a huge monitor. The 17" on the Maho seems somewhat too big.
I like the slide out keyboard though.
I recommend mounting the USB ports on the bottom, any where else they eventually fill up with chips.
Good point Ormund.
Two thoughts: keep your E stop well clear of things and on the left. I reckon go separate manual jog wheels for Z and X on a lathe, but go different sizes and shapes and even positions so that they approximate cross slide and apron handwheels.
I did think about making up a box with real wheels driving the encoders, but couldnt find reasonably priced 100PPR hollow shaft encoders, and it would take me forever to do that project. I will stagger the jog wheels though.
On my lathe the hand wheels are the same size and both on the same level. It didn't take long to get used to them. I found a high count encoder (500PPR or more) makes it feel more like a manual handwheel. If you use encoders you have to have some sort of brake/friction mechanism otherwise the wheel will turn due to vibration or even gravity if it's not perfectly balanced.
@@LesNewell Thanks for that. I had already ordered a couple of they standard 100PPR hand wheels though.
I'd use two jog wheels, z on the front, x on the right. So the direction the top of the jog wheel moves is the same direction the tool moves.
That is the way Schaublin did it on the 125CCN with the Fanuc control. But that would reverse my hands. I am used to X with the LH on the Boley.
I like your videos and the way you solve problems and explain at the same time.
Can i give you a sugestion?
Put the control of the collet open/close next to XZ jog, also one increment selector for each axel next to them also.
Gut gemacht 👍🏻
Thanks for the tip. I realised afterwards that I dont need collet clamping on the panel. It is a footswitch I already set up.
I am not sure if LinuxCNC supports separate jog increments per axis, or if I would want that. It seems more predicable if both are the same.
Also you should connect a braking resistor inyo your inverter. You will be able to stop your spindle within one second and you will save a lot of time
There is an electric brake integrated into the main spindle motor. I just haven't got around to wiring it up, and integrating into the control system yet. I figured I'd use the mechanical break rather than electrical.
@@RotarySMP Electric brake doesn't have any wear, and you don't need any special wiring. Just put in a resistor, and tell the VFD to use it. It basically shorts the motor to the resistor, creating magnetic drag.
You can always combine it with the mechanical brake, which has it's own benefits. For example, servo drives with mechanical normally-closed brakes don't need any homing - although probably not relevant for your spindle, maybe even counterproductive.
@@graealex I dont really have space for the breaking resister, as some moron chopped down that control cabinet so small that it is full up :/
I guess I could put one in the next bay, I'll keep that as plan B if the mechanical brake thing doesn't work for me.
@@RotarySMP You can always put the resistor next to it, as long as the wire gauge is heavy enough.
True. How big are the braking resisters for a 3KW VFD? I need to look into it.
I use 'Front Panel Designer' for my panels
I need to look into that.
@@RotarySMP I've been using it for 6-7years now. 3d view. Clean, simple software.
Nice work so far.
I have a lathe at work I use that has jog encoders mounted to the carriage, and it also has a jog joystick mounted between the encoders. The joystick is set to a faster jog rate so far rapid moves it's quite nice. I use the joy stick nearly every time I use the lathe, as well as the jog encoders. I think it's valuable to have both options!
Thanks for the real world inputs from experience.
For things that can be G-code controlled (on my Holbrook this is only the coolant, but in your case it could include the collet, if you were to fit a bar puller) I have three-way switches for on, auto, off.
It's nice to be able to force the coolant on or off if the G-code omitted it (or has it on when it should be off)
So you intercept the signal from the UI and or it with the physical switch output before controlling the relay?
@@RotarySMP Yes, with a Mux that chooses between G-code output and fixed values.
@@andypughtube Nice. I even have some three way switches which use the standard contact blocks I design around. I can see overriding coolant being really handy. I have often made a mess because the CAM set it automatically.
I just continue to be awed by your ability to put all this together. Very impressive. And I know there are people out there even more versed in this which just makes me feel inadequate! Nice work. And the occasional aircraft tidbits are most welcome.
Thanks for the kind feedback Glenn. I also feel like I have little idea, as there are so many smart people out there. I really appreciate when they take their time to share their knowledge.
I like the overall plan. Reusing the factory box and cutting it down is a great idea but it seems like it'll still be on the big side. Maybe a flip up keyboard for when not in use? Separate encoders are a given instead of the switch. I'd be constantly moving the wrong axis.
FYI that Bridgeport series 2 I mentioned before is now almost free but finding a place for that 15K pound 7500KG beast and the cost of getting it here... Still no film to drop off in Naxos yet but I'm trying to come up with something else for you. Will be there on the 8th next month.
Yeah, the bigger the machine, the cheaper it is. At least here.
Enjoy Naxos. We are there in mid June.
Thanks for the video update, no sparks and swarf is totally forgivable. I wish I could offer some useful insights regarding the control panel layout but I'll be reading the other responses because I have no idea. Hopefully there are plenty of good suggestions inbound. - Heather
Thanks for making it back to my channel on Easter weekend Heather :)
In my opinion the keyboard definitely needs to be in front (closer to the user) of the button block. I would recommend a little palm rest if you plan or expect to spend any amount of time typing.
At work we have one machine where the keyboard is arranged between the screen and the buttons, that is really annoying to use because you have to hover your whole arm over the keyboard while typing to not press any of the buttons or stab your wrist on the key-switch.
Thanks for confirming that. It is what I suspected.
Yes, having no way to actually rest your arms/palms without pressing buttons or manipulating dials when using the keyboard is cumbersome. Keyboard always needs to be the nearest thing. And the membrane keyboards don't really suffer from accidental key presses either.
@@graealex Good points. Thanks. There was a good suggestion to put the keyboard n a retractable drawer under the buttons.
This was also my first idea :)
But typimg on the keyboard may also be boring with buttons in front of it.
Retractable may be the good solution.
@@RotarySMP Drawer is probably the best idea. You could even go for a real keyboard then, instead of the membrane one. It's just going to be a lot of work.
I have the solution for cleaning the tag plate, just remove it, build a vaporblasting cabinet, buy a big compressor and then vaporblast it. And as a bonus you can now vaporblast parts. I built mine so it also can be used for dryblasting.
... or just keep picking away at it, as I still have about another 50 videos which need content :)
I use a joystick for X and Z jogging and love it. The speed of jogging is proportional to the stick deviation. Would hate going back to wheels. Ciao, Marco.
What machine and controller type it that?
@@RotarySMP Made a video about it 8 years ago, here it is... th-cam.com/video/ixOsdKvbKlk/w-d-xo.html&ab_channel=EZshop I also use it with my lathe, obviously the buttons are mapped differently in that use case.
I'm looking forward to playing with these Mesa cards when I finally get started on my Deckel FP4A!
They are a great system. If you speak german, check out Peter, (Talla83) channel on youtube. If you dont speak german, switch on subtitles and check out Peters channel anyway, as he has the best MESA knoeledge of anyone, and shares it generously.
Also PWC, a designer from Mesa is active on the LinuxCNC forum, so you can get support direct from the source their.
I've worked a while on a Weiler C30 lathe, it had a joystick which could latch in X+/X- or Z+/Z- direction while turning, which made it feel kinda similar to a manual lathe with feed engaged. It also had two jog wheels, one for each axis. Maybe worth looking into.
Thanks for providing examples.
I like the idea of using a joystick for coarse jogging (with a rotary encoder for finer work).
Thanks for your feedback.
I liked the trouble shooting. I have no suggestions regards the control layout, however making a cabinet could be an interesting project and will let you set the angle you want for the screen and controls. Perhaps have it on a adjustable mounting?
Good idea, I will have to see if I can do that. The new screen will have a good viewing angle, but reflections could be an issue.
Hi over the ditch. I would use the Maho control unit. I would move all the control buttons onto the slop use the larger screen with no buttons next to it. It makes for a cleaner unit. Place the key board at the bottom.
Thanks for the feedback. I really like the physical soft keys though.
@@RotarySMP Also if touchscreen fails, you can still use physical keys :)
At work we used one robot pendant with wireless mouse because screen died and we just had keep machine working to finish rest of batch. It was working, but it really wasn't comfortable.
There actually were physical keys to start, stop and restart cycle + to move all axis. Problem is that to move axis you have to switch to different screen on touch panel and then back to run in automatic mode. That was one thing which wasn't present on physical key and only reason we had to use mouse at all. Also to change program, but that batch had to be running for two more weeks.
@@DreitTheDarkDragon That sounds like feeling around in the dark :) Gotta do what is needed to get the product out.
@@RotarySMP Yeah, I guess sometimes that happens everywhere. Funny thing is that I suggested to buy replacement pendant for that robot some months before it happened. We even got price offer but somehow nobody ordered it.
As a hobbyist, I understand the frustration of not knowing which way to turn the encoder. I've considered using a combination joystick and encoder as a solution. The joystick would only determine the direction, and then the encoder would move in that direction, regardless of which way you turn the encoder. However, it could build bad habits on other machines. Just thinking....
When I use the Boley, I never have to think, the, handles just turn naturally, so I would like to have the encoder rotation the same.
For the screen, I would go with the larger screen and build small extensions for the soft keys. The increased screen size is going to be more valuable.
I have the larger 17" screen on the Maho, and dont find I really need so much realestate for a machine control
@@RotarySMP that’s what I thought too when I built my CNC router. By the time the UI was completed though Icons were too small and the gcode viewer was effectively useless. I was thinking more about your eyesight though. My near vision plummeted between 45 and 50 so maybe I’m a bit biased then.
Hi. The only 2 things that I can think of adding is some more control over the spindle speed (since you have that complicated variator + vfd + gearbox setup, it might be nice to have finer control over each part) and a switch for a light (maybe adding one of those gooseneck lights to the crosslide, so that you can more easily see what's happening on the cutting edge). Other than that, looking great! :)
Thanks for reminding me. I was also thinking that it would be nice to at least have a switch from hi-auto-low on the gearbox, as it is a bear to turn the chuck by hand when in back gear.
I got two nice goose neck lights with the machine. You remind me that I need to see if I can still mount one on the original mount pad on the back of the machine, with it closer to the wall.
The two controls that I find very useful are the jog speed and jog turtle mode. I have an analog pot hooked up to the 7i73 to adjust jog speed. The turtle button effectively gives you a low range with finer adjust. I would put this on your dual encoder panel on the lathe apron. You probably want the jog step increment controls on that panel as well.
Yeah- it would be super annoying not to have the step/increment with the wheels.
Good point.
Yes the layout for the Schablin have the same as the Maho (Heidenhain control 320,530 etc) and your keyboard far from the screen between, you have to put the machine buttons , and it is NOODstop not nudestop the only Nude we have is a street in the city Utrcht tells the Dutch Monopoly game 😉👍👍
Thanks for your feedback and correction :)
Hi Mark , regarding Handwheels and Jog control , Fanuc, Meldas, and Seimens, all arrange the handwheel as you have done on the minilathe
Jog control is done with simple momentary contact joystick, panel push button ,or membrane panel button . jog speed is controlled either with
the 0-200% feed rate override or a second jog speed selector with specific mm/rev settings, just like a manual lathe and probably the prefered
approach for a toolroom style lathe.
Joysticks were discontinued a long time ago,but I suspect it had more to do with ecomomics and not really being needed in a production environment.
Perhaps ,being as Joysticks and pushbuttons are not very expensive a removable plate with either a joystick or 4 buttons could be done to find out
which you prefer the electrical connections would be identical.
FYI japanese builders have always used omron a16 series buttons ,they are very nice ,and as accessable as potato chips in canada.
alco and fuji also make identical switches.
Regarding sense of feel,that just doesn't exist in cnc, electronic leadscrew or even powerfeed on a 100 year old lathe .
What I have seen with industrial cutting tools is that the feed and speed information that they provide is absolutely true and reliable ,
if you have done your setup correctly , it will work.
This along with being vastly more stable, is why jog instead of handwheel. try aligning something with the jog control instead of the
handwheel and you will see what I mean.
what you do feel is a sense of doom if something is going wrong😱
Thanks for your detailed feedback on that. I am collecting all the various advice for a second deign iteration.
I would prefer to have the additional controls and knobs on the right side of the keyboard, also i would put the stop feed hold and single line step buttons on the right side as they are what i use most of the time when testing my gcode! Great job as always!
Thanks for the input Stefano.
Good information. I would use 2 hand wheels with 3 steps for travel speed.
Thanks, that is about my idea as well.
Look Hurco milling machines they have similar control layout.
@@akfarmboy49 Thanks Dale. I'll look some out.
@@RotarySMP I had spelling wrong
Hurco. Made in US.
Large control panel.
I heard about airplane collection in Nelson NZ. Lot of ww2 and 50’s
Vintage. He kept it private and hidden from public.
@@RotarySMP here is a Mazak control
th-cam.com/video/Bj8eP1l0aI8/w-d-xo.html
I would lean very hard towards putting all "manual control" style switches/controls on the carriage (don't even bother with the front rail)... Wheels, e stop, jog increment, even spindle control.
I don't have a CNC but I drive my lathe entirely by DRO, and having everything within easy reach of the work area is critical for me. I know that with CNC it's a bit different, but for checking, zeroing, setup, etc, I'd want as much control near the workpiece as I could get. Redundancy doesn't hurt so hopefully you can optimize your controller but also duplicate the necessary functions onto the carriage!
Thanks for your input. There have been a few comments in that direction, so I think I will explored that more closely.
@@RotarySMP I can't wait to see! I'm considering buying a small CNC router to get my feet wet in CNC, because i'd like to do a conversion on my own lathe - even the soft keys you're implementing are an eye opener for me! I'll probably get a wireless pendant to start (I'll likely go Centroid instead of LinuxCNC) and using it enough on the router should hopefully give me the experience needed to design a control panel like the ones you've done!
@@hogtownpens LinuxCNC can seem daunting with it's learning curve, but the guys on the forum are really helpful and knowledgeable. I would recommend it to anyone.
While you’re running programs I’m concerned you’ll rest your wrists on the keyboard, so perhaps put it on a drawer under the control unit? It would also tidy up the unit significantly.
Thanks Hugo. That is a really good idea. I hadn't considered that.
Is there enough space to put the screen up higher, and fit the E-Stop button and other controls below, on the vertical panel, rather than on the sloped base? If needed, cut down the sloped console to the lower the keyboard, or use a drawer, as others have suggested. That would also allow reducing the depth of the sloped console, so you aren't reaching out too far, to get to the buttons and touchscreen. The sloped console then only needs to be deep enough for the keyboard, with a little clearance vs. the buttons above, and perhaps something of a palmrest. Definitely like the sliding pendant idea!
Thanks for the ideas Mike. I will need to do a CAD (Cardboard aided design) mock up and see.
Grate video! looks like a nice layout
Thanks a lot Luke.
I would recommend a 24" DELL P2418HT Touchscreen Monitor we have them on our CNC wood saws and other machines and i think it's the right size.
Do you have those monitors as stand alone, or integrated in to a control panel? I really dont have space for such a large monitor in a control panel.
@@RotarySMP There are more or less standalone
Oh and they are HP
drive.google.com/file/d/1DCzpnbDZqSOTm9Y0bO0zH_L7hlln-3G-/view?usp=drivesdk
Always, always go with the keyboard layout you're used to (usually qwerty), you wont realize how much a non-standard key layout will slow you down and make you crazy.
Yeah true. I have used the german QUERTZ my whole working life, and english keyboards drive me nuts. I can never find the special symbols. At least I dont need them for CNC.
The pendants would be nice if you can put then in the front, then they feel more like in a real lathe.
was to quick your solution is nice
Thanks Patrick.
Tool selector "button": four+ way rotary selector, as per your setup.
Thanks for your feedback.
Try looking up DMG's old Millplus control, and check out the layout
Thanks for the tip. Will do.
From experience ... you probably want the jog increment control right next to the jog wheels themselves. Going back and forth between the two will get annoying otherwise.
Good point Peter. Andy Pugh also had the same observation.
Just use the cabinet as-is. Use the monitor you already have-- remember you said you wanted a bigger screen... As for the softkeys, just add buttons around the monitor that "hang out" on the outside of the cabinet!
Argh... that sounds horrible :)
Needs dark mode, dark gray background white letters, like that keyboard at the end.
Our kiwi here really isn't a software guy, but you're right, dark mode would look a lot nicer.
@@graealex I dont really like dark mode screens. I like the lighter look of Gmocapy for sure.
@@RotarySMP Yes, I heard of savages that prefer light mode, but I've never seen one in the wild. What other barbaric traditions do you frown upon? Using analog 0-10V signals to drive a VFD? Admit it!
@@RotarySMP 👍
I think for tool room work a jog button and jog lock button is really useful for each axis. It gives an equivalent function as engaging the power feed on a manual machine. Some tool room machines will link the X and Z axis to machine tapers manually without a program using jog buttons, but I don't think that's very useful.
Gmoccpy locks out jogging when you are not on the jogging page, so if you bump a jog encoder while on the MDI or program page the input will be ignored.
The Schaublin 125CCN with the Fanuc control can also link axis and drive along a taper with a single jog button, but I will be implementing Andy Pughs Lathe Makros, so cutting a taper will be easy enough in MDI.
Möglicherweise wäre die angeschrägte Oberseite für alle Hardware Knöpfe gut und nach vorne einen Schubladen Auszug um die Tastatur zu verstauen. Die Schaublin Kiste wäre groß genug um so etwas zu machen.
Die Idee mit einem Schubladen für die Tastatur ist eine gute. Danke.
Could you tell me what software you're using for your wiring diagram?
It is the opensource QELECTROTECH. seems very powerful. I am just a newbie.
I think either a pair of 8 jog buttons or a joy stick would be grate for quick and dirty turning „manual style“ and the 45 deg movements for easy Chamfers. A joystick with guided movements is ideal I would guess, so you don’t have to worry about engaging the wrong feed. On the Deckel FP‘s it’s sometimes a bit flimsy 😂 on some Euchner 8 way joysticks there is a push button on top, maybe could be nice for feed override.
Thanks for the feedback on that Lorenz. Do those machines also have jog encoders?
@@RotarySMP I mean the manual Deckels 👍🏻 on my Fp1 you can engage the Z and the X feed at the same time (actually not to sure if there was a part preventing this, which was removed)
I don't know a thing about CNC-machines, so I can't halp you with this.
But I enjoyed the video and did hit the "like" button. I hope that helps somehow :)
Sures does thanks Rusty.
good video rotary smp
Thanks again.
Have you tried a sharpened wooden skewer in a cordless drill with lapping paste to clean the nameplate? I’m developing a twitch when I see clips of it. Love the content as always. Cheers man.
Oh, you mean like a tooth pick, Wouldnt it be more accurate to CNC that taper :)
Thanks for watching.
Hi Mark I assume that you have seen the early version of the 125 ccn where the handwheels, jog controls etc are mounted on the apron?
same as you were talking about in the video.
Hi Steve,
Yeah a mate here in Vienna has an early CCN like that. It is sort of what I was modelling those thought on.
Hi interesting as always, look forward to another episode.
Thanks for the feedback
hello very nice video as always can someone give me the name of the controller for milling machine? because I don't understand English very well.
I am using the Gmoccapy interface for LinuxCNC.
linuxcnc.org/docs/html/gui/gmoccapy.html
I guarantee you crash machines less with a dedicated feed hold button! I use my feed hold button all the time and never hit the e-stop on our Haas. Even in an "emergency" I hit feed hold and reset.
Thanks Nathaniel. I think I will put a dedicated feed hold, and also make the button uniquely shaped.
I type this as i watch the video so if you mention this before the end forgive me.
With the feedrate override switch instead of a pot, go a twist switch with resistors for adjusting the value ie 0% 1% 5% 10% 25% 50% 75% 90% 95% 100% 105% and so on.
This is how the big machines do it and its so hard to switch between the two plus you learn the spot and its amazing
I did the overrides with encoders rather than pot's on the Maho, and it provides very similar steps to that. I was going to do the same again.
@@RotarySMP I think the benefit he is describing is that a twist-switch would have a proper mechanical stop, while an encoder is just relative. So with a twist switch, you could actually put some proper markings on the panel, and be able to visually confirm the current setting.
You could use a binary twist switch, and use only a few input lines to encode many states. 3 lines = 8 states, 4 lines = 16 states, 5 lines = 32 states, etc.
The advantage of a rotary switch instead of an encoder is that you can set it to a specific position by feel. That way you can keep your eyes on the machine while you have your hand on the feed override. Being able to quickly dial it down and then up again is a nice safety thing, to prevent crashes, for new programs. The important thing is to be able to do out by feel.
@JohnSL - Random Products exactly this. Keeps eyes on machine not on the screen
Can you provide a link to the keyboard you used (the black one) or at least a model number!
thanks
It is an old one I got of the german classifieds.
NSI P/N KSM103B0001-W-Mc1
What software are you using at 18:51? And is it worth doing the cabinet in CAD and sending it out to be laser or waterjet cut?
Software is Qelectrotech
@@graealex Thanks!
I drew up the Maho cabinet in CAD and had it lasered. That is the choice, modify what I have or start from scratch?
@@RotarySMP Start from scratch. More CAD practice, get it exactly the way you want.
@@JCWren That is a good point. Also dont have to dick around with paint removal.
a very importan choice making a useful control panel, i think make a new one is better that modifying the old one also a nice thing to use is buttons whit led so you have more feedback
I was going to use contact block, behind home made buttons. I thing LED feedback might be nice on some functions though.
@@RotarySMP good idea
What about rotary controls with haptic feedback (via BLDC motors on the control knob)?
It is a nice idea, but were do you get your input for force required from? Servo current is more speed and inertia dependent than cutting force dependent.
@@RotarySMP I am not sure all the cool things one might use it for, but they let you do things like program the acceleration of the knob, so e.g. if you turn the knob "harder" (because you programmed resistance to increase past the 3 o'clock position) the jog goes faster. Or programmable haptic "detents." Or input from a current sensor on the spindle motor for operations where you risk stalling the spindle.
There are a lot of possibilities and BLDC motors as haptic input controls are a relatively new idea that is relatively underexplored. Lot of space for creative innovation.
@@RotarySMP Also check out th-cam.com/video/ip641WmY4pA/w-d-xo.html for some ideas in this space!
What now? Measure before cutting? Is that what I heard between the lines?
I heard that it some do it that way. I was curious :)
What about some lighting up buttons for mode selecting so you can out of the corner of your eyes quickly see what mode it in ( mdi. Check ( where everything is running at 25% and g-code can be changed well running. Like on a citizen lathe that I use to work on and loved that feature) and auto ) that's my idea for the controller
And a light on off button on the side if you also put some lights on the machine
Since this will never be a production machine, I doubt I would ever need the high speed side of it. I am trying to practice dialing back the feed rate and living with the hand on the feed hold with untested programs.
@@RotarySMP yeah but that's not as good for the tool life. since insert tools are ment/designed for a serten feed pr. rpm and temp at the cutting edge (830 C in what's called the flow zone. where the insert is still at full hard/stiffness but the material gets softer and gets pushed away/seperated from the main stock and forms the chips
so at least make sure to keep the feed pr. rpm constant
and was just personal recomendation from my experience. and what i had put on if it was me cnc
but it's your cnc mate and the most important point is that your happy with how the machine turns out
@@flikflak24 I appreciate your inputs on this. As you can see I am not a machinist, so my ideas are pretty half baked.
Some of the best ideas presented in this thread (and there are a lot), may be beyond my ablity to implement or code in LinuxCNC.
@@RotarySMP that's also fair. i am however glad to finally see the idear i came up with inteams of air ( if form of mist ) that i came with some time ago finnaly come into excistance
by the way are you planing on useing coolant or just straight oil ?
the oil is only slightly worse at cooling ( in my experience ) but never needs to be changed at all. it lubs the tools way better so they last way longer ( we whend from about 800 parts pr conrer to 3500 parts pr corner of the insert ( same feed.speed and deapth of cut )) and only needs topping up from the oil that splash out of the machine/part/ships. nothing eraperates away over time ( at least not compared to coolent ) and it also constantly oil the ways and chuck. so need to oil them by hand or any chance of rust
I would seriously consider DUAL jog wheels, vertically positioned with the lower one for lathe Z and the upper one for lathe X.
That would be fast and intiutive because it would be like the 2 handwheels on a manual lathe. In some cases you could even drive it like a manual lathe using the two jogwheels, like for simple facing tasks without needing to run a program...
That is what I showed n the video right? Except I put them next to each other. Probably more intuitive if I place then diagonally. X FWD Higher, and Z lower back, as that would be the same as my Boley.
@@RotarySMP ... whoops sorry to confess I only watched the first half of your vid then was called away to do other stuff. I'm a dufuss! 😁
I like Tom S suggestion the best, put the 2 dials ON the machine itself, and possibly a couple other jog-critical controls like a feed rate switch and reversing switches. All in a place where your hands instinctively find the controls so you dont have to take your eyes off the cutter during setup or simple turning tasks.
@@wizrom3046 That is kind of where I was going with the two encoders down in a "pendant" running on the rail at the front of the base.
I hear a lot of negative about putting dual encoders on a lathe. I would have thought that to be intuitive?
@@RotarySMP ... I cant see why they think it's a negative.
Ultimately its your baby so if you want your baby to have two heads then that is the right amount! 😎
For me... nothing gives personality to a tool / panel like having physical buttons and switches. Personally I enjoy having light indicators in / above the buttons, but that also introduces a point of failure and maintenance. Are the trade-offs for the buttons worth it? A few encoders would also be nice, however just like valves and switches you really want there to be muscle memory for the controls. So having a single function for each control or atleast for each mode is best. Also note... the four corners, and then the edges are the most valuable locations because those can easily be found without having to look over at the controls. Good luck, hope you find a mix you can enjoy... there is always rev2 if you don't like it. =D
ugg... i went down and looked at the comments there is plenty down there on the subject... sorry. atleast a nod for Tim Hunkin is nice to see.
I appreciate your comments. The one about the edges being most important is a good one.
Would the estop be better on the front (vertical) panel?seems quite far back. I’ve seen some jog dongles have jog wheels on different faces e.g so they are at 90 degrees to each other does that make it more intuitive? Interesting video 😀
Thanks. Schaublin did the two jog wheels at 90° on the 125 CCN. I should go and try them out on my mates one. Good point.
HI. Where did you purchase the nice membrane keyboard shown at about 9:20?
Hi Henk, I got it from eBay-kleinzeigen.de
I can only give you a feedback on what I'm currently designing...
My aim it to convert a small chinese lathe to a semi-automatic one with electronic gearbox, removed compound and some predefined macros for feeding and threading.
All configuration data is handled by a stationary panel with keyboard, small encoders etc, but the operational controls are on a small mobile box that can be hand-held or attached to the carriage (as mechanical ones).
I've found that I could live with a single wheel coupled to 2 (X1-X100) or 3 (X1-X10-X100) "speed selectors" pushbuttons and 2 mode selectors, one for jog, the other for manual move/macro definition, both of them providing Neutral/Z move/X move/+X+Z diagonal/+X-Z diagonal.
This way I can both handle 30° manual jog for threading (emulated compound) or angled XZ macro for tapers...
I really don't feel like using 4_joysticks unless you can lock at times one axis - which defeats the purpose of a joystick...
Nice work anyway from Maho to Schaubling - I wish I could handle such big projects...
Thanks for the feedback Jean. Just question, once you get that far along the retrofit path, it would be almost no additional work to just use LinixCNC and have a full CNC. Why the half step from manual to CNC?
@@RotarySMP I agree, that may seem strange, but I don't feel that much interested in a full-cnc lathe.
That's completely different for milling/routing.
In parallel I'm reviving my first home-made cnc-router (late 1990, 1200mm x 700mm work space, ball screws) for which I will most probably move from grbl+custom python to linuxcnc (I'm particularly interested by your experience with Mesa cards ...).
And if I'm still alive, do the same with my small lathe...
To be fair, what worries me somewhat is how (well) linuxcnc can cope with backlash on conventional acme screws (of unimpressive precision)...
Thanks for answering !
@@jeanbarbier9448 The mesa hardware is incredible. I think over three installations, I only managed to damage one board, on the mini-lathe, with sheer ignorance, and even that was pretty easily repairable.
The Mini lathe has horrible backlash. LinuxCNC can be programmed to take up the backlash on reversing an axis. It does it well as you can, but it is never a good solution.
Absolutely brilliant video
Glad you enjoyed it
Cup holder, for the tea;-)
Lucky I am not a tea drinker :)
Love your videos mate. This is just pure information, sorry if it comes off like I'm being a dick:
I probably do 500 setups a year as a machinist on a variety of CNC machines.
To me, it looks like you're making all the same mistakes again.
1) Screen too small. You're going to be actually needing this screen when you're stood at the spindle. Think about how big your disance to go numbers are going to be when you're stood there.
2) You MUST have a dedicated cycle start and cycle pause button. If you want proof, just look at your previous video where the knurling tool crashed. You couldn't stop it. You've got to be able to have your index finger on cycle start, middle finger on cycle pause, and use them while looking at your workpiece. I can't impress how vital this is.
My suggestion would be to have a setup like Guildermeiter NEF 520.
A display up top, with soft keys on the side and bottom+keyboard(for programming and setup)and then a seperate box at hand level with jog handles, cycle start and stop, feed override, spindle override, coolant on/off and E stop.)
Honestly I have no idea how you don't crash all your machines all the time without the ability to switch rapid speed and cycle start/pause without looking at the control. It must be terrifying.
I'm more than happy to run through some ideas with you if you like :) What I've circled in red here is what I would call minimum viable, I would take some design advice from these two pictures imgur.com/a/4T2C7my
Hi Samuel, I really appreciate your feedback.
To 1/ remember this is a pretty small machine. It only has 400mm of Z travel, so I think in practice there will be less than one step between headstock position and controller pos. But it is a good point. I'll check out what it looks like once the new screen come in.
To 2/ I have had a feeling that I need what you say. On the MAHO I have learned to run programs with my fingers on those physical soft keys of the function especially feed hold. Although the use of the feed override knob cranking quickly to zero is also useful.
I havent done much turning on the Schaublin yet, as it is really terrible to run with the keyboard as you have identified. That was only ever a temp thing.
Thanks, I'll check out that link and get back to you.
Any chance you'd share your wiring diagram from Qelectrotech? Im building a CNC mill at the moment and trying to get my head around best practices (or any practices!) for laying things out cleanly and effectively. Cheers!
Hi James, I've upload it to my LinuxCNC Forum thread.
forum.linuxcnc.org/26-turning/41498-schaublin-125-cnc-retrofit?start=490#268658
You need to rename from TXT to QET, as that forum only accepted certain file types.
Dont learn best practices in wiring schematics from me. Mine suck ! You will see that this is very much work in progress.
@@RotarySMP Cheers, looks quite good so far! :) Will post about mine eventually.
I am very impressed about your progress, so that I can follow you with actions to my maschine under restauration. You are asking for ideas for the human interface. Isn*'t it a question how how to use which technology. Mouse driven or key driven or a rotary wheel with a selection button?
I find your way quiet interessting. Keys allow direct access to parameter and one can set different levels of paramters in relation to the frequency of using the specific parameter. This gives you the opportunity to direct access a parameter that needs to be change. Automotive electronic made a similar journey over the last years. Single technology parameter setting by a wheel and a single key or the combination with distingt paramater setting for specific use. It is easier to learn a single technology, but the combination offers much room for improvement and single user specific layout. Keys can be places in horizontal an vertical spheres, wheels need more attention especially how the movment will take place when using right or left hand. As a role of sum I would position those parameters which will be need to get different modes of operation by just switching one or some parameters. Examples may be reversing of rotation as toggle or form feed setting by numeric key as second input etc.
regards
Klaus
Thanks for sharing your ideas on this Klaus.
Add a button for Air Blast next to the buttons for Flood and Mist.
Thanks for your input Tim.
Why not to use flat vertical module with screen, keyboard etc like big industrial machining centers?
I think the industrial machines are normally used very differently to mine. I do mainly one offs, never series production, and often code right on the machine. Typing on a vertical keyboard is not comfortable at all.
@@RotarySMP than it could be done with flip down keybord. Vertical screen and jog wheel might be more comfy than 45 degree. But i might be wrong.
Good points.
Joystick (actiual analog joystick from gamepad, not 4 buttons you show in video) is much much better than jog wheel. For Z/tailstock you can add second joystick locked to only one direction.
But that's for fast travel, for actual manual machining it's better to mimick actual lathe wheels (i mean different size and positioned diagonally).
With control box that high it's probably better to make new case and put encoders/buttons on the front and leave only keyboard on slated top.
This is what I plan for my small 3018.
As a quick fix before building this, I will use an Arduino Pro Micro connected via USB, to simulate keyboard.
Ajusting the XY position with mouse or real keyboard is boring on Candle.
@@nalinux I am getting a few recommedations to move most of the control buttons down to the apron area. I considered building the apron with real lathe wheels, but couldn't find 100PPR hollow shaft encoders for a decent price, and the project would take me ages.
@@RotarySMP I don't know what 100PPR are. I'm just an amateur with a 3018 and a minilathe :)
My plan is using stupid regular analog joystick from a recycled drone remote control.
Even new, there are cheap.
I think I will have to modify and recompile Candle to have variable speed, but it may not be too difficult.
@@nalinux 100 PPR is a spec for rotary encoders meaning 100 pulses per revolution. It's basically the resolution of the encoder
@@UnreasonableSteve Thanks for the info.
I've often used encoders, but never payed attention to this, since in my usual application PPR is not a concern.
It seems mine are 20PPR.
Looks like that’s going to be a nice fancy control panel. I don’t think I can add anything, maybe just if you give yourself the ability to make changes for future improvements…
That is a good point. Although I tend to be very resistant to change something one I have done it, even if I did a crap job :/
@@RotarySMP I was thinking the exact same thing, if I left something off, I’d probably live with it and just be annoyed…
I understand the desire to use what you have on hand, but laser cutting a new case will eliminate significant compromise and deconstruction allowing you to focus on assembly and wiring. Deconstruction will have costs in time, cutting disks, electricity, paint sanding, hole-making.
Yeah, especially the mess of paint stripping. Cutting disk = sparks. Everyone on YT likes sparks.
You said you always forget the direction of the jog wheel, the joystick should resolve that problem.
I was going to set up dual jog wheels with the same roatation as the manual Boley lathe I have. There it is intuitive.
Have a look at how ThisOldTony mounted MPG wheels on his Maho.
He can operate those wheels in the dark and never make a mistake which one is being turned.
Yeah, I remember that build :) Schaublin did it similarly on the later model 125CCN.
You have to make a teach pendant to jog. Do not install the jog encoders into the panel. If you dont you will regret
That is my idea for the jog wheels on the front of the stand. I wont put them up on the control panel.
18:43 Could someone tell me which tool is used for the wiring diagram? Thank you!
I added a link to the description.
I'd personally make it exactly like Heidenhain turnplus IT, or fanuc T control.
Thanks for the tips. I will look them up for ideas.
@@RotarySMP I really like the layout of heidenhain, but the amazing simplicity and ease of an old fanuc is attractive too.
@@max_eley Thanks Max.
Which one is that soft for wiring diagrams ???
Qelectrotech. It is an open source SW. I am a real beginner with it, but it seems powerful.
Is it an option to lose the jog wheel, and install a joy stick to move both axes? That seem pretty intuitive to me.
And another thing: your videos are never too long. No worries there.
and pots or encoders for controlling the jog speed that is how it was done by Fanuc in their controllers
Hi Willem, I find the two hand wheel two axis concept very intuitive, at least with my manual lathe. I felt the opposite with a joystick.
@@RotarySMP Joystick on the panel, jog wheels on the pendant? Point is, if you are really in need of closely dialing in something, you probably are going to lean over the machine, especially near the spindle/chuck, anyway.
@@graealex That is what I thought, and there I'd want encoders for the click, click microns or hundredths. I am not sure when I would use a joystick.
I'm using the joystick for rapids, kind sucks to traverse the whole machine with a jog mpg. also lets you get rid of the feed rate control. rapid over with the joystick, fine move with mpg.
That's a valid point in favor of joysticks. You can just hold a joystick in one position, while a wheel needs rotation as long as you want the machine to move.
Another option is a self-centering knob that controls the direction by which direction you turn it and the speed by how far you turn it. You get the functionality of a joy stick, but the feel of a knob. You would use one for X and one for Z.
@@jamiestotz2516 My shuttle contour jog wheel which I use for video editing has one of them. I dont use it. I tried setting it up for timeline zoom, but it never felt intuitive to me.
Well my experience is quite the contray.
Never been happy with joystick jog on my router, as fast moves imply overshoot of target.
Now (for my lathe) I stick to the weel with (preferably) 2 push-buttons, 1 for "normal" '0.01mm/detent => 1mm/turn' speed, and 1 for "fast" "1mm/detent => 10cm/turn" speed.
That'is more intuitive IMO and when you release the push-button there is no risk of unexpected moves..
@@jeanbarbier9448 That is why I will use dedicated encoder wheels.