The BEST description of "lead screws and threads" I have EVER heard/seen in my 87 yrs on this planet. You are truly a genius. WOW! Can't wait to see each new episode of what was once a mystery to me. And I will think hard and long about purchasing your kit. This is awesome. I thank Jesus for giving you these "gifts" that seems to come so easy. WOW!
@Barry Manilowa I happen to know, on good authority, that you just made 2 false assumptions against this man. What happened to the "judge not" that you people like to use against Christians? Or is that only okay when the heathen does it? Hypocrite!
@@DixieDee And where does the "judge-not" enter into the equation. There is a huge difference between "judgement" and "assessment!" A difficult concept for the cognitively challenged....I have that on good authority! The man made an "assessment" based upon his observation of the presenter. You seem to be cognitively confused, when you pose a question like " or is that only ok when the heathen does it?" You are directing the question at yourself, as Pat Dee is obviously a man of faith.....you....faith in yourself. Everyone, with a modicum of common sense, would assess that as being "pride-filled" HYPOCRITE?????? - that's where you are obviously casting judgement! Move along. You'd do well on the Disney Channel.
@Barry Manilowa Can you please be less rude to this person who was just trying to complement the author on his hard work that went into this project? If you wouldn't say it to his face on national television, don't hide behind the anonymity of the internet to insult the man.
With a little more programming, you could have an electronic version of the threading setup on a Hardinge HLV. Then you would not need to worry about half-nuts and threading dials. No difference in technique between inch and metric, and high speed threading into shoulders. Have the controller keep track of the Z axis position, and set a "thread start" and "thread stop" position. Then it could automatically disengage the lead screw drive at the "thread stop." You would move the tool out of the cut and push a button, and the controller would rapid the tool back to the starting point. Then you would move the tool back in, push the button again, and the controller would re-engage the screw at the same angular relationship to the spindle.
For all the videos I've watched on using a lathe, this is the one that actually explained how threading works in way I actually understood. It kind of makes the entire project make a lot more sense to me as well.
I really liked how you described how the threading dial worked with the lead screw. For me, I think the most salient point was not to disengage the half nut when cutting hybrid threads. I'm a software person with a recent fascination with machinist tools. While most of the videos I've been watching are "what happened" instead of "how to", of the "how to" videos I've seen, I don't remember anyone getting deeper than "Engage the half nut when this dial points to this number. The number depends on what thread you're cutting." And almost nothing on cutting metric on an imperial lead screw.
Living in the UK, I now have a metric lathe with a 3mm lead screw pitch but over here we live with the legacy of British Standard Imperial measurements and threads. Being of early 60's vintage and therefore brought up in the era of "UK metrification" during the 70's, I flip between these two measurement systems with ease. However, when cutting Imperial threads on my lathe which I do more than you would think, I've often wondered about disengaging the half-nuts and re-starting on the same thread indicator number instead of reversing the lathe but have been too "chicken" as you call say... After watching this video, now I know "chicken" is good! Awesome channel, love your presentation and insight behind everything, including the foibles!
I suppose the only challenge left for the control now is to cut multi starts. That has always been a giant pain on a metric lathe trying to cut imperial. Never tried it the other way but if you could divide the encoder pulse count per revolution it would be possible. Even a 3 start as you are running the spindle encoder faster than the spindle. Even if they are running in sync a pulse either way wont matter considering the resolution you have with the 4000 pulse encoder. It could be an excellent selling point for those in the US doing stuff like gunsmithing. Also a great explanation for lead screw use. Cheers
I haven't spent much time thinking about it. For many threads, it's just a matter of setting the absolute thread pitch and making multiple passes with the halfnut engaged at different points. But the halfnut isn't really designed for this in most cases. It might be easier to leave the halfnut engaged and use software, as you suggest. I still haven't tackled that problem.
@@Clough42 The answer would be to start the threading on a different pulse count than the first thread. Would just need a little selector switch for "how many starts you want".
Multi start threads are not that difficult. Never disengage the half nut, the same as with any non native thread. Keep the compound slide parallel to the work. Cut the first start with the compound on zero and advance the compound by roughly half the infeed until the desired DOC is reached. Advance the compound by the thread pitch/number of starts, repeat till finished. If you get the compound advances correct and use the same infeed for all threads then the threads will be correctly indexed. An angled compound will also work with the appropriate sin cosine correction.
@@ValExperimenter Yes of course. Set the compound at 90 and divide the lead by starts. The idea was that a very accurate version can be done with the spindle encoder that's all. It was a convenience thing. The main thing I was trying to get to is if the spindle had the option of an index. Cheers
If you have another input for a quadrature-encoder on the microcontroller, you could have another encoder on the threading dial and a microswitch that detects if the half-nut is engaged. Then you can unlock it, and move around and your microcontroller keeps track of essentially how far we got and can calculate, knowing the pitch of the lead screw. Then when you re-engage, you can do a little adjustment move with the stepper motor to be back at the multiple of whatever thread-pitch you're doing, and we're good to go from there. Advantage also: not only for metric but for any pitch you go no worry about looking at the threading dial needed anymore as everything will be handled automatically. Disadvantage: having to mount an encoder at the threading dial where it is much more exposed to coolant, oil and general machining crud.
Love the way you got into detail on your other vids in this series about the firmware. Usually people say hey look I just programed this oled display and you scratch your head and think howd they do that. No head scratching with these detailed videos. Nice combination of firmware and hardware.
Thanks! I'm trying to document everything in a way that would allow someone else to follow along and learn. As you point out, there are a ton of videos showing "what I did" and far fewer showing "how I did it" or "how you can do it". I'm attempting to fill that niche.
There are videos out there of people making their own lathes out of all manner of interesting materials--including concrete. I don't think the threading gears are the biggest challenge, though.
Hi James. As one of the metric proponents - Thank You! I, of course, have a metric leadscrew. I had thought that covering all combinations of US and metric leadscrews with US and imperial threads on the user interface might get a bit tricky, but no, you've done it. That was a good explanation of the machinations of threading dials, and congratulations on recognising that bit of the World outside the USA! Cheers. BobUK.
With this method in metric mode another simple safer method of turning right hand thread is as with reverse threads left to right but with tool upside down and motor running in revers this prevents crashing the tool in the gutter with the conventional method. Joe Pie does a great video on this.
Fantastic job. Was going to do the same thing to my lathe, and this reference is invaluable. I really appreciate the very thorough series and how you worked through it, rather than just a finished product 1 video gig. Also, this was a very great instruction/refresher on threading using a lathe - better than some of the others I've seen out there. Thank you!
That was the clearest description of why threading works and not just how. I have a 9x20 with VDC and Tumbler reverse, but the reverse sounds like it will fall apart at any minute. 16 TPI lead screw. I do a lot of plastic threading in metric sizes, So am very familiar with the change gears. Looking forward to every video in this series. Thank you! I purloined a threading chart that has all kinds of combos for change gears in both Imperial and metric. The 9x20 uses a 120/127 combo to switch from Imperial to Metric
Have you considered a "swing up threading tool"? It's made for reversing back out without pulling the cross slide back. When running forward, the tool is pressed down against a stop and cuts. When the machine is running backwards, it lifts the tool so it's not cutting, and it runs back in the cut.
I feel so lucky to be one of the 11-thou people who found out about this channel. I hope that number goes up by 10x in the coming year, because you're putting so much effort into these videos that it's super inspirational. Thank you :)
Thanks for explaining how the lead screw is used on the lathe. I'm fascinated by the project and understand everything else, but I have no hands-on experience with lathes, so I was a bit lost. Now that I understand how the lead screw is used/engaged and how each thread pass stays aligned it all makes sense. Very cool series.
I will never and should never go near a lathe. But what you are doing is very interesting. And thanks for explaining how the threads work. I've watched other channels cut threads and I never understood how they keep it straight. Thanks!
I try to avoid Metric threads unless I have to on my lathe simply because of having to change the intermediate gears and not being able to unlock the half nuts. Even with your system where the half nut needs to be left engaged, the system looks far easier to run as there is no farting around with gears. Nicely done.
Great job. Love seeing the areas where there is crossover between manual and cnc. Always makes you investigate how you as the operator interact with the machine. Thanks for the video
James, I really would like to do the same thing to my old Atlas 101.07403 lathe as you have done to yours (electronic lead screw) but I am not knowledgeable enough to do it. If, however, you do get a kit developed that would work on my lathe I surely would be interested in purchasing one from you as soon as they become available. I am really impressed with your videos as I have told you before, I have never seen anybody produce a video as good as yours. The clarity, the placement of the camera, the perfect audio and everything is just super amazing. Please keep up the outstanding work that you do. I have learned so much from you because of this and the fact that you explain everything that you do and why you did it that way and nobody can do that as well as you. Thanks James. I would love to hear the story about how you lost so much weight so fast. I need to loose about 100 pounds but so far I haven’t had enough willpower.
Thanks! As you clearly understand, video production is much harder than it looks. I've thought a lot about weight loss and what enabled my success, and I think it comes down to system thinking. I can't really just "lose weight" because that isn't something one can do. You can't just get up one morning and tell yourself "I'll spend the next hour losing weight". But I can build a system (get an app, count calories, research restaurant menus, plan meals, go to the gym, lift weights on a certain plan, get 10000 steps on my fitbit today, etc.) Those are all things I can do.
sorry to be offtopic but does anybody know of a method to log back into an instagram account?? I stupidly lost my password. I would appreciate any tricks you can give me.
@Israel Shiloh I really appreciate your reply. I found the site thru google and I'm waiting for the hacking stuff atm. Looks like it's gonna take quite some time so I will reply here later with my results.
You wanted me to let you know what I think. I think this is a fantastic project. Electronic leadscrew control has been very much in my thinking for a long time and here is a nicely designed solution. One thing I want to share with you on the rougher finish when you cut the left hand thread compared to the right hand. I frequently use a similar insert threading tool and tip to the one you use here. The threading tip is formed with a helix angle on the projecting cutting part. If the form is slanted for right hand thread as manufactured then the geometry is not as suited to cutting left hand. Under the tip is a shim seat and these are available with differing angles of side slant. I think a different shim seat could prove more suited to left hand cutting.I liked this so much I subscribed... Howard.
I wish I had seen you videos a few months ago, would of saved me a lot of work. I have just completed a very similar conversation but I also had a rotary encoder on the lead screw and a limit switch for the cross table travel. This enabled me to cut any thread either metric or imperial without the issue you have with the engage nut
Awesome! Feature creep is always rampant in diy projects. Which isn’t always a bad thing. Suggestion is when designing a board, break out every io to a pin header. Say there is 4 unused outputs, build the trace to either an existing header, or another header then if you use it for say spindle stop, the source is already built into the board, wires are easy to change on a header, not so much with pcb and layout. Keep up this awesome project!
Excellent video. Well done on outing "Metric" 😉. In the 1970s the USA was a machine tool powerhouse, and changing to metric, would have been expensive in time and equipment, but machine tools last a long time. I'm from the UK, and with the UK joining Europe, ( called the Common Market then), this was done to sell there. Best regards and thanks for sharing.
Yeah, I've always been a little confused about the US not switching to metric. There are lots of times when a short-term win is worth the long-term cost, but as you point out, machine tools last a long time.
Awesome video and a great project. I drop everything to watch when you release these videos! If cutting hybrid threads were something you were doing frequently, there is a lot you could do on the software side to streamline the process. There’s no real reason you need to reverse the lathe, other than because the software is simulating the change gears. Consider a software function where you position the tool in the gutter and press a “mark” button of some sort. As long as the half-nut is engaged, the software knows the absolute position of the tool. Now you back out the tool and activate a software function to drive the lead screw in reverse (all with the lathe running forward) until the tool is to the right of your workpiece. Now the software stops the lead screw, you reposition the tool for your cut, and then tell the software to re-engage the lead screw. It can now do this at exactly the right angle of rotation for your current thread. Now when it reaches the gutter again it can stop automatically, waiting for you to indicate that the tool is clear and it can return to the start of the cut. That would make it so much more than just a replacement for the change gears!
You can do that with some but not all thread pitches. As I understand it now it doesn't leave sync with the spindle and only adjusts rotation speed to suit pitch. To do what you want it will need positional control and then it becomes a teach lathe or an NC/CNC lathe. Cheers
@@dazaspc I don't think any thread pitch limitations apply. With this setup he has complete control of the Z axis as long as the half-nut is engaged. So it can sync up with the running chuck at the start of a new pass.
dazaspc Yes - you’re right it is more like a CNC lathe, but with manual X-axis control. As @somun points out, this setup would allow any thread because it doesn’t depend on the thread pitch of the lead screw at all - the angle encoder has, if I recall correctly, about 4000 divisions per turn, so it can start the movement at whatever angle of rotation the software instructs it to.
@@Somun-a In theory it's possible, in practice keeping count of commanded drive pulses and count spindle pulses is beyond the equipment. To maintain sync with a random disconnection event without a home pulse its impossible. Tim wanted to disengage the half nut and or reverse the lead screw. To do it read my latter explanation
@@TimFSpears This might get a bit long and would be easier to explain with a run state diagram but here it is in words As it sits now the lathe has what i call Positional Command. The motors drive has some compensation for Commanded position leading and trailing error, but it doesn't get feedback from the encoder on the spindle {only changing commanded speed} or the lead screw. This works fine until sync is lost or varying loads are introduced that go beyond the simple drive compensation. With a stepper motor this is their downfall as even though they have more torque in practice they don't monitor the position of the motors rotation so if rotation is lost to load unless the drive picks this up via a change in voltage on the input wires it has lost sync. The Hybrid servo/stepper motor in use here I believe uses this form of compensation monitoring. It is far from perfect but better than nothing. On a professional grade CNC The spindle motor will have a resolver to monitor and give feedback for commanded speed. It uses belts to transmit drive to the spindle, belts can slip so to keep up the desired spindle speed measured by the spindle encoder/resolver it will over or under speed. There is a bit more to this but this gives the idea. Most modern decent CNC can key axis movement to spindle rotation. Depending on brand and generation this is done several different ways Ill explain from the Home pulse forward. If you want to know more look up how encoders and resolvers work. For rough or non critical work CNC machines can be and are usually commanded to feed in distance/time. There is no sync required so this takes far less controlling and critical compensation. Say you have a many parts to make and these need to be more tightly controlled size and finish wise you use feeds keyed to spindle position as each part will have identical feed marks and tool loads, a screw cut with out making thread. To do this the spindle control picks up its home pulse as do the axis controls. The overall controller will send a sync command to the drives. Depending on brand/generation of controls they usually cross monitor but can just adjust the axis to stay in time with the spindle. There is no reason to keep sync between cuts here so depending on the g code it just restarts on radi and different face/axis travel. If the G code for screw cutting is used the sync is monitored and is usually started on home pulse or a given count from it, depending on thread starts command. During these keyed operations the drives will try to compensate the motors a lot harder at rates well over %100 of rated motor input current. It is not uncommon to see %400 overload for moments when something is trying to catch up. To do what you wanted in the first comment to either reverse the lead screw or break the half nut and return manually spindle sync is lost. To regain sync the Spindle and lead screw need a known position to return to. The motor for the lead screw does not have this feedback and neither does the spindle I believe so it is impossible to do it with the current configuration. Control of the spindle and lead screw is required. This is done by having key unique pulses coming from resolver/encoders on both to enable the control to pick up a known position on each and then restart with everything working together in the same relationship. I commented on the multi start thread option because it is possible with Positional Command and turning the spindle by hand. The spindle encoder would count off the required number of pulses whilst the drive for the lead screw would wait for the start signal. Once the required amount had been counted off the lead screw could be restarted in the desired sync with the spindle. To do this on the fly at speed would require equipment that would make it cost prohibitive. It is done on newer CNC equipment though. Twin spindle mills that can Rigid tap at rpms of up to 2000 rpm are out there. These have no mechanical link between motors and just use electronics to keep all in sync. 3/8 unf Rigid tapped holes on the same Z axis running the taps at 2000 rpm. These motors start and stop in sync then reverse in sync. Oh and those were thread form taps not thread cutting taps, they have even less margin for rotational error. Twin spindle lathes do something similar when they sync to swap a part between chucks at speed although this is a lot easier to achieve as they don't need to compensate to much for different loads.
Yay, metric! Even some of us Americans like it that way. :) And... super cool project. I’m binging the series right now, and if/when I eventually actually get a lathe, you can expect I’ll be back, inquiring about kit availability... or studying up to make my own equivalent.
As I have just come across this fascinating series regarding electronic control of a machine, I stand in awe of the mind of a man who thinks the way he does. I have only one comment, which has to do with user safety. Has the author ever considered using some kind of hearing protector, as you only get issued with two ears, and the machine noise must be in the higher decibel range?
Do you realize that you can do the lead screw timing in your micro controller without reversing the spindle or unlocking the half nut? Use a spare button, start your thread. At the end of the cut, push the button and have the lead screw stop. Back out the cutter, then press the button again and have your lead screw reverse. Press the button a third time to have the lead screw stop again. Set the cutting depth and press the button again. Now here is where the magic happens. Don't have the lead screw start again until the math lines up. Your micro has more then enough grunt to pull this off. Although, writing the code to pull that off would be more involved then just speed gearing. Awesome project by the way.
I'm consistently blown away by your instructional videos, James! This one just takes the cake. I am waiting anxiously for the next batch of PCBs, as I have bought an older 12x36 lathe with a badly crashed qcgb. I'd be driving a metric Acme leadscrew and splined feed shaft. Anyone have any ideas about how to figure the required torque for the hybrid servo?
Brilliant video as always. Thanks for the great explanation of the leadscrew. I have a metric lathe without leadscrew dial but this helps to understand a lot.
Now you just need to add a laser distance sensor to get the position of the carriage so that you can figure out which thread the half nut is locked in on. You already have the exact orientation of the lead screw, if you had a sensor with an accuracy of 0.1% and enough range to cover the entire lathe bed then at 12 tpi that would be accurate enough to come up with the exact position of the carriage whenever the half nuts are engaged. Then you wouldn't even need to cut a gutter at the end of the threads, just add a button on the controller for an end stop and another button to "engage" the electronic lead screw again once it was synced back up to where it was. You wouldn't even need to stop the lathe, just manually come back out, disengage the half nuts, move back to any position before the start, engage the half nuts wherever, bring the cross slide back in and feed in the compound for the next cut and press "engage" to make the next pass all the way to the end stop.
I, as a software dev think that you can do so much more in software considering the hardware investments cost that this setup require. for example: -why revers the lathe when you can set the the stepper in reverse -why not setting an end point to the thread at witch the steppe stop moving (never more to worry about stopping the lathe in time) -why not setting a reverse end point where the stepper would stop moving after the restraction this features could lead to a nice workflow for threading: -set start and end point -set tpi -enjoy your life ! ; press a button to start the pass, wait for it to reach the end point, retract, press a button to back-feed, wait for it to reach reverse end point, advance depth, press a button to start the pass and so on the lock nut would never need to be touched in metric or in inch regardless of your lathe.
Would be quite nice yeah. Except for the extra buttons, mostly a software feature. I dont think he's done any acceleration code in the current version, but certainly possible. But for now i think all acceleration "management" is done via physical inertia of the spindle.
@@LunaticCharade yes, indeed, the user interface would need to be upgraded a bit (it already is i think). for the acceleration problem i completely forgot about it, but switching to a servo would solve it, i don't think there is much energy in that long screw and carriage at such low velocities anyway, the rotor of the motor must not be ignored either but yeah, my point is not so much inertia in this system compared to friction.
You could add a rotary position sensor on the threading dial and 'pause' the carriage with the half nut engaged until the correct position comes up, come to think of it that would turn into a ghetto dro then.
This would only work if the half-nut stays engaged. Without it the software can’t know how many lead-screw threads you’ve skipped, and hence can’t know where the carriage is.
As others have commented, there are lots of things like this that could be added; It's very easy to turn a $200 electronic change gear project and turn it into a $2000 CNC conversion.
@@Clough42 I think the real trick would be to make it able to cut threads like a hardinge HLV. I'm not entirely sure how the hardinge does it (mechanical), but you engage the half nuts once and are then able to drive the leadscrew forward/stop/backward with a lever while keeping the correct timing for the thread and without reversing the spindle. Once you cut threads on one you never want to go back to waiting for the number to come around. Seems like this should be possible, as long as you start at the same degree of spindle rotation every time? Also, adding the ability to stop the leadscrew would let you mount a switch for auto stop. Great work btw, loving the videos. I'm totally planning on implementing this on my south bend.
Here i am 1 year later and i think this could be done without any additional hardware. There already is a counter on the main spindle, so from the moment when the lead screw is stopped, the electronic knows by about 1/10 of a degree how far of it is from reengaging the exact same thread. Or in other words, with 4096 points of measurement on that main spindle sensor, it can run 4096 different offsets of the same thread pitch and it should basically be able to repeatedly engage the same one over and over again after it moved the crossslide by a whole multiple of the pitch. Only i cannot program it to do so, most difficult might be to run a ramping-phase where the lead screw travels at a slightly different speed until it matches the exact point, and do that while the tool is still cutting air, as the stepper or servo cannot get to full speed in an instant.
Couldn't be any happier for you the way this is turning out! This is truly your forte and you could produce longer videos if you're ever so inclined. And, yeah, your guns are getting there...try to toss in some heavy days too. ;p
@@greybeard3759 That one was a top single. My current program usually starts with a heavy single and then backs off to five or six sets of 4-6 around RPE 8, say, 320. Depends on the day.
@@Clough42 I still lift in this, my 7th decade, but completely for nostalgia now. There are many joints telling me I probably should have ignored those personal bests long ago. ;)
Great project! It seems to me that your controller could give you a virtual dial gauge with a virtual 127 or 254 teeth to allow you to disengage the half-nuts. I'd use the LEDs to illustrate the indicator moving to show you when to engage the nut.
Hmmm - he'd have to write a routine to change the phase of the leadscrew relative to the spindle, depending on which thread you are going to drop the nut into. Otherwise you'd be waiting an age for the next opportunity to engage the nut. A practical solution for hybrid threading in cases where you need to open the nut (say when threading close up to a shoulder is to hit the brake at the same time you open the nut, then you can use the threading dial to go back, not just to the same number on the dial, but in the same revolution of the dial (in other words, the half nuts lock into the exact same place on the leadscrew). You may have to turn the chuck back a bit if your lathe didn't stop quickly enough.
Excellent! You have made so much progress while explaining all the steps that drove your decisions. You are well organized and smart. You obviously have a EE background, with your knowledge of scopes and circuits. I’m an optical physicist but can’t understand electrical stuff no matter how hard I try. The link for the servo motor on eBay doesn’t go through. Can you give me the specifics on the motor and driver? Thanks, Paul
Mod tip, add input from linear DRO glass scale sich that the thread dial can be disengaged and re engaged when doint metric threads with an imperial pitch leadscrew.. e.g. (input about position of apron vs rotation of leadscrew) add something like a led that lights up to show when to engage.
Great work again. But I think metric treat has some calculation issue. in this case some arduino Code and LCD screen and easy sets might solve this problem. Forexample you wanna go 1 mm pitch, controller will know how many revulution per desired pitch distance. By doing that you may also add safety issue. Forexample limiting the sipindle speed while desired pitch distance increases. Other safety issue is if servo failed and error led fired, all system could shut off. Thank you for valuable input and effort.
Now I wait to see you make the lead screw stop at the shoulder based on the counts? Maybe halfnut always on, drive carriage back with a button. Keeping track of counters will be a challenge, but it would be fancy. Also there is an urgent need for modular pitch, to make your own metric gear hob.
This design opens the doors to so many projects :) Wouldn't it make life easier simply to start on the 1 of the dial indictator? Actually it should work on all numbers as long as you always use the same one That way you always engage on the same position
love it! well done! ok you have a controller you can calculate the angle f the chuck, you can also count the turns (well number of degrees) so you can mark in memory just before the start of the thread and you can count to the end of the thead and you can electronicaly stop the tracking so you can repeat the cut automaticaly using the controler to engage at the correct angle and start point and disengage at the end cutting its own gutter or even tracking back along the thread! maybe two buttons , onr to mark start and one to mark stop?
Here in Europe we use the metric system however if I want to cut imperial threads on a lathe with metric leadscrew it is only necessary to replace some gears on the left side of the lathe. It is much easier than replacing the leadscrew. If most of the threads were imperial then I bought a lathe with Imperial leadscrew. This process is even easier if the lathe is equipped with norton gearbox, just make the combination of levers.
Thanks for the GREAT explanations ! Question : couldn't you "emulate" the halfnut with your Electronic LeadScrew too ? when "disengaged" you stop advancing the carriage, add some buttons to "fast move" it, and when reangaging you first resynchronize by how much step you've changed modulo the thread pitches /main number of turns ?
I practically was thinking the entire time: couldn't one solve the entirety by counting turns of the spindle and pausing the lead screw until the right one comes around? But also: with either a real or ghetto DRO one could automatically stop right in the gutter, couldn't one?
No need for a DRO for that. The stepper-driven leadscrew IS a DRO of its own, the controller knows exactly how much it moved you as long as you don't disengage the nut; so all you'd need to do is program a button as a "stop leadscrew now" (instead of stopping the whole lathe), retract the tool, hold another button to ask the controller to jog back to the other end, re-set the tool, then just tell the controller to start turning the leadscrew again whenever it would be appropriate to keep it in sync, and go all the way to where you stopped it the last time; rinse, repeat...
@@AttilaAsztalos I was thinking about the possibility, that backlash and missed steps might mess this up. But maybe that would turn out to not be a problem. But I tend to always search the worst case ;)
Backlash is not involved - you would obviously always start actual thread cutting in the same direction (only the return move goes the other way and that cuts nothing), which would take up backlash before touching the thread; as for lost steps, there are not supposed to be any, ever - if they do happen, that electronic leadscrew would be a failure (or taking a much too heavy cut); no amount of lost steps is "ok", but a reasonably oversized motor is simply not supposed to lose steps...
I now want to make a clockwork gear change for the indicator for lead screws. There seems no reason why you couldn't have a change gear for both cutting types if you have a multi geared transmission in your indicator. Which really only needs a single bar for moving between the two sets of gears, and possibly a change of the face.
James, once again you have done an excellent video! Now my question is: Will I be able to add 1 extra encoder to the system? One that is mounted on my thread chasing dial and input into the mathematics and produce a "count" output to a an LCD. Here is how I cut metric threads on my imperial lathe: The gearing between the imperial threaded lead screw and the spindle is 127:100 A 1.0mm pitch thread produces precicely 127 threads per 5.0 inches. The lead screw has a 4TPI thread. 5.0 inches is precicely 20 Leads-screw-threads (LST) The Thread-Chasing-Dial (TCD) counts 16 leads-screw-threads (LST) per revolution of the dial. While ever the half-nut is closed the TCD does not rotate. But as soon as the carriage gets to the end of the cut and I disengage the half-nut, the TCD starts rotating and I have to watch it, because I leave lathe spindle and lead-screw running continuously. Now the half-nut is disengaged and the dial turning (counting threads on the lead-screw) I move the cariage back to the right and the TCD rotates even faster ( i'm tripping over lead-screw threads). 127 threads cut on the work piece equals 5 inches (OK i didn't actualy cut 127 threads, I probably only cut 25, but pretend there are another 102 "virtual" threads back to the right of my workpiece. I need to move the cariage and re-engage the half-nut at 20 LST's to the right. Every full rotation of the TCD is 16 LST's. So I count one full rotation plus another quarter of a rotation and then engage the half-nut. If I overshoot the 20th LST I can "catch up" by moving the cariage back to the left at a faster rate than the lead screw is turning, the TCD then rotates the other direction and I have to keep a close count on what is happening, and hopefully engage into thread number 20 before I reach the work piece. If the lathe is turning too quick for me and I miss it I have to go another 20 LST's and pick it up on the 40th if I haven't lost count of the TCD! For a 1.5mm pitch thread, 127 threads on the workpiece is precisely 7.5 inches, and that is equals 30 LST's (7.5 x 4), 1 full turn plus seven eigths of a turn. of the TCD. Can also pick it up at 60 LST's (15 inches = 2 x 7.5) on 3 and three quarter turns of the TCD, etc (60 / 16) A 0.8mm pitch thread, 127 threads cut, equals 4 inches, or exactly one turn of the TCD, or any multiple of it. so what I need is the number of count of the thread chasing dial displayed, so I don't have to concentrate so hard. An "Electronic Thread Chasing Dial".
@@Clough42 Wouldn't it be possible to read a linear scale that is attached to the carriage to get the information about the position of the cutter in regards to the thread? This would free the leadscrew to be a pure device of moving the carriage right/left at any speed your electronic device desires, with the benefit of your device being in the know where the carriage is located in comparison to the thread that has been/will be cut on the spindle. The TPI of the leadscrew would become completely irrelevant then and only needed for servo motor sizing vs torque and speed.
@@joansparky4439 I believe this could be done. It would be complicated in software, since the backlash and flex in the system would be in the feedback loop, but certainly possible.
@@Clough42 You sure? Manually engaging the half nut when the number comes up has that very same flex in there.. backlash shouldn't matter. Works like this: - your ELS knows position of the carriage (linear scale) and rotation of spindle (rot encoder) and LS (servo drive) - 1st cut of thread, LS controlled by ELS (same flex and backlash as manual) - at end of cut, disengaging of the halfnut, manual reversing of the carriage back to before start, ELS registers this new position - engaging of halfnut, ELS controls LS rotation to match thread cutting with spindle and 2nd cut of thread is happening (same flex and backlash as manual) The only thing different between doing it manually with the thread dial and your ELS would be, that the ELS is matching the LS rotation to match the spindle rotation at any point you engage the halfnut, while if you do it manually you're hooking up the carriage to the LS which already is being hard synced to the spindle by the gears. Might need a switch to detect half nut engagement.. but that would probably only be needed for a first version as this info should be available via the linear scale from the carriage as well, as the ELS would be able to recognize that it is driving the carriage and not the user. Or after engaging the halfnut the user hits a button (halfnut engagement detector) and the ELS syncs the LS rotation to the spindle rotation to get the tread cut, so one engages the halfnut on a non-rotating LS in this case. Neither flex or backlash should be different from doing this manually. At least I can't see how that would be.. Your ELS PID-controller would then steer the carriage position/speed instead of the LS rotational speed.
Ya know, you don't have to reverse the entire lathe. The lead screw is motorized; just reverse the lead screw motor. An index on the spindle encoder can keep you synchronized, or a programmed counter that resets after 4096 counts.
You can disengage the half nut, you just have to re-engage it on the same number every time, and don't get off by a tooth. Much easier to keep it engaged and reverse the spindle.
not when cutting a metric thread on an imperial lead screw. the change gears only approximate the divisions for metric threads. you will accumulate error for every time that lead screw rotates while disengaged from the half nut. try it and see.
Hi James. Here in the UK we have a monthly publication "Model Engineers' Workshop Manual". The editor has been kind enough to include my e-mail to him about your video series on ELS in the just released volume. Issue 283, August 2019, page 18. BobUK
For BOTH situations where... 1) lathes don't have the threading dial, 2) Metric/Imperial combinations of Leadscrew/Thread or vice-versa are used, and the engaging of the half nut becomes impossible to know then the solution could be to implement a half nut engaging countdown on the ELS display together with sound beeps for the last 3 countdown numbers like.... ..., 5, 4, 3, 2=ShortSingleBeep, 1=ShortSingleBeep, 0=LongBeep -> engage the half nut!
There have been several comments about some variant of this feature. The issue is that it isn't just about the angular position of the screw, but also about the physical position of the carriage. The controller would have to know both to be able to track this. So maybe with a Z scale on the carriage it could be done. I haven't spent a lot of time thinking about it.
I mean, a 1mm pitch metric tread is just a 25.4 TPI imperial thread, and if you're doing everything electronically, that's just a simple (yet perhaps random-seeming) calculation. With the right motor/servo, you could even do away with needing to engage and disengage the half nut and work entirely based on if you're spinning the lead screw or not.
Very informative, well done. I would add an interlock to the forward reverse switch so that you cannot select the wrong direction by accident. How about adding an automatic stop when you threading towards a shoulder?
These are all good suggestions. Stopping automatically introduces the challenge of resynchronization, which so far I've been able to avoid tackling. Someday.
@@Clough42 automatic stop is exactly what I was thinking this whole video - I'm imagining basically a microswitch on a mag base. Set that so the carriage hits it when it gets to the thread "gutter". Microcontroller stops the leadscrew immediately when the switch gets hit and stats counting pulses from the spindle. You stop the lathe, then put it in reverse, the microcontroller counts back down and turns the leadscrew back on (in reverse) once the spindle counter gets back to zero. Wouldn't solve everything, but would take the stress out of stopping the lathe while metric threading and would be fairly simple code
I think, you are an universal genius. Absolutely great stuff. Programming the controller, fusion 360 etc. I started to copy your little plant, but i am not nearly as genius as you. I need a wiring diagramm. ?!? Can you help me?
Clough42 I’m so glad I came across your channel. I’ve watched all these videos multiple times. I’ve ordered the TI LaunchXL and other components. Where do we put our names on a list to buy boards and anything else that you decide to sell? Mechatronics is a hobby for me, trying to learn as much as I can as fast as I can. Is there any forum where us fans of your project can share ideas and questions specific to just this project or do we stick to asking questions here on TH-cam? By the way - in case you didn’t get the message -- you are a bloody genius! Thank you for letting us travel this journey with you.
Hi James. Thank you very much for this project...this is truly a gift for all of of the hobbyists community. I´ve been cheching the firmware you´ve posted on github but I´m a little bit lost between the different files. I am really excited with the possibility of get rid of the feed gears and I have almost all the required components, but unfortunately the controller part is the one that causes me the biggest of the doubts. Would you post for us a video about how loading the files to the controller? Thank you very much!!!
I get this comment a lot. I think I want to keep the manual lathe, just so I can turn the handwheels and make something quickly without fussing over feeds and speeds at the computer.
Looking great. Can't wait to ditch my change gears. I have a lathe/mill combo, equivalent to a Grizzly G4015Z. I'm wondering whether the software could accommodate driving the leadscrew at a selectable set speed, to provide a power feed for the mill. The change gear setup doesn't allow this but it would be very useful.
It would be massive overkill, but if you were using an encoder on the mill spindle, you could calculate and run feed-per-tooth for your cutter. I'm trying really hard not to get distracted by cool stuff like this. :)
@@Clough42 Good thinking. :) A second encoder on the mill spindle would do the job. It might be possible to find a way to drive one encoder when either mill or lathe is selected, but the clutch arrangement makes it somewhat difficult.
Any thoughts of a CSS modification? With the spindle encoder and a cheap digital scale CSS would be relatively easy to do and the surface finishes would lifted to the next level. CSS is the reason CNC turned parts always look amazing.
CSS is kind of a different animal. In this case, I'm controlling the screw based on the spindle position. CSS requires controlling the spindle RPM based on the X position of the cross slide. That would interesting, but kind of out of scope, I think. Unless I'm misunderstanding you.
If you encoder is also providing you with one-pulse-per-rev, then you could use this as a sync feature for when to start threading and you would not have to leave the half-nut engaged all the time. At least that is what I suspect :D would be cool if you could try that out :)
Would not work, because the encoder has no idea where the carriage is when you engaged the half nut. You would also need a very precise linear encoder on the carriage.
I guess that I'm missing something here, as it seems that if you can change the relative speeds of the spindle and the lead screw, as long as you engage the half-nut always on 1, you should be able to cut metric threads.
Hi, I enjoy your vids, I am new to your channel and likes the description on threading and halfnuts. One thing crossed my mind, you have lathe and mill? and buy gears, why?
I believe it's just a matter of cost vs. time... As inexpensive as aluminum timing pulleys are these days, it probably just makes more sense to purchase them. I'm stubborn and pinch a nickel probably more often than I should. I often find myself spending hours making something I could buy for $10... then I kick myself. ;)
Those aren't gears. They are pulleys. I don't think you're realizing how difficult those would be to cut yourself. Did you see the "rim" around the edge? That means you can't single-point shape/broach them. If you wanted to kill them with an end mill (it would have to be a tapered EM, or a form tool if it's not a straight taper, but in not sure what the form of a pulley is) you would have to make it wider than the belt to account for the radius of the EM when you come up against the "rim." That would introduce tracking issues and might bind up the belt. Not to mention tee time it would take.
Instead of reversing the motor could you have the controller advance the lead screw a specific distance and then reverse the lead screw leaving the main motor on? Maybe have a cycle go button that when pressed advances so far then returns, you can then increase the cut and hit go again
Yes. 1st press: record start position, start threading. 2nd press: stop moving, record end position. 3rd press: rapid to start position. 4th press: threading until end position. Goto step 3. (Plus a way to reset) Of course all threading must be synchronised to tracked spindle angle+pitch
@@ivanblogs good point, the acceleration in the reverse direction might be an issue, advancing wouldn't be a problem you could set the start point far enough back that it would be up to speed by the time it starts cutting, might still be worth a shot
@@scottkennedy7741 Yeah... I glossed over the acceleration issues. I believe the current implementation ignores speed/acceleration issues, and assumes everything is bounded by the spindle acceleration. It's not trivial to implement, but also not rocket science. (Unless you're threading a rocket...)
I am doing my best to resist the temptation to keep adding features. If you want a polished commercial product, they're out there. I'm targeting something much less expensive that uses mostly off-the-shelf parts. No matter what I do, though, the motor and driver are going to be a bit part of the cost.
Hello James. Thank you very much for your content... It's excellent as usual. Just a question... If you engage the threading dial at same position, you could move the carriage and still make any kind of threads right? I mean, you will be finding always the same grove on leading screw. So you would have a option to disengage the locking nut in that case. I'm I mistaken? Greetings from Brazil. A metric country but imperial bolts and raw materia widely available also. 😁👍
Yes. On the *exact same position*, not forward or backward any rotations. Can be handy if you need to stop the carriage in a very short distance, but your lathe spindle takes too long to spin down.
The BEST description of "lead screws and threads" I have EVER heard/seen in my 87 yrs on this planet. You are truly a genius. WOW! Can't wait to see each new episode of what was once a mystery to me. And I will think hard and long about purchasing your kit. This is awesome.
I thank Jesus for giving you these "gifts" that seems to come so easy. WOW!
@Barry Manilowa I happen to know, on good authority, that you just made 2 false assumptions against this man. What happened to the "judge not" that you people like to use against Christians? Or is that only okay when the heathen does it?
Hypocrite!
Thanks!
@Barry Manilowa well, once again, the secular sideline beaks-off about what precious little they know themselves. At least he can read & comprehend!
@@DixieDee And where does the "judge-not" enter into the equation. There is a huge difference between "judgement" and "assessment!" A difficult concept for the cognitively challenged....I have that on good authority! The man made an "assessment" based upon his observation of the presenter. You seem to be cognitively confused, when you pose a question like " or is that only ok when the heathen does it?" You are directing the question at yourself, as Pat Dee is obviously a man of faith.....you....faith in yourself. Everyone, with a modicum of common sense, would assess that as being "pride-filled" HYPOCRITE?????? - that's where you are obviously casting judgement! Move along. You'd do well on the Disney Channel.
@Barry Manilowa Can you please be less rude to this person who was just trying to complement the author on his hard work that went into this project? If you wouldn't say it to his face on national television, don't hide behind the anonymity of the internet to insult the man.
With a little more programming, you could have an electronic version of the threading setup on a Hardinge HLV. Then you would not need to worry about half-nuts and threading dials. No difference in technique between inch and metric, and high speed threading into shoulders.
Have the controller keep track of the Z axis position, and set a "thread start" and "thread stop" position. Then it could automatically disengage the lead screw drive at the "thread stop." You would move the tool out of the cut and push a button, and the controller would rapid the tool back to the starting point. Then you would move the tool back in, push the button again, and the controller would re-engage the screw at the same angular relationship to the spindle.
That seems to bay the Numobams lathes with electronic leadscrew work
For all the videos I've watched on using a lathe, this is the one that actually explained how threading works in way I actually understood. It kind of makes the entire project make a lot more sense to me as well.
Very cool. Thanks!
hell of an electronic and mechanical engineer, articulate pedagogic stud LOL. I must do some after-class study to digest it :)
Pedagogic stud, huh? Nice.
Can’t wait for the kits to be available! :D
clough42..... That brings me back to my Makerfarm days! Thank god 3d printers aren't wood anymore!
I have made a number of improvements to my printers, and I'm getting good results, so it's hard to justify upgrading.
I really liked how you described how the threading dial worked with the lead screw. For me, I think the most salient point was not to disengage the half nut when cutting hybrid threads.
I'm a software person with a recent fascination with machinist tools. While most of the videos I've been watching are "what happened" instead of "how to", of the "how to" videos I've seen, I don't remember anyone getting deeper than "Engage the half nut when this dial points to this number. The number depends on what thread you're cutting." And almost nothing on cutting metric on an imperial lead screw.
Living in the UK, I now have a metric lathe with a 3mm lead screw pitch but over here we live with the legacy of British Standard Imperial measurements and threads. Being of early 60's vintage and therefore brought up in the era of "UK metrification" during the 70's, I flip between these two measurement systems with ease. However, when cutting Imperial threads on my lathe which I do more than you would think, I've often wondered about disengaging the half-nuts and re-starting on the same thread indicator number instead of reversing the lathe but have been too "chicken" as you call say...
After watching this video, now I know "chicken" is good!
Awesome channel, love your presentation and insight behind everything, including the foibles!
Fear sometimes keeps us from new discoveries. And sometimes it keeps us alive.
In the last one I kept telling internally "Cut those 1"-3/4 down to an M6 thread to show off! " Glad you're doing it. Show off, lol.
Great demo and explanation, enjoyed. The system is working great. Thank you James.
Thanks!
I suppose the only challenge left for the control now is to cut multi starts. That has always been a giant pain on a metric lathe trying to cut imperial. Never tried it the other way but if you could divide the encoder pulse count per revolution it would be possible. Even a 3 start as you are running the spindle encoder faster than the spindle. Even if they are running in sync a pulse either way wont matter considering the resolution you have with the 4000 pulse encoder. It could be an excellent selling point for those in the US doing stuff like gunsmithing.
Also a great explanation for lead screw use.
Cheers
I haven't spent much time thinking about it. For many threads, it's just a matter of setting the absolute thread pitch and making multiple passes with the halfnut engaged at different points. But the halfnut isn't really designed for this in most cases. It might be easier to leave the halfnut engaged and use software, as you suggest. I still haven't tackled that problem.
@@Clough42 The answer would be to start the threading on a different pulse count than the first thread. Would just need a little selector switch for "how many starts you want".
@@TheWireEDM and a routine to move the slide over to the next start after each previous is done
Multi start threads are not that difficult. Never disengage the half nut, the same as with any non native thread. Keep the compound slide parallel to the work. Cut the first start with the compound on zero and advance the compound by roughly half the infeed until the desired DOC is reached. Advance the compound by the thread pitch/number of starts, repeat till finished. If you get the compound advances correct and use the same infeed for all threads then the threads will be correctly indexed. An angled compound will also work with the appropriate sin cosine correction.
@@ValExperimenter Yes of course. Set the compound at 90 and divide the lead by starts. The idea was that a very accurate version can be done with the spindle encoder that's all. It was a convenience thing. The main thing I was trying to get to is if the spindle had the option of an index.
Cheers
That is a hell of a project success!
Thank you, James. Interesting as ever. The most lucid explanation of the threading indicator I've ever heard! Regards Al (UK).
I'm glad you liked it. I spent a bunch of time trying to put together an animation to make it clearer, but ran out of time. There's too much to learn.
If you have another input for a quadrature-encoder on the microcontroller, you could have another encoder on the threading dial and a microswitch that detects if the half-nut is engaged. Then you can unlock it, and move around and your microcontroller keeps track of essentially how far we got and can calculate, knowing the pitch of the lead screw. Then when you re-engage, you can do a little adjustment move with the stepper motor to be back at the multiple of whatever thread-pitch you're doing, and we're good to go from there. Advantage also: not only for metric but for any pitch you go no worry about looking at the threading dial needed anymore as everything will be handled automatically. Disadvantage: having to mount an encoder at the threading dial where it is much more exposed to coolant, oil and general machining crud.
Most interesting thing going on in the home shop size machinist community. Looking forward to every episode.
Thanks! The summer is pretty slow in this community, but this project seems to be keeping people interested.
Love the way you got into detail on your other vids in this series about the firmware. Usually people say hey look I just programed this oled display and you scratch your head and think howd they do that. No head scratching with these detailed videos. Nice combination of firmware and hardware.
Thanks! I'm trying to document everything in a way that would allow someone else to follow along and learn. As you point out, there are a ton of videos showing "what I did" and far fewer showing "how I did it" or "how you can do it". I'm attempting to fill that niche.
This is a great project. It really makes the idea of building your own lathe an easier proposition.
Spotted the optimist...
There are videos out there of people making their own lathes out of all manner of interesting materials--including concrete. I don't think the threading gears are the biggest challenge, though.
Hi James. As one of the metric proponents - Thank You! I, of course, have a metric leadscrew. I had thought that covering all combinations of US and metric leadscrews with US and imperial threads on the user interface might get a bit tricky, but no, you've done it. That was a good explanation of the machinations of threading dials, and congratulations on recognising that bit of the World outside the USA! Cheers. BobUK.
Awesome!!!! Thanks for all of your hard work to make it possible to cut imperial or metric threads on just about any metal lathe.
With this method in metric mode another simple safer method of turning right hand thread is as with reverse threads left to right but with tool upside down and motor running in revers this prevents crashing the tool in the gutter with the conventional method. Joe Pie does a great video on this.
Can't wait til a kit is available!
Thanks James very interesting and informative video 👍
Fantastic job. Was going to do the same thing to my lathe, and this reference is invaluable. I really appreciate the very thorough series and how you worked through it, rather than just a finished product 1 video gig. Also, this was a very great instruction/refresher on threading using a lathe - better than some of the others I've seen out there. Thank you!
Thanks! That is my goal: to show the whole process.
That was the clearest description of why threading works and not just how. I have a 9x20 with VDC and Tumbler reverse, but the reverse sounds like it will fall apart at any minute. 16 TPI lead screw. I do a lot of plastic threading in metric sizes, So am very familiar with the change gears. Looking forward to every video in this series. Thank you! I purloined a threading chart that has all kinds of combos for change gears in both Imperial and metric. The 9x20 uses a 120/127 combo to switch from Imperial to Metric
The Grizzly G0602 uses the same 120/127 combo. I hope to never use it again. :)
Have you considered a "swing up threading tool"? It's made for reversing back out without pulling the cross slide back. When running forward, the tool is pressed down against a stop and cuts. When the machine is running backwards, it lifts the tool so it's not cutting, and it runs back in the cut.
I feel so lucky to be one of the 11-thou people who found out about this channel. I hope that number goes up by 10x in the coming year, because you're putting so much effort into these videos that it's super inspirational. Thank you :)
That would be cool. Thanks!
Thanks for explaining how the lead screw is used on the lathe. I'm fascinated by the project and understand everything else, but I have no hands-on experience with lathes, so I was a bit lost. Now that I understand how the lead screw is used/engaged and how each thread pass stays aligned it all makes sense. Very cool series.
Impressive!! I admire your skills.
Man, this is the most awesome DIY feature for a lathe I ever see. Congratulations!
Thank you very much!
This series is so BADASS!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
I will never and should never go near a lathe. But what you are doing is very interesting. And thanks for explaining how the threads work. I've watched other channels cut threads and I never understood how they keep it straight. Thanks!
I try to avoid Metric threads unless I have to on my lathe simply because of having to change the intermediate gears and not being able to unlock the half nuts. Even with your system where the half nut needs to be left engaged, the system looks far easier to run as there is no farting around with gears. Nicely done.
Great job. Love seeing the areas where there is crossover between manual and cnc. Always makes you investigate how you as the operator interact with the machine. Thanks for the video
You're welcome.
James, I really would like to do the same thing to my old Atlas 101.07403 lathe as you have done to yours (electronic lead screw) but I am not knowledgeable enough to do it. If, however, you do get a kit developed that would work on my lathe I surely would be interested in purchasing one from you as soon as they become available.
I am really impressed with your videos as I have told you before, I have never seen anybody produce a video as good as yours. The clarity, the placement of the camera, the perfect audio and everything is just super amazing. Please keep up the outstanding work that you do. I have learned so much from you because of this and the fact that you explain everything that you do and why you did it that way and nobody can do that as well as you. Thanks James. I would love to hear the story about how you lost so much weight so fast. I need to loose about 100 pounds but so far I haven’t had enough willpower.
Thanks! As you clearly understand, video production is much harder than it looks. I've thought a lot about weight loss and what enabled my success, and I think it comes down to system thinking. I can't really just "lose weight" because that isn't something one can do. You can't just get up one morning and tell yourself "I'll spend the next hour losing weight". But I can build a system (get an app, count calories, research restaurant menus, plan meals, go to the gym, lift weights on a certain plan, get 10000 steps on my fitbit today, etc.) Those are all things I can do.
Once again you did such an amazing job of explaining and creating such an amazing video. Thanks again James!
Great job, very impressive 😊
sorry to be offtopic but does anybody know of a method to log back into an instagram account??
I stupidly lost my password. I would appreciate any tricks you can give me.
@Santana Kameron Instablaster :)
@Israel Shiloh I really appreciate your reply. I found the site thru google and I'm waiting for the hacking stuff atm.
Looks like it's gonna take quite some time so I will reply here later with my results.
@Israel Shiloh it did the trick and I finally got access to my account again. Im so happy:D
Thank you so much, you really help me out :D
@Santana Kameron Glad I could help =)
You wanted me to let you know what I think. I think this is a fantastic project. Electronic leadscrew control has been very much in my thinking for a long time and here is a nicely designed solution. One thing I want to share with you on the rougher finish when you cut the left hand thread compared to the right hand. I frequently use a similar insert threading tool and tip to the one you use here. The threading tip is formed with a helix angle on the projecting cutting part. If the form is slanted for right hand thread as manufactured then the geometry is not as suited to cutting left hand. Under the tip is a shim seat and these are available with differing angles of side slant. I think a different shim seat could prove more suited to left hand cutting.I liked this so much I subscribed... Howard.
I wish I had seen you videos a few months ago, would of saved me a lot of work.
I have just completed a very similar conversation but I also had a rotary encoder on the lead screw and a limit switch for the cross table travel. This enabled me to cut any thread either metric or imperial without the issue you have with the engage nut
Awesome!
Feature creep is always rampant in diy projects. Which isn’t always a bad thing.
Suggestion is when designing a board, break out every io to a pin header. Say there is 4 unused outputs, build the trace to either an existing header, or another header then if you use it for say spindle stop, the source is already built into the board, wires are easy to change on a header, not so much with pcb and layout.
Keep up this awesome project!
I am the Feature Creep. :)
Excellent video.
Well done on outing "Metric" 😉. In the 1970s the USA was a machine tool powerhouse, and changing to metric, would have been expensive in time and equipment, but machine tools last a long time. I'm from the UK, and with the UK joining Europe, ( called the Common Market then), this was done to sell there.
Best regards and thanks for sharing.
Yeah, I've always been a little confused about the US not switching to metric. There are lots of times when a short-term win is worth the long-term cost, but as you point out, machine tools last a long time.
Awesome video and a great project. I drop everything to watch when you release these videos!
If cutting hybrid threads were something you were doing frequently, there is a lot you could do on the software side to streamline the process. There’s no real reason you need to reverse the lathe, other than because the software is simulating the change gears.
Consider a software function where you position the tool in the gutter and press a “mark” button of some sort. As long as the half-nut is engaged, the software knows the absolute position of the tool. Now you back out the tool and activate a software function to drive the lead screw in reverse (all with the lathe running forward) until the tool is to the right of your workpiece. Now the software stops the lead screw, you reposition the tool for your cut, and then tell the software to re-engage the lead screw. It can now do this at exactly the right angle of rotation for your current thread.
Now when it reaches the gutter again it can stop automatically, waiting for you to indicate that the tool is clear and it can return to the start of the cut. That would make it so much more than just a replacement for the change gears!
You can do that with some but not all thread pitches. As I understand it now it doesn't leave sync with the spindle and only adjusts rotation speed to suit pitch. To do what you want it will need positional control and then it becomes a teach lathe or an NC/CNC lathe.
Cheers
@@dazaspc I don't think any thread pitch limitations apply. With this setup he has complete control of the Z axis as long as the half-nut is engaged. So it can sync up with the running chuck at the start of a new pass.
dazaspc Yes - you’re right it is more like a CNC lathe, but with manual X-axis control. As @somun points out, this setup would allow any thread because it doesn’t depend on the thread pitch of the lead screw at all - the angle encoder has, if I recall correctly, about 4000 divisions per turn, so it can start the movement at whatever angle of rotation the software instructs it to.
@@Somun-a In theory it's possible, in practice keeping count of commanded drive pulses and count spindle pulses is beyond the equipment. To maintain sync with a random disconnection event without a home pulse its impossible. Tim wanted to disengage the half nut and or reverse the lead screw. To do it read my latter explanation
@@TimFSpears This might get a bit long and would be easier to explain with a run state diagram but here it is in words
As it sits now the lathe has what i call Positional Command. The motors drive has some compensation for Commanded position leading and trailing error, but it doesn't get feedback from the encoder on the spindle {only changing commanded speed} or the lead screw. This works fine until sync is lost or varying loads are introduced that go beyond the simple drive compensation. With a stepper motor this is their downfall as even though they have more torque in practice they don't monitor the position of the motors rotation so if rotation is lost to load unless the drive picks this up via a change in voltage on the input wires it has lost sync. The Hybrid servo/stepper motor in use here I believe uses this form of compensation monitoring. It is far from perfect but better than nothing.
On a professional grade CNC The spindle motor will have a resolver to monitor and give feedback for commanded speed. It uses belts to transmit drive to the spindle, belts can slip so to keep up the desired spindle speed measured by the spindle encoder/resolver it will over or under speed. There is a bit more to this but this gives the idea. Most modern decent CNC can key axis movement to spindle rotation. Depending on brand and generation this is done several different ways Ill explain from the Home pulse forward. If you want to know more look up how encoders and resolvers work. For rough or non critical work CNC machines can be and are usually commanded to feed in distance/time. There is no sync required so this takes far less controlling and critical compensation. Say you have a many parts to make and these need to be more tightly controlled size and finish wise you use feeds keyed to spindle position as each part will have identical feed marks and tool loads, a screw cut with out making thread. To do this the spindle control picks up its home pulse as do the axis controls. The overall controller will send a sync command to the drives. Depending on brand/generation of controls they usually cross monitor but can just adjust the axis to stay in time with the spindle. There is no reason to keep sync between cuts here so depending on the g code it just restarts on radi and different face/axis travel. If the G code for screw cutting is used the sync is monitored and is usually started on home pulse or a given count from it, depending on thread starts command. During these keyed operations the drives will try to compensate the motors a lot harder at rates well over %100 of rated motor input current. It is not uncommon to see %400 overload for moments when something is trying to catch up.
To do what you wanted in the first comment to either reverse the lead screw or break the half nut and return manually spindle sync is lost. To regain sync the Spindle and lead screw need a known position to return to. The motor for the lead screw does not have this feedback and neither does the spindle I believe so it is impossible to do it with the current configuration. Control of the spindle and lead screw is required. This is done by having key unique pulses coming from resolver/encoders on both to enable the control to pick up a known position on each and then restart with everything working together in the same relationship.
I commented on the multi start thread option because it is possible with Positional Command and turning the spindle by hand. The spindle encoder would count off the required number of pulses whilst the drive for the lead screw would wait for the start signal. Once the required amount had been counted off the lead screw could be restarted in the desired sync with the spindle. To do this on the fly at speed would require equipment that would make it cost prohibitive.
It is done on newer CNC equipment though. Twin spindle mills that can Rigid tap at rpms of up to 2000 rpm are out there. These have no mechanical link between motors and just use electronics to keep all in sync. 3/8 unf Rigid tapped holes on the same Z axis running the taps at 2000 rpm. These motors start and stop in sync then reverse in sync. Oh and those were thread form taps not thread cutting taps, they have even less margin for rotational error. Twin spindle lathes do something similar when they sync to swap a part between chucks at speed although this is a lot easier to achieve as they don't need to compensate to much for different loads.
Yay, metric! Even some of us Americans like it that way. :)
And... super cool project. I’m binging the series right now, and if/when I eventually actually get a lathe, you can expect I’ll be back, inquiring about kit availability... or studying up to make my own equivalent.
As I have just come across this fascinating series regarding electronic control of a machine, I stand in awe of the mind of a man who thinks the way he does.
I have only one comment, which has to do with user safety.
Has the author ever considered using some kind of hearing protector, as you only get issued with two ears, and the machine noise must be in the higher decibel range?
You mean the lathe? It's pretty quiet, unless you have a tool misbehaving or a part ringing.
Do you realize that you can do the lead screw timing in your micro controller without reversing the spindle or unlocking the half nut? Use a spare button, start your thread. At the end of the cut, push the button and have the lead screw stop. Back out the cutter, then press the button again and have your lead screw reverse. Press the button a third time to have the lead screw stop again. Set the cutting depth and press the button again. Now here is where the magic happens. Don't have the lead screw start again until the math lines up. Your micro has more then enough grunt to pull this off. Although, writing the code to pull that off would be more involved then just speed gearing. Awesome project by the way.
I'm consistently blown away by your instructional videos, James! This one just takes the cake.
I am waiting anxiously for the next batch of PCBs, as I have bought an older 12x36 lathe with a badly crashed qcgb. I'd be driving a metric Acme leadscrew and splined feed shaft. Anyone have any ideas about how to figure the required torque for the hybrid servo?
I like the way you work, great content as usual.
Thanks!
Brilliant video as always. Thanks for the great explanation of the leadscrew. I have a metric lathe without leadscrew dial but this helps to understand a lot.
Really enjoying these videos on the electronic lead screw project - now I’m thinking about converting my Lathe! John🇬🇧
Cool! If you do, be sure to share.
This project is advancing very nicely. I noticed that it appears time to put that controller in a proper case so you can label the buttons! :)
Yes. So much to do! :)
Now you just need to add a laser distance sensor to get the position of the carriage so that you can figure out which thread the half nut is locked in on. You already have the exact orientation of the lead screw, if you had a sensor with an accuracy of 0.1% and enough range to cover the entire lathe bed then at 12 tpi that would be accurate enough to come up with the exact position of the carriage whenever the half nuts are engaged. Then you wouldn't even need to cut a gutter at the end of the threads, just add a button on the controller for an end stop and another button to "engage" the electronic lead screw again once it was synced back up to where it was. You wouldn't even need to stop the lathe, just manually come back out, disengage the half nuts, move back to any position before the start, engage the half nuts wherever, bring the cross slide back in and feed in the compound for the next cut and press "engage" to make the next pass all the way to the end stop.
Again a great explanation. Also with the function of the dial. Only missing: picture in picture while turning the threads. Thx ;)
Good idea. Now I'll need a fifth camera....
Dudes in shape
I, as a software dev think that you can do so much more in software considering the hardware investments cost that this setup require.
for example:
-why revers the lathe when you can set the the stepper in reverse
-why not setting an end point to the thread at witch the steppe stop moving (never more to worry about stopping the lathe in time)
-why not setting a reverse end point where the stepper would stop moving after the restraction
this features could lead to a nice workflow for threading:
-set start and end point
-set tpi
-enjoy your life ! ; press a button to start the pass, wait for it to reach the end point, retract, press a button to back-feed, wait for it to reach reverse end point, advance depth, press a button to start the pass and so on
the lock nut would never need to be touched in metric or in inch regardless of your lathe.
Would be quite nice yeah. Except for the extra buttons, mostly a software feature. I dont think he's done any acceleration code in the current version, but certainly possible. But for now i think all acceleration "management" is done via physical inertia of the spindle.
@@LunaticCharade yes, indeed, the user interface would need to be upgraded a bit (it already is i think).
for the acceleration problem i completely forgot about it, but switching to a servo would solve it, i don't think there is much energy in that long screw and carriage at such low velocities anyway, the rotor of the motor must not be ignored either but yeah, my point is not so much inertia in this system compared to friction.
You could add a rotary position sensor on the threading dial and 'pause' the carriage with the half nut engaged until the correct position comes up, come to think of it that would turn into a ghetto dro then.
This would only work if the half-nut stays engaged. Without it the software can’t know how many lead-screw threads you’ve skipped, and hence can’t know where the carriage is.
I thought that too. It's not much further off just adding servos to everything and turn it full CNC :P
As others have commented, there are lots of things like this that could be added; It's very easy to turn a $200 electronic change gear project and turn it into a $2000 CNC conversion.
@@Clough42 I think the real trick would be to make it able to cut threads like a hardinge HLV. I'm not entirely sure how the hardinge does it (mechanical), but you engage the half nuts once and are then able to drive the leadscrew forward/stop/backward with a lever while keeping the correct timing for the thread and without reversing the spindle. Once you cut threads on one you never want to go back to waiting for the number to come around.
Seems like this should be possible, as long as you start at the same degree of spindle rotation every time? Also, adding the ability to stop the leadscrew would let you mount a switch for auto stop.
Great work btw, loving the videos. I'm totally planning on implementing this on my south bend.
Here i am 1 year later and i think this could be done without any additional hardware. There already is a counter on the main spindle, so from the moment when the lead screw is stopped, the electronic knows by about 1/10 of a degree how far of it is from reengaging the exact same thread. Or in other words, with 4096 points of measurement on that main spindle sensor, it can run 4096 different offsets of the same thread pitch and it should basically be able to repeatedly engage the same one over and over again after it moved the crossslide by a whole multiple of the pitch. Only i cannot program it to do so, most difficult might be to run a ramping-phase where the lead screw travels at a slightly different speed until it matches the exact point, and do that while the tool is still cutting air, as the stepper or servo cannot get to full speed in an instant.
Excellent and comprehensive explanation, thanks. And congrats on a project well engineered!
Thank you
Fascinating series of videos, well-presented and thought out. Interested in anything more on this, and any kit materials when available. Thank you.
Couldn't be any happier for you the way this is turning out!
This is truly your forte and you could produce longer videos if you're ever so inclined.
And, yeah, your guns are getting there...try to toss in some heavy days too. ;p
Thanks! The 380lb deadlift I pulled the other day felt kind of heavy, but maybe I was just tired. :) I'm a work in progress.
@@Clough42 Reps? :)
@@greybeard3759 That one was a top single. My current program usually starts with a heavy single and then backs off to five or six sets of 4-6 around RPE 8, say, 320. Depends on the day.
@@Clough42 I still lift in this, my 7th decade, but completely for nostalgia now.
There are many joints telling me I probably should have ignored those personal bests long ago. ;)
Great video. Really good explanation of the threading dial and associated maths! Keep 'em coming!
Great explanation of how the TDI works Cheers
Excellent expained and excecuted! Congrats on this modification!
Thanks!
Great project! It seems to me that your controller could give you a virtual dial gauge with a virtual 127 or 254 teeth to allow you to disengage the half-nuts. I'd use the LEDs to illustrate the indicator moving to show you when to engage the nut.
This seems like a really good idea!
Hmmm - he'd have to write a routine to change the phase of the leadscrew relative to the spindle, depending on which thread you are going to drop the nut into. Otherwise you'd be waiting an age for the next opportunity to engage the nut.
A practical solution for hybrid threading in cases where you need to open the nut (say when threading close up to a shoulder is to hit the brake at the same time you open the nut, then you can use the threading dial to go back, not just to the same number on the dial, but in the same revolution of the dial (in other words, the half nuts lock into the exact same place on the leadscrew). You may have to turn the chuck back a bit if your lathe didn't stop quickly enough.
Great video, as always! Thanks for sharing.
Thanks i am looking forward to the next video.
Seems like the reverse feature would also be handy for threading away from the chuck with the tool right side up.
Excellent! You have made so much progress while explaining all the steps that drove your decisions. You are well organized and smart. You obviously have a EE background, with your knowledge of scopes and circuits. I’m an optical physicist but can’t understand electrical stuff no matter how hard I try.
The link for the servo motor on eBay doesn’t go through. Can you give me the specifics on the motor and driver?
Thanks,
Paul
Mod tip, add input from linear DRO glass scale sich that the thread dial can be disengaged and re engaged when doint metric threads with an imperial pitch leadscrew.. e.g. (input about position of apron vs rotation of leadscrew) add something like a led that lights up to show when to engage.
One word ..... AWESOME !!!
Great work again. But I think metric treat has some calculation issue. in this case some arduino Code and LCD screen and easy sets might solve this problem. Forexample you wanna go 1 mm pitch, controller will know how many revulution per desired pitch distance. By doing that you may also add safety issue. Forexample limiting the sipindle speed while desired pitch distance increases. Other safety issue is if servo failed and error led fired, all system could shut off. Thank you for valuable input and effort.
Love your videos. Didn't think I would, but I do. Subbed.
Now I wait to see you make the lead screw stop at the shoulder based on the counts? Maybe halfnut always on, drive carriage back with a button. Keeping track of counters will be a challenge, but it would be fancy. Also there is an urgent need for modular pitch, to make your own metric gear hob.
This design opens the doors to so many projects :)
Wouldn't it make life easier simply to start on the 1 of the dial indictator? Actually it should work on all numbers as long as you always use the same one
That way you always engage on the same position
love it! well done! ok you have a controller you can calculate the angle f the chuck, you can also count the turns (well number of degrees) so you can mark in memory just before the start of the thread and you can count to the end of the thead and you can electronicaly stop the tracking so you can repeat the cut automaticaly using the controler to engage at the correct angle and start point and disengage at the end cutting its own gutter or even tracking back along the thread! maybe two buttons , onr to mark start and one to mark stop?
Here in Europe we use the metric system however if I want to cut imperial threads on a lathe with metric leadscrew it is only necessary to replace some gears on the left side of the lathe. It is much easier than replacing the leadscrew.
If most of the threads were imperial then I bought a lathe with Imperial leadscrew.
This process is even easier if the lathe is equipped with norton gearbox, just make the combination of levers.
Thanks for the GREAT explanations ! Question : couldn't you "emulate" the halfnut with your Electronic LeadScrew too ? when "disengaged" you stop advancing the carriage, add some buttons to "fast move" it, and when reangaging you first resynchronize by how much step you've changed modulo the thread pitches /main number of turns ?
I practically was thinking the entire time: couldn't one solve the entirety by counting turns of the spindle and pausing the lead screw until the right one comes around?
But also: with either a real or ghetto DRO one could automatically stop right in the gutter, couldn't one?
No need for a DRO for that. The stepper-driven leadscrew IS a DRO of its own, the controller knows exactly how much it moved you as long as you don't disengage the nut; so all you'd need to do is program a button as a "stop leadscrew now" (instead of stopping the whole lathe), retract the tool, hold another button to ask the controller to jog back to the other end, re-set the tool, then just tell the controller to start turning the leadscrew again whenever it would be appropriate to keep it in sync, and go all the way to where you stopped it the last time; rinse, repeat...
@@AttilaAsztalos I was thinking about the possibility, that backlash and missed steps might mess this up. But maybe that would turn out to not be a problem. But I tend to always search the worst case ;)
Backlash is not involved - you would obviously always start actual thread cutting in the same direction (only the return move goes the other way and that cuts nothing), which would take up backlash before touching the thread; as for lost steps, there are not supposed to be any, ever - if they do happen, that electronic leadscrew would be a failure (or taking a much too heavy cut); no amount of lost steps is "ok", but a reasonably oversized motor is simply not supposed to lose steps...
@@AttilaAsztalos Or add another encoder to detect the step loss and at least stop the lead screw (and start beeping maybe ;p).
Great project well done !
Thank you for the very understandable exile nation.
I'm glad it was understandable. I wasn't sure how it was going to turn out.
I now want to make a clockwork gear change for the indicator for lead screws. There seems no reason why you couldn't have a change gear for both cutting types if you have a multi geared transmission in your indicator. Which really only needs a single bar for moving between the two sets of gears, and possibly a change of the face.
James, once again you have done an excellent video!
Now my question is: Will I be able to add 1 extra encoder to the system? One that is mounted on my thread chasing dial and input into the mathematics and produce a "count" output to a an LCD.
Here is how I cut metric threads on my imperial lathe:
The gearing between the imperial threaded lead screw and the spindle is 127:100
A 1.0mm pitch thread produces precicely 127 threads per 5.0 inches.
The lead screw has a 4TPI thread. 5.0 inches is precicely 20 Leads-screw-threads (LST)
The Thread-Chasing-Dial (TCD) counts 16 leads-screw-threads (LST) per revolution of the dial.
While ever the half-nut is closed the TCD does not rotate. But as soon as the carriage gets to the end of the cut and I disengage the half-nut, the TCD starts rotating and I have to watch it, because I leave lathe spindle and lead-screw running continuously.
Now the half-nut is disengaged and the dial turning (counting threads on the lead-screw)
I move the cariage back to the right and the TCD rotates even faster ( i'm tripping over lead-screw threads).
127 threads cut on the work piece equals 5 inches (OK i didn't actualy cut 127 threads, I probably only cut 25, but pretend there are another 102 "virtual" threads back to the right of my workpiece.
I need to move the cariage and re-engage the half-nut at 20 LST's to the right.
Every full rotation of the TCD is 16 LST's. So I count one full rotation plus another quarter of a rotation and then engage the half-nut.
If I overshoot the 20th LST I can "catch up" by moving the cariage back to the left at a faster rate than the lead screw is turning, the TCD then rotates the other direction and I have to keep a close count on what is happening, and hopefully engage into thread number 20 before I reach the work piece.
If the lathe is turning too quick for me and I miss it I have to go another 20 LST's and pick it up on the 40th if I haven't lost count of the TCD!
For a 1.5mm pitch thread, 127 threads on the workpiece is precisely 7.5 inches, and that is equals 30 LST's (7.5 x 4), 1 full turn plus seven eigths of a turn. of the TCD. Can also pick it up at 60 LST's (15 inches = 2 x 7.5) on 3 and three quarter turns of the TCD, etc (60 / 16)
A 0.8mm pitch thread, 127 threads cut, equals 4 inches, or exactly one turn of the TCD, or any multiple of it.
so what I need is the number of count of the thread chasing dial displayed, so I don't have to concentrate so hard. An "Electronic Thread Chasing Dial".
Unfortunately,I can only give 1 thumbs up!
That's okay. Thanks for watching.
@@Clough42
Wouldn't it be possible to read a linear scale that is attached to the carriage to get the information about the position of the cutter in regards to the thread?
This would free the leadscrew to be a pure device of moving the carriage right/left at any speed your electronic device desires, with the benefit of your device being in the know where the carriage is located in comparison to the thread that has been/will be cut on the spindle.
The TPI of the leadscrew would become completely irrelevant then and only needed for servo motor sizing vs torque and speed.
@@joansparky4439 I believe this could be done. It would be complicated in software, since the backlash and flex in the system would be in the feedback loop, but certainly possible.
@@Clough42
You sure?
Manually engaging the half nut when the number comes up has that very same flex in there.. backlash shouldn't matter.
Works like this:
- your ELS knows position of the carriage (linear scale) and rotation of spindle (rot encoder) and LS (servo drive)
- 1st cut of thread, LS controlled by ELS (same flex and backlash as manual)
- at end of cut, disengaging of the halfnut, manual reversing of the carriage back to before start, ELS registers this new position
- engaging of halfnut, ELS controls LS rotation to match thread cutting with spindle and 2nd cut of thread is happening (same flex and backlash as manual)
The only thing different between doing it manually with the thread dial and your ELS would be, that the ELS is matching the LS rotation to match the spindle rotation at any point you engage the halfnut, while if you do it manually you're hooking up the carriage to the LS which already is being hard synced to the spindle by the gears.
Might need a switch to detect half nut engagement.. but that would probably only be needed for a first version as this info should be available via the linear scale from the carriage as well, as the ELS would be able to recognize that it is driving the carriage and not the user.
Or after engaging the halfnut the user hits a button (halfnut engagement detector) and the ELS syncs the LS rotation to the spindle rotation to get the tread cut, so one engages the halfnut on a non-rotating LS in this case. Neither flex or backlash should be different from doing this manually. At least I can't see how that would be..
Your ELS PID-controller would then steer the carriage position/speed instead of the LS rotational speed.
Ya know, you don't have to reverse the entire lathe. The lead screw is motorized; just reverse the lead screw motor. An index on the spindle encoder can keep you synchronized, or a programmed counter that resets after 4096 counts.
You can disengage the half nut, you just have to re-engage it on the same number every time, and don't get off by a tooth. Much easier to keep it engaged and reverse the spindle.
not when cutting a metric thread on an imperial lead screw. the change gears only approximate the divisions for metric threads. you will accumulate error for every time that lead screw rotates while disengaged from the half nut. try it and see.
Hi James. Here in the UK we have a monthly publication "Model Engineers' Workshop Manual". The editor has been kind enough to include my e-mail to him about your video series on ELS in the just released volume. Issue 283, August 2019, page 18. BobUK
Very cool. Thanks!
Cutting lefthand and righthand threads suggests that you could use this to make cut knurls. Maybe use a tool with a greater than 60 degrees?
For BOTH situations where...
1) lathes don't have the threading dial,
2) Metric/Imperial combinations of Leadscrew/Thread or vice-versa are used,
and the engaging of the half nut becomes impossible to know then the solution could be to implement a half nut engaging countdown on the ELS display together with sound beeps for the last 3 countdown numbers like....
..., 5, 4, 3, 2=ShortSingleBeep, 1=ShortSingleBeep, 0=LongBeep
-> engage the half nut!
There have been several comments about some variant of this feature. The issue is that it isn't just about the angular position of the screw, but also about the physical position of the carriage. The controller would have to know both to be able to track this. So maybe with a Z scale on the carriage it could be done. I haven't spent a lot of time thinking about it.
@@Clough42 You are right!
I think that it than becomes half a CNC and thus outside the basic scope of the ELS.
Keep up the good work.
I mean, a 1mm pitch metric tread is just a 25.4 TPI imperial thread, and if you're doing everything electronically, that's just a simple (yet perhaps random-seeming) calculation. With the right motor/servo, you could even do away with needing to engage and disengage the half nut and work entirely based on if you're spinning the lead screw or not.
Very informative, well done.
I would add an interlock to the forward reverse switch so that you cannot select the wrong direction by accident.
How about adding an automatic stop when you threading towards a shoulder?
These are all good suggestions. Stopping automatically introduces the challenge of resynchronization, which so far I've been able to avoid tackling. Someday.
@@Clough42 automatic stop is exactly what I was thinking this whole video - I'm imagining basically a microswitch on a mag base. Set that so the carriage hits it when it gets to the thread "gutter". Microcontroller stops the leadscrew immediately when the switch gets hit and stats counting pulses from the spindle. You stop the lathe, then put it in reverse, the microcontroller counts back down and turns the leadscrew back on (in reverse) once the spindle counter gets back to zero.
Wouldn't solve everything, but would take the stress out of stopping the lathe while metric threading and would be fairly simple code
I think, you are an universal genius. Absolutely great stuff. Programming the controller, fusion 360 etc. I started to copy your little plant, but i am not nearly as genius as you. I need a wiring diagramm. ?!? Can you help me?
There's some information on the github page about which pins are used for what. I'll have more schematics coming with the PC boards as well.
Clough42 I’m so glad I came across your channel. I’ve watched all these videos multiple times. I’ve ordered the TI LaunchXL and other components. Where do we put our names on a list to buy boards and anything else that you decide to sell? Mechatronics is a hobby for me, trying to learn as much as I can as fast as I can. Is there any forum where us fans of your project can share ideas and questions specific to just this project or do we stick to asking questions here on TH-cam?
By the way - in case you didn’t get the message -- you are a bloody genius! Thank you for letting us travel this journey with you.
I wonder if there are any step marks under magnification? At some points it looks like I can see steps, but the spindle rotation looks smooth.
i wouldn't imagine there would be any significant marks due to the inertia of the system
Hi James. Thank you very much for this project...this is truly a gift for all of of the hobbyists community. I´ve been cheching the firmware you´ve posted on github but I´m a little bit lost between the different files. I am really excited with the possibility of get rid of the feed gears and I have almost all the required components, but unfortunately the controller part is the one that causes me the biggest of the doubts. Would you post for us a video about how loading the files to the controller?
Thank you very much!!!
This is on my to-fo list.
@@Clough42 Niiiiice!!!!
Almost forgot, i can see this lathe fully CnC with your talent your half way there.
I get this comment a lot. I think I want to keep the manual lathe, just so I can turn the handwheels and make something quickly without fussing over feeds and speeds at the computer.
Looking great. Can't wait to ditch my change gears.
I have a lathe/mill combo, equivalent to a Grizzly G4015Z. I'm wondering whether the software could accommodate driving the leadscrew at a selectable set speed, to provide a power feed for the mill. The change gear setup doesn't allow this but it would be very useful.
It would be massive overkill, but if you were using an encoder on the mill spindle, you could calculate and run feed-per-tooth for your cutter. I'm trying really hard not to get distracted by cool stuff like this. :)
@@Clough42 Good thinking. :) A second encoder on the mill spindle would do the job. It might be possible to find a way to drive one encoder when either mill or lathe is selected, but the clutch arrangement makes it somewhat difficult.
Any thoughts of a CSS modification? With the spindle encoder and a cheap digital scale CSS would be relatively easy to do and the surface finishes would lifted to the next level. CSS is the reason CNC turned parts always look amazing.
CSS is kind of a different animal. In this case, I'm controlling the screw based on the spindle position. CSS requires controlling the spindle RPM based on the X position of the cross slide. That would interesting, but kind of out of scope, I think. Unless I'm misunderstanding you.
I was thinking of it as a separate project. My reference to the encoder muddied the waters.
If you encoder is also providing you with one-pulse-per-rev, then you could use this as a sync feature for when to start threading and you would not have to leave the half-nut engaged all the time. At least that is what I suspect :D would be cool if you could try that out :)
Would not work, because the encoder has no idea where the carriage is when you engaged the half nut. You would also need a very precise linear encoder on the carriage.
I guess that I'm missing something here, as it seems that if you can change the relative speeds of the spindle and the lead screw, as long as you engage the half-nut always on
1, you should be able to cut metric threads.
Hi, I enjoy your vids, I am new to your channel and likes the description on threading and halfnuts. One thing crossed my mind, you have lathe and mill? and buy gears, why?
I believe it's just a matter of cost vs. time... As inexpensive as aluminum timing pulleys are these days, it probably just makes more sense to purchase them. I'm stubborn and pinch a nickel probably more often than I should. I often find myself spending hours making something I could buy for $10... then I kick myself. ;)
Those aren't gears. They are pulleys. I don't think you're realizing how difficult those would be to cut yourself. Did you see the "rim" around the edge? That means you can't single-point shape/broach them. If you wanted to kill them with an end mill (it would have to be a tapered EM, or a form tool if it's not a straight taper, but in not sure what the form of a pulley is) you would have to make it wider than the belt to account for the radius of the EM when you come up against the "rim." That would introduce tracking issues and might bind up the belt. Not to mention tee time it would take.
PERFECT, Your a real PRO.
No question but we already knew that it would do metric. ;-)
Instead of reversing the motor could you have the controller advance the lead screw a specific distance and then reverse the lead screw leaving the main motor on? Maybe have a cycle go button that when pressed advances so far then returns, you can then increase the cut and hit go again
Yes. 1st press: record start position, start threading. 2nd press: stop moving, record end position. 3rd press: rapid to start position. 4th press: threading until end position. Goto step 3. (Plus a way to reset)
Of course all threading must be synchronised to tracked spindle angle+pitch
@@ivanblogs good point, the acceleration in the reverse direction might be an issue, advancing wouldn't be a problem you could set the start point far enough back that it would be up to speed by the time it starts cutting, might still be worth a shot
Scott Kennedy that's exactly how it's done on Hardinge HLV. It's all implemented in mechanics there.
@@scottkennedy7741 Yeah... I glossed over the acceleration issues. I believe the current implementation ignores speed/acceleration issues, and assumes everything is bounded by the spindle acceleration.
It's not trivial to implement, but also not rocket science. (Unless you're threading a rocket...)
James, Great mod. If and when you plan to offer complete kits, do you have a ball park dollar amount in mind?
I am doing my best to resist the temptation to keep adding features. If you want a polished commercial product, they're out there. I'm targeting something much less expensive that uses mostly off-the-shelf parts. No matter what I do, though, the motor and driver are going to be a bit part of the cost.
@@Clough42 - Mmm, feature creep...
Always excelent James!!!!!!!!
So, have you thought about making your system open source?
Thanks for sharing.
there is a github link in the description
Hello James. Thank you very much for your content... It's excellent as usual.
Just a question... If you engage the threading dial at same position, you could move the carriage and still make any kind of threads right? I mean, you will be finding always the same grove on leading screw. So you would have a option to disengage the locking nut in that case.
I'm I mistaken?
Greetings from Brazil. A metric country but imperial bolts and raw materia widely available also. 😁👍
Yes. On the *exact same position*, not forward or backward any rotations.
Can be handy if you need to stop the carriage in a very short distance, but your lathe spindle takes too long to spin down.
So just set up a hard stop to the right of the carriage and use that as the engage position using the same number each time?