ฝัง
- เผยแพร่เมื่อ 16 ต.ค. 2024
- I must have watched more than 30 DIY CNC build videos/series on TH-cam, and never saw anyone else having these frustrating electrical problems... I naively figured this part would take about 20 minutes, but by now I have just about spent as much time on this crap as I did building the whole fucking machine itself, and it's still not fixed!!
![[UNCUT]"ทนายตั้ม" แฉยับ! ความลับสวรรค์ ใครคือ "บิ๊กเทวดา" ของบอสพอล l คนดังนั่งเคลียร์ l 15 ต.ค. 67](http://i.ytimg.com/vi/ru9fpSaUwdA/mqdefault.jpg)
I had a similar problem with serial communication due to interferences. logic signal fails very quick next to that much interferences. what I was told to do in that moment and it worked it was to wind with a gnd cable each of the serial wires between the Arduino and the module. and after dining that I had no more errors and a very clean signal on the osciloscope. the ground might be connected to both ends, one end to the Arduino o and one to the module. so what I had was 2 wires of ground and two for the signal. hope you won't have any more problems. cool mill 👍
intresting!
The motor leads are acting as an aerial (antenna) with about 100v at the PWM switching speed. I suggest a shielded cable but only connect the shield at one end, as don't want to make several earth current routes. I would try the power supply earth where the arduino is referenced first. A small valure capacitor across the motor may help in addition. If you have an AM radio it would be intresting to see what interference the CNC makes.
On a less technical note, nice to see you break stuff in the video! And props on the humor. I like it
In my DIY CNC router I shielded the stepper motor cables and the spindle motor cable. Haven't had rf problems since.
Try putting one of those ferrite clips on your USB, power and other comms cables - that will act as a common mode choke and hopefully solve the RFI from the motor.
Lol, you've made me laughing with that "wrong clip":)))
I'm loving this series! And the comedy at the end was pure gold! Great job iforce2d. Please don't get too frustrated or give up, I await the next video.
I too had similar issues with my CNC router. The main fix that I found in the end was to use a high quality shielded USB cable, El cheapo USB cables seemed to have this issue but the higher quality shielded ones did not. The results were consistent with different cables, longer usb cables or usb extension cables caused issues too.A note on grounding if you are shielding the motor cables, tie the shield to ground on one end only or otherwise you can get circulating currents which can cause your electronics issues too.Another thing you can do is add diodes to between the across the cables to disipate the back emf caused by the switching of the speed control. (Most drivers should have this internally anyway but something to try maybe).My 100V spindle sucked up mdf dust pretty early on so I replaced it with a 1.5kw spindle with a vsd, I used sheilded vsd cable and have had no issues like this ever again.Hopefully this is of some use.
Thankyou! Yes, a slightly shorter USB cable and caps on the motor terminals is working fairly well right now.
it is an arduino clone? if it is a clone has it have the usb interface chip ch340g or the atmel one?
I have a similar problem on my mini CNC 3018 (lost connection ). Your machine's hardware is very good. Thanks for sharing and keep it up :-)
make all leads as short as is practical. route the motor leads away from the stepper leads. shield all the leads with well grounded shielding. Twist the motor pairs and also twist the stepper coil pairs. The twist means that any interference picked up should be equal in each lead and therefore cancel out.
Twist your wires, shield them, put ferrite chokes on the cables, use a buss bar to ground/earth everything, filter where applicable (capacitors/resistors), do a really long test with no spindle to verify it's the problem. I've had to do this type of stuff with big drones. Don't despair, you'll get it.
try adding a capacitor of a value from somthing pf to nf across motor connections and see how it goes. Also add a diode to motor connections. connect katode to positive connection and the anode to the negative.
Are you powering up arduino from a different power supply (isolated from motor power supply); no ground connections connected together ?
12:36 I've had that same problem with my CNC machine when I used a too long and un-shielded flat cable between my computer and CNC controller acting as an antenna. It would add random steps to the motors while running a job and screw it up.
Shortening the cable to the proper length and shielding it solved my problem.
Just saying.... I would shield all cables especially the 100v spindle cable it is transmitting big time. Then find a common ground point and connect all grounds to that one point as mentioned below.
Great job mate!! According to yours problem with interferences try use metal sheath around stepper motors cables and router cable and ground it to the CNC frame and power supply. Try use the ferrite cores for all cables. Good luck!!
Maby check if you got a few ferrites you can use to wind the cables around and check with the scope if it does anything?
PWM impresses large supply rail distortion at lower speed settings. It is especially disruptive when supplied by a SMPS due to instantaneous (square wave) current harmonics relative to the switchmode supply switching frequency. Best solution is a separate supply for the motor drive, not a separate output from single supply. Also high PWM frequency causes masses of near field emi, again at low speed settings. I would run balanced supply to spindle motor in shielded wire, it would require opto coupling the arduino to PWM speed controller for control. Another beneficial trick is to make a double wound toroid filter to fit as close to the motor as possible. Also, do you have a super fast shottky diode on the spindle motor, not at the controller end. You need to snub back emf's as close to the source as possible. Diode must be rated to full motor driven current too.
More coils and capasitor to filter power lines... Could be spindle causing high frequency noise on power lines or io- lines
those 2 volt spikes are plenty enough to cause problem with microcontroller..
And you could use mesh shielded cable to power the spindle. Just remember to connect shield mesh to power supplies ground. Such cable seems to be standard cable for motors in industry installations..
Also aluminium need cutting fluid... Helps keep tool temperature low.
Bah I was Binge watching your vids. Hope you upload soon. Thx for this, I am enjoying and learning at the same time. Especially since I just got my first CNC. Can't wait till it's up and running to it's fullest.
You can try to place capacitors across the motor leads. That's what we did growing up before brushless motors came available for rc cars. Brushed motors put out a lot of RF
your videos are awesome! i really like the way you talk through the issues and take time to explain things. thanks!
I don't know what you ended up doing with this, but I would install and isolated mount for the spindle motor. Years ago, I had to iso-mount fork truck terminals from the chasis of the propane driven fork trucks. Their spark plug noise was preventing the radios from working properly. Floating grounds are always a problem, but iso mounting will help a lot. Back EMF diode on the motor might help.
It is sometimes difficult to get rid of RF interference, especially with unknown chinese power supply and regulators. Also, check on the scope your laptop PSU and GND of the arduino. Your computer may bet interference over charging cable too (serving as a loop antenna, or even worse all garbage from the mains). If your mains is RF 'dirty', maybe due to pulses from motor regulator, then you should sense with O-scope all possible RF pathways, and then get rid of all of that RF noises. It will be good also that motor regulator has those yellow choke as seen in PSUs, with two wire wounded at once, and from each wire to the GND some 100 nF capacitor to the ground, in this case common ground connecting everything as a single ground, then to the mains and outlet ground. Use some AM radio as a 'helper' to listen possible RF noise sources. Then you will know what has the most RF power, then block it. Good job anyway!
Hi. The scope is not isolated? When you hook it to the arduino, it has same ground as the motor driver, and motor current travels through the scope probe ground and is picked up by the probe i think. noise is a hard nut to crack. seems like the star ground trick helped :) I love this problem solving you have here.
ferrite core on the usb cable?
I had to trouble shoot a problem that sounds very similar with the theatre production Devolution several years ago. Your problem is most likely caused by the chopper drive, and also induction from the motors. You need to more than adequately filter the system. Driver chips need filter caps as close as possible to the pins plus a power diode for each chip. This is to prevent spikes and rubbish getting back into the power supply. Filter the power and GND pins on the micro's with a filter cap - this is important - you have to stop EMI/RFI in its tracks, and that can only be done by filter caps. Do not cheap out, use good quality capacitors. Segregate the GND for all parts of the system. Run individual grounds for laptop, arduino, driver board and power supply and connect them together at only one point. You need to eliminate looping earth. Test your earth connection point and make sure it is sound. Lastly, twist the motor wires - once every cm should do. This will help prevent radiated interference. Remember that components are made as cheaply as possibly - my experience has shown that filtering is always a problem, especially in situations like yours which will naturally produce noise on the lines. Good luck :)
So, you think the PWM spikes can just be filtered out? I did too. Then after a year of trying I wasn't so sure anymore. So I finally put some optocouplers on my control lines and lo and behold, they fixed it. Coincidence? I think not!
Congratulations on solving the problem. Please remember, you have the advantage of having the electronics and equipment in front of you and can trial and error test your theory first hand, Those offering advise over the net can trouble shoot only from the information given to them by yourself, We can relate that data to problems we have encountered that were similar, and share solutions that worked for us. There is no guarantee, only a willingness to offer advise that may assist another person with their problems. I am sure that any competent engineer with a CRO, working on your equipment first hand, would have found a solution to the problem in 5 minutes.
You can use aluminium foil to make a makeshift shield. Surround the arduino in it and that should protect it somewhat from the RF disturbance.
It took me about a year to figure out noise issues I was having with my stepper motors. I finally solved it by optically isolating the step and direction lines of the drives. I don't know if your drives are isolated, or not, but if they aren't then you're better off adding that. If you're interested I can give you the schematic.
Hi there buddy what stepper motors did you use nema17? And spindle any links? Cheers
yes nema17
docs.google.com/spreadsheets/u/1/d/1zyGDZSG_wHfn7GA0wKzwLYhK88SAIyLOe-4c9KuL23A/edit?usp=sharing
I forced thanks and the rest 8mm road and couplings?
I suggest increased the speed of drill and reduce the speed of axis movement, maybe use oil or water for reduce the friction between pieces for have better milling, and for the electrical problems change the usb or serial connector cable and add more capacitors for better filtering, and reduce the long of cables for have low interference.
Try a Faraday cage around the Arduino and shield. Is the motor AC or DC? The speed controller section of the power supply may be the problem, SCR based controllers can be incredibly "noisy", do you have any way of running the motor at full speed without using the supplied PSU? Shield everything you can, ferrite toroids, as mentioned elsewhere, on every power cable may also help but identifying the source of the interference/spikes and isolating it somehow .....
I know this is probably solved by now since I am seeing this in June. I admire your efforts and applaud the struggle. I can't help but wonder if one of the Brushless DC spindles might do better?
Congratulation!! Your machine itself is sturdy enought to cut ally but your bit broke because it was had too many flutes, too long and too small causing massive tool deflection even on the ply, tool travel too slow because you need the tool to actually cut a chip out and if it can't get a bite it just rubs and bends as the spindle traverses causing heat and softening the ally to the point that when it does cut it gums onto the end of your tool then no exposed cutting edges. I use single flute 3mm centre cut bits from China, less that a dollar each delivered ( I break many more that I get to wear out ). Also, that ally looked like something from Bunnings and only some ally can be machined easily, some of it machines like bubble gum. I do admire your perseverance and patience with the electronics, I gave up after a week and running out of Uno's, shields and drivers so I bought four $20 drivers and a $15 parallel port breakout board, blew the dust out of an old PC with XP, installed Mach3 ( you can run upto 500 lines of code without buying a licence) and had the machine wired up, calibrated and running in one long day with a lot of help form Google. The only problem I had was noise causing the limit switchs to trip an estop so solved by removing them: on small machines they aren't necessary if working in the centre of the table and also makes it easy to square the Y axis by just driving the gantry onto the end stops. All the best with your build and I am willing to bet that it won't be your last.
looks like the tool had a built-up edge, try increasing the cutting speed and decreasing the feed rate, you could use an speeds and feeds calculator if you have any way of measuring the spindle speed , there are also special aluminium cutting tools that you could use
Check you are still only getting a max of 100v from the spindle power supply, I've tried most of the things you have checked and while initially I got a max of 100v now the power supply goes upto 130v, I think while I was messing about with different grounds, I've fried something in the power supply.
at a guess i'd say its due to noise being put on the line by the power supply - at low speeds its chopping up that voltage by rapidly switching the output on and off... try adding some good size 150v caps on the power supply output (after the speed controller)
also grounding your spindle/frame may help, if you haven't already.
Keep at it dude! So awesome watching your adventure / problem solving into making a CNC :)
Had similar problems getting an arduino to run my car ignition , damn thing would work fine up to 2000rpm then die/reset . Stuck arduino in metal box (biscuit tin actually) and earthed the box, pretty much solved my issues . Also a low esr capacitor on any dc supplies helps.
is your endmill center cutting?
How deep was your cut? check with your tooling manufacturer guidelines, but I would start with a cutting depth of 25% of tool diameter and adjust in 5% increments. a tach for your spindle would be really handy right now. there are formulas used to figure spindle speed and feed rate.
I saw your poor aluminium cut results. I had the same poor results and broken endmills on my first try. But then I used ethanol to cool/lubricate the cutting, that was a game changer. With ethanol I get very smooth cutting surfaces and no more broken mills.
two way capacitors on the motor!
keep it up man, i like your videos. for what its worth, i'd suggest trying to resolve the EM interference issues, leave buying new expensive parts for last. the interference problem might cause the cutting offset too
Maybe putting Ferrite cores on the cables would help?
you also need some lubricant to cut. Even soap and water can help cut aluminum. Also slow the spindle down.
Is the power supply for the spindle isolated? Is it even a Mean Well or a chinese cheapo ? Looks like back emf, but it is ground loops or via the spindle enable ? Either way id replace that spindle power cable with something shielded or drain.
Have your tried to decouple the notebook or the driver shield?
The problem can arise from a few sources:
- electromagnetic interference, i doubt this but I'm working with 433mhz 1w modules and some arduinos are going literally crazy when i switch it on
- high voltage spikes on spindle wire crosstalk with other wires (arduino shield, uart)
- huge spikes coming back from spindle eventually getting back to arduino raising/lowering thresholds of digital inputs and causing missinterpreted uart commands or steps the motor
- interference on uart/usb cable
I would try to use
- a bluetooth serial adapter for the arduino-pc communication (~1usd on ebay)
- use a "choke inductor" plus a big ass buffer capacitor on the entry side of the shield/arduino be awar that choke+cap act as LC component and thus can oscillate or amplify the noise (but in that case you would know what to tune.
- wrap the arduino shield into a papertowel then alufoil and shield it (or just put it in a metal casing)
- optical decoupling for the arduino and the driver shield
Have you tried to google the issue?
"cnc spindle interference" gave me a lot of results and many suggestions solved the issue.
Are there any capacitors on the motor connections. I don't know much about this stuff, but all of the old brushed motor model vehicles have them. Shield motor wires and only ground on one end of shield like Graham Wise suggests might do it. This is like my son's problem with humming in his car stereo. He has purchased new Pioneer radios and of course has amps and woofers. Different people installed 3 different times and the alternator has been replaced. He feels like you do about this situation.
So if its not a rude Q..... whats the cost of the build and will you be letting your plans for your design out so we can make one?....... nice build BTW
You could use an external usb to uart converter and isolate the TX and RX of the arduino from the TX and RX of your usb uart converter using optocouplers
have you tried making a shield for the arduino like just making a grounded metal case for it.
Disconnect the motor and measure resistance between the wires and between the metal on the motor (and between the wires).
Just to see that it isn't faulty. Also check the resistance while slowly turning the motor.
The smaller the bit are, the faster you need to turn it. So a bit larger diameter will get its outer cutting part to move faster. And it will be a bit stronger.
Someone already probably mentioned this but not all end mills are suitable for plunging cutting, they need to be center cut.
hahah awesome that you to put some humor in at the end xD.
Try adding three capacitors to the motor. I used to use 0.1uf caps on my rc cars. One from positive to negative, one from positive to the can of the motor, and one from the negative to the can of the motor.
Maybe low quality usb cable that looses connection when the arduino gets exposed to vibrations of a certain frequency (it seems to be attached to the same base plate like the motors).
I would try another usb cable and (at least for testing) put arduino and laptop on a different surface to get isolated from vibrations.
the other alternative is to use opto-isolators with transistors and a battery with the arduino to completely isolate it.
i got a router spindle on my cnc and same as you when i turn it on the arduino crushes. My solution is to plug the router to a different electrical line it happens when i am using tb6600 and drv8825 driveers. But when i change my dr ver to dm556 it doest happen anymore...
have you tried connecting all the grounds so everything is at the same potential?
yes, that's what I did in this video...
I think grounding ( main's Earth) the body of the motor and the holder will be enough for shielding the motor.
minute 2.37 epic, got me laughing thru the rest of the video, very nice work altrough and i apreciate your perseverance, sory for bad english
Try wrapping the chord of the spindle motor onto a ferrite toroid. That's how laptop chargers, and pretty much any good smps does to remove transience(that little cylinder near the end of smps wires is a ferite toroid :D). 3-5 turns should be enough, also try to put the toroid closer to the motor. If that doesn't do much try wrapping the stepper motor wires onto toroids and put them near the arduino.
And your bit broke because you were spinning way to slow.
Looking at the noise spikes at time 7:32 of the video, it shows about 16 kHz frequency. It is very likely poor EMI filtering in the switching power supply. Try using a Linear power supply to test. When you touched the motor to the plate it showed the noise even though the spindle was not running. A linear supply should fix that. It seems that the EMI is conducted and not radiated. Best of luck to you.
I also lose connection to my arduino, and i am using a Dremel spindel connectet to 230V direct.
I have the same problem with arduino. No communication when I run the spindle. I have used 3 different spindles. And the effect is the same.
Did you somehow fix it ???
I think so. Caps on the motor, ground everything properly. Just keep watching the series and read all the helpful comments :)
dirty earth?
if all connected to common ground, (frame too) then does it disappear?
is the motor rf creating noise back through earth.
faraday cage on arduino (alfoil)
So would you recommend this motor/power supply? :P
You can make a zero probe, by connecting sda or scl (Not shore witch on) to ground.
Can you try to mill a pcb? Love your stuff keep it up
A second thought - you may have better luck driving the cutting motor with a wound core transformer - much less likely to induce EMI/RFI into the power system than a switchmode.
Ferrite on your cables should take the edge of the RF.
Just use some shielded cable for the spindle. I know my reply is late and it probably has been said before, but it may help others
disconnecting the PC just make it worst, everything have "floating" ground. You should connect all ground potential together. (for switching powersupply it's not recommanded to connect the enclosure to the ground but you should check if it's possible to connect the 0volt output to gnd). your issue seem to appear when you turn the motor off, maybe a flyback diode could fix it
How about those Arduino driver shield? Work ok with that CNC stepper?
Not sure what you mean by "those", but you can see here the shield I'm using: th-cam.com/video/OuhgZMDXDdY/w-d-xo.html
I had almost my pants full of shit ... muahahahahahahahaha!!!! :-D Same issues here! And I felt with you!! This animations and the video of a professional machine ... really ... still laughing :-D :-D
Very nice design and vids. You in nz?
I finally found soeone with the same project, setup and problem as me :D I´m looking forward to see if you found a solution. Thanks for sharing (I know I´m late)
I eventually gave up on the wired connection and I now use bluetooth between the laptop and arduino. Never had a single problem since. It also lets me move the laptop around more conveniently. Many laptops come with bluetooth devices built in (as mine does) so you don't even need to plug in a dongle.
I’m not alone! I have the same problem. Same spindle. How have you fixed this?!?!?! I was laughing as I went through a very similar diagnosis.
I eventually used a bluetooth connection between the computer and the arduino, have not had any trouble since then.
Hi, i have the same spindle and power supply and the power supply is the absolute garbage, it will murder every electronic device in the vicinity with electromagnetic interference, the noise on the wires between the spindle and the power supply is in the range of 1 to 1.5 kV!!! There is absolutely no inductive spike protection there.
Get the power supply and the spindle wires as far away from the driver as possible.
Forget about controlling the spindle from the arduino.
Also, it took me about a year to get my own fully diy cnc to work as i want it, so do not get discouraged, there are way more problems ahead of you :)
I have that same spindle and PSU. I have the PSU screwed right to the side of my machine. No problems.
I don't know if this would help, or if someone has said it earlier... I would suggest an earth ground to your machine, my thought is that the EM energy would go to ground.
remember to cool the aluminium with alcohol, so it don't get soft and "grabs" the tool and the tool breaks...
that's why u should use a 5v microcontrollers (or 10v if available ) instead of 3.3v , use filter capacitors of 4700 micro at power supply outputs and 220microfarads at aurdino input supplies , and why ur power supply showing mains voltage on case etc etc strange , change the supply to smps desktop ones, H bridge is also an issue with spindle motors and steppers , use a reverse emf protection (clamping schkttoy diodes) H bridge circuit , make the h bridge ur self . These spikes hang the microcontrollers ....4700uF and 220uF are minimal values given , u need to go 3 times higher ...
please try to connect all ground together like usb ground to the external power supply ground..this is serial communication problem...
all grounds must be connected together. possible intermittent connection, bad pot or bad motor.
ferrite beads may help. run your wire thru the bead.
use lubricant with aluminum, few drops could help in that drill wrecking situation
Hi would you make a new version of the gps rth rc car
Is the spindle motor isolated (rubber gasket) from the whole frame?
Hell no. The whole point is to keep everything as rigid as possible :)
rigid thin plastic will do.. a gasket won'y make it less rigid
the point is you should'nt be getting power off the the frame or the power supply case.. That's a no brainer
Plastic is less rigid than aluminium, I'm not gonna argue about that sorry.
Lol but it goes between two peices of aluminum.. it'll only make your aluminum even stronger.. your bolts are what you should be concerned about more than the plastic
How much cnc?
Feedrates ?
You cant tell how accurate the machining is if you hold it by hand
Yeah I wouldn't have my hand anywhere near the machine when it is running. Especially if it is running flaky. But hey they're his hands.
Banggood seems to list a new product that might be the solution to your problem, worth taking a look anyway. Search for "ADUM4160 CNC USB Isolator"
Thanks for the tip, I'll give it a try.
Oh, yes and the bare mains connection and hanging live wires freak me out big time...
What is the name of the software used in the video？
qcad, estlcam, chillipeppr
Thank you
Que bom que fez esses teste. pois eu ia faze-los e ia quebrar minhas fresas. haha! abraços.
the issue is magnetic & rf ... not physical electrons... an opto-coupling circuit for the motors to be isolated from each other, and any VFD P/S you might install.
Seek others' help... but I think your issue is common.
I tried optocoupler and still had the problem. Eventually used bluetooth connection, no problems since.
22:37 - you now know that using a normal drill for cutting is not a good idea. You have to use "millig cutter"?
when you have two powersupplys you must conect Their
2 minus togethers, or all your mesurements will be wrong
11:05 my favorite
Woohoo! Score! Have you got a patreon going? Or a Paypal donations or whatever
2:23 made me LOL :D
you ned one Common minus, not case or Wall outlet Ground.