Stepper motor driver - which one is not to get - avoiding TB6600 at all cost
Vložit
- čas přidán 26. 08. 2024
- #CNCPlasma
#DIYCNCMill
#3018CNC
#Fusion360
#Mach3
#Tutorials
I upgraded my 3018 CNC using 3 TB6600 stepper motor drivers because they are cheap and the current rating seemed good (more than enough for any nema 17 motors).
I started noticing some drifting during some long milling jobs. The drifting showed more clearly in the X and Y axis (cannot really tell in the Z axis).
This problem didn't occur when I used the same stepper motors with the original stepper motors on the original board of the machine (very small and not powerful).
I decided to swap the TB6600 with 3 of the DM542 stepper motor driver. They are bigger, more expensive but they worked as expected. The drifting disappeared after the change.
So, please stay away from the TB6600 stepper motor drivers. They are cheap but you get what you pay for - problems!!!
Thank you for turning the volume down when the machine was cutting, THANK YOU VERY,VERY MUCH. It’s bad enough having to wear hearing protection when I’m using my machine, I don’t want to wear it when watching an instructional video. Thank you. And thank you for the advice and the great videos I’ve seen you do so far, well done 😎👍🏻
As a self-taught practitioner of digital and power electronics and indeed of motion control for nearly 40 years, I have a few comments:
In a stepper motor system, it seems relatively straight-forward: Connect your microprocessor step and direction signals to the corresponding inputs on a step-driver module. Apply power (!) and you're good to go. It's basic stuff to those who have done some study and practice in electronics, but evidently mysterious to newbies who follow the typically incomplete - and sometimes downright incorrect instructions lying around on the web. Including the (so-called) 'documentation' from many of the hobby and maker suppliers. There are numerous 'fine points' to learn about stepper motors, drivers, signals and interfaces which need to be covered to get good results. These include the power supply, power distribution, grounding scheme, logic levels and drive capability, noise immunity, and a host of other factors. This knowledge is clearly missing here for the most part. Therefore I recommend you all NOT give any weight to pronouncements about this is good/this is bad.
I DO recommend that if folks want to 'do' electronics, they should make a serious study of the subject.
I indeed agree with you about that, but believe it or not this case is true and i trust this man,the driver is actually bad but it might be just a bad particular lot or cloned driver, I'm saying this because I'm having the same problem currently, i just built my cnc plasma table and I'm using NEMA 23 motors with gear boxes, for y-axis I'm using two motors one on each side with a dm542 driver for each one, for my x-axis the load is too light so i used a tp6600 driver at a current of 2.5amp, while trying cutting some samples a small drift was noticed but i didn't give much care ,
now when i tried cutting a 2m x 0.7m entrance door the cuts were shifting in a concerning way, now when the machine finished i found a really bad job accuracy, homing the machine showed the x-axis (tp6600) shifted around 1.5 cm WHICH IS A LOT while y-axis (dm542) had absolutely 0 cm shift.
By the way this CZcamsr helped me a lot with his videos, but the only true reason why I'm here is because i just googled " tp6600 are they any good? " and this video came on top of the list .
the tp6600 driver has a really good reviews but some people say the chip inside might get manipulated and might not really be a tp6600.
By the way, by your experience could there be a solution for such a problem or do suggest changing the driver?
Hello, i want say that you are totally right, the main issue is that this driver is very sensitive to interference, my case was solved with a filtering capacitor at the pulse inputs of the driver.
@@osamahnajjar2824 Tremendous! Glad to hear you have success! And thank you for your commendation. By the way, industrial-quality stepper drives usually have opto-isolated signal inputs to deal with the noise issue. As well as ground loops. In your case, you may want to protect other inputs to the drive (if any? - ah Reset perhaps?) with capacitor as you have done with Step and Direction signals. Good fortune !
@@gyrogearloose1345 Thanks for the advice i will be doing that, as for the opto isolators, the tb6600 has them for sure as i have seen online, but for the pulse inputs there is a special isolator different from the direction and enable inputs, that one i think is supposed to be a very fast isolator as it deals with high frequency step signals and maybe that's why it is more sensitive by it's nature,
By the way in my case the interference was coming from the start arc of my plasma machine, even though it's duration is just a split of a second the drift is unnoticeable but when the design has more than 300 starting points the drift accumulates to more than 15mm.
@@osamahnajjar2824 I totally agree with Gyro Gearloose. The basic concept how this works is really simple, but since low signal and power electronics are combined things get very complicated easily. This is why understanding how electroincs work, how EMI can effect things, what happens with a piece of 10cm wire, which is supposed to be a solid connection...
The chips may be counterfeit, I don't want to debate in that, but even is they are just more sensitive, with proper scheme it still should work.
May very well be the case that not (just) a capacitor would have needed in those sensitive inputs but some loading resistor to decrease the input impedance which will also make harder for EMI to induce levels that disturb the real signal.
Optocouplers won't magically make everything immune to noise, when you tie it's groud point right to the PowerGND. It might take just a few mA to activate the inputs and if some unwanted current induced either in the ground wire or the signal wire, it still can disturb the signal.
Good practice is to use shielded wires for both the motors and the control signals, and still do not route them parallel in the same loom. Distance is the most guaranteed EMI reduction possibility. Separate in space the high power parts and it's wiring, including their supplies from the small signal control side.
Grounding is a very hard topic. I do not like the term "ground loop", because it suggests that the problem is with the loop. No, it is not. You can make a very unstable wiring with a star grounding just as well.
The whole grounding problem is caused by just a simple issue: the receiver side of a signal interprets the signal according to it's own ground. If it isn't the same as the sender side's ground, then it is obvious that something will happen. In analog circuits, it causes a continuous measureable offset, noise, and all sorts of stuff, but might be the bandwith limits it to an acceptable level at least. In digital circuits, it might just work most of the time, as long as there isn't some transient that push a signal out of the levels. And usually there is such a transient, if the ground wires are long, unshielded and other currents than what belongs to the signal flow through the wire. High frequency switching signals easily induce enough noise, and digital circuits are usually very fast, fast enough to capture these pulses.
The best case would be when every signal would have been carried through a balanced differential line, this would eliminate a lot of EMC/EMI disturbance. But when this isn't available, the second best option is the optocoupled inputs. But that only works good, when you supply your own small-signal ground, and not tie it together with PGND, nor it tied together inside the driver. If it is, then it is an inherently bad design, the optos doesn't help much noise-wise other than protecting the input from overvoltage.
Long story short, this is easy to understand, yet very hard to solve it right and things get very complicated fast as you increase the number of interconnected modules, especially when not you are the one who designed how they expect connections to the rest of the system.
Thank you very much for this video. I can confirm the problems with TB6600 stepper motor drivers. Drift and step losses up to total driver failure. The additional costs for DM542 are worthwhile in any case. It is simply better for the machine and the nerves.😄
I took apart my first laser engraver thinking the bearing setup was too heavy for the stepper motors because the drivers kept stopping in the middle of any engraving. After found out the problem, it was not my design or stepper motors, it was the drivers all along. I'm still mad about it!!! And yes, just go for at least dm542, still working fine so far.
thank you for posting this video i cancelled my tb6600 and ordered a couple of the other dm542
I'm glad that you canceled your Tb6600 order since other viewers also confirmed the problem (check the comments)
What driver your recomemd for gbrl.... I have problem to milling pcb 0,2mm thanks
It really depends on what type of stepper motors you have on your machine. Anything other than Tb6600 and have the rated current higher than your stepper motors will be fine.
Thank you, you gave me some hope. I also changing right now from the cheap ones to DM516T, hopefully it will solve my problems.
If your problem is drifting, then I think the better drivers will solve it
Thanks for this advice Minh! I came across the very same observation from a Polish reviewer: TB6600 drivers' response may differ from batch to batch. Looks like it's an industrial issue... After all, we get for what we pay for.
(FR) Merci d'être aussi rapide pour relever le manque de fiabilité des TB6600 souvent relevé par d'autres monteurs de diy CNC. Mais j'étais incapable d'imaginer une telle dérive à ce point... grave. On atteint les désagréables surprises du "cheap" ;) Mais ton diagnostic semble coller. Reste à voir ce que ça donne dans le temps ;). Bein! Repose-toi et prends soins de tes proches. Je suis heureux: comme je n'applique pas tout de suite tes modifications, j'ai le temps de les voir murir ;) et heureux que tu sois attentif à leur évolution. Bon week end l'ami!
I got the drifting problem since I was done with my laser engraver, but I never thought that the problem came from the motor drivers. I was thinking my home design of the linear bearings was too heavy. But now, thinking about it, that wasn't it because I was able to get the fast move feed rate at 30m per minutes. It was the stepper motor drivers.
(FR) c'était du temps perdue mais on apprend des choses pour les futurs projets. Au moins c'était pas mon laser qui a été mauvais. Bon week-end mon ami.
Tb6600hg is genuine Toshiba chip
Which comes in tb6600 v1.2 driver , which are without that black casing those black case drivers feels sketchy to me since there's no where written they are tb6600 where else on this driver tb6600 it is actually and written and there is Toshiba chip tb6600hg inside
Hello, i also confirmed and had the same problem BUT NOT ANY MORE!, i had the same problems of drift and every thing but it was for sure less than yours, mainly the problem is some kind of interference caused by another component like the spindle motor (mine was the high frequency starter arc for the plasma tourch), after some hard work the solution was just a filtering capacitor wired in parallel with pulse + and - of the driver, it's value is 100nf (it's marking is : 104) and every thing worked flawlessly and the whole issue literally disappeared,
I hope this helps somebody :) .
Awesome, I'm glad you solved the problem and provided a simple solution for everyone. I don't use those drivers anymore but surely will help others for low cost machines
@@minhsmancave9448 I'm glad to help others back,as it's the least i could do, i really learned a lot from people around the internet and from you particularly, thanks man!
As you confirmed that your solution worked. I'll ask for your permission to make an update video about this driver. I'm sure it will help many people who already bought them. Unless you're planning on making a video yourself.
@@minhsmancave9448 Sure, i will be more than happy about the update, go ahead and I'm waiting for your results 😊
I had the same issue with the TB67S109A stepper driver that they sell it as an upgraded version of the TB6600. I almost got crazy for all the things i have checked and replaced and only one capacitor fixed the problem! I have ordered the DM542 thow cause its working for now but i dont know what else will they do in the future! Thanks again both of you! Akis from Greece
90 % of the time, the problems you experience are form connections:
are all zero volts connected together?
Is your power supply very stable?
Do you use power tools at the same time you use your CNC?
I would still say that the drivers are not good. I found out they are not original Tb6600s. Now, you can buy the real ones but the problem is that you don't know which ones are real. I have been using dm542 and dm556, no problem at all, running my saw, shop vac, etc. While milling, nothing shifted.
It is because there is no TB6600 chip of Toshiba in these drivers. there is a much smaller and different chip inside it.
If you will open it you will see the small chip. calling this driver "TB6600" driver is a big mistake and it is for sale all over the net.
I wish i had seen this video before my build. The machine was very noisy even at standstill. All motors were humming. I thought it was something wrong with my overall design. I switched to DM542 and the machine was silent and ran smoothly. I can confidently tell everyone to stay away from TB6600
I heard about this before I purchased my 6600, since they al only £9 UK (Amazon), I opened myne up to find no dry joints but there is no thermal compound on the small aluminium square, which makes contact to the big aluminium heatsink, perhaps this is the problem?
THANK YOU!!!!!!! I WAS ABOUT TO BUY IT!!! HAHAHAHA you changed my mind for the better
Very glad that you saw the video before deciding buying those. Lot of others, including myself, just have to throw those drivers away. It wasn't just money wasted, it was also time to figure out what was the problem.
That's the notorious problem with the TB6600. They last about 6 or 7 projects and die quickly with the same troubles...Lost steps. I wasted time rewiring everything with upgraded shield cable and ground blocks only to learn the problem was the drivers and not an EMI problem. Improved shielding is a bonus but I could have done that at another time when the budget allowed for it. Listen to the masses. TB6600's are absolute garbage!
Get me some 542 drivers what a big difference I will never go back to the Tb 6600
That's right my friend!!!
Same issue replaced thanks now it's excellent replaced with tb6560
I've been having terrible problems with drift. I'm using TB6600 based drivers. I hadn't considered it could be coming from the driver. Maybe I should try another driver. Thanks!
Thank for the video. Has some problem, but my motors much bigger so I am getting that issued regularly, it's happening after few minute cutting. losing x or y axis position for few millimeters. I am thinking it's from driver. quality or heating problem
bought 4xDM556t for nema 24 hope it will resolve my problem. I will leave additional feedback when I was install them
@@sergeyarustamyan6361It is a little late, but did you got the results?
I agree. Do not use TB6600 I use DM556 digital drivers. Supper quiet.
how did you connect these to the controller
Pulse and direction pins from controller to pulse- and dir- on the motor driver. 5v to pul+ and dir+.
Thanks !!
Yikes! I have the 35$ special you purchased as well :( I am building a larger CNC, using my 3018 and 3d printers. I may go ahead and get two for the X & Y and just keep the 6600 on the Z for a bit. I am still working on the base anyway, so no rush. Thanks for the info and great video as usual.
Yikes indeed. I don't know what to say. Maybe you can just try them first, maybe you will get lucky. If the problem happens to you, you know where it comes from.
@@minhsmancave9448 True and may just do that. No rework really to swap them later. Did you notice any axis that was worse than another?
@@minhsmancave9448 I wonder at the higher amperage if the driver is losing steps due to heat? I know on my first 3d printer (a reprap) that the extruder motor would start losing steps due to heat on the driver and the motor. This happened on long prints. I added a fan to my Ramps board and a heatsink on my extruder stepper motor with a fan. I'll add a fan for the drivers and see if that helps if I see the issue. Thanks again for bringing it to our attention!
I don't think it was the heat problem. Both the drivers and motors were barely warm. With the dm542, the motors are hot but still ran correctly. I set the current at 2a for the Tb6600 and 1.9a for the dm542. So, most likely it's the Tb6600 encoding problem
@@minhsmancave9448 I have set up a test rig and have duplicated the problem. It starts to lose/gain steps quickly. It could be it does not handle noise that well. I have a "control" subject using an Easydriver board and it runs a very long time with no loss or gains in steps seen (yet). I am going to play with som shielding and chokes on the wires and circuit board. Something to occupy my time for a bit :)
Who was the manufacturer of the DM542 driver board please mate.
I have no idea bro, Chinese people, I guess.
@@minhsmancave9448 Just found some for 30 dollars Au? Do you think fake?
I got mine $11 each, they work fine
You are not the first I had heard from about these things. There are upgraded models but, of course, cost more so most people go for these cheap things.edit: The DM542 I have on my 50 watt laser machine on 3.5a Nema 23 motors and never an issue with them and you know a laser is so tiny that if it lost a step, even a single step, it would be highly noticeable.
I'm just mad because it costed me a day trying to see what was wrong with my machine. Changing feed rates, accelerations, etc. Nothing worked.
Been there so many times and it is infuriating. You would think with all of the issues with the TB6600 for all of these years that they would have ceased to sell them but China merchants don't care and they take advantage of people who just don't know. 542, the last time I checked, was 30-35 a piece but that was a year ago.
@@thebeststooge I just got 2 DM542 from ebay for $15 per piece. And on AliExpress, I found some for about $11/piece. Not too bad
WOW, those have really dropped in price as I suspected since so many things use them now (due to the TB6600 causing so many varied issues). My t-nuts for the cnc project will arrive on Monday from China (finally) but I have ceased everything until the panic is over.
@@thebeststooge I have stopped on mine as well, I just hope to get the parts I had already ordered. In the US now though the tracking says!
DM542 not support nema 17 with 200 microstep right?
No reason not to. I used then for nema 17 before
@@minhsmancave9448 so what spec of your stepper motor.
The one with 2A current
Em hiện tại cũng đang dùng TB6600.Cắt hình tròn nhưng không được tròn xoe, vẫn có dung sai 0.5mm. Qua quá trình anh sử dụng anh thấy TB6600 nó bị lỗi gì anh ?
Nếu cắt những Gcode dài thì sẽ lệch dần giữa các lớp cắt. Nếu cắt lâu thì driver sẽ bị lỗi và cắt nguồn.
@@minhsmancave9448 thank anh
Have a source for a low price for the dm542, 24v power supply, and a controller? I am only using Nema 17 motors for now but will go to 23/34 in time if I decide I like it.
I just bought a pack of dm556 for $13 each from ebay. It might take about a bit more than 2 weeks to arrive. If you wanted the same, I can test them first and give you the link. They were priced at £12 and I gave the best offer for £10 and the seller accepted. They should be coming in a few days (just cleared the custom in New York). I don't have one for the power supplies, the cheap ones are usually not reliable.
What type of controller are you going for? If you don't need too many inputs, the RnRmotion board that I'm using is not bad, only cost $11 from China and about $20 from the US. Sometimes I do want to have more inputs so I can have all the sensors installed. 6 inputs will be ideal for me (1 for emergency stop, 1 for tool probing, 1 for limit switches and 3 for homing X, Y, Z. But this controller only have 4, I have to ignore the homing switches for now.
@@minhsmancave9448 Well, I don't think the 556 will work due to the minimum amps since these are tiny Nema 17 motors from a 3d printer I had. Basically 1-1.5a is about as much as they can handle or they get very hot.
The controller is the Mach 3 one you went to with the red board and comes with a disc and cable. I finally found a video on how to hook everything up with that (except for the endstops and motor speed control) after six hours of searching.
Dm542 won't work well with nema34. Nema23 will be the limit. In my last video, the driver overheated and stopped working ( I used max current at 36V on nema 23 425oz). Last time I got 2 dm542 at $15 each on eBay and shipped from the US. $11 each shipped from China. Let me check the link then I'll send both to you.
@@thebeststooge Here's the link to get buy DM542 that I used last time. I can't find any cheaper. The link I bought on eBay is no longer existed. www.aliexpress.com/item/32951509951.html
I am ordering the TB6600 to use, what should I do? Is there no solution?
Not that I know off for now. Some viewers told me that you can try increasing the pulse width to 20us. I can't try since I don't have Tb6600 anymore. I guess that you will find out and let us know.
It seems that most of 6600 driver's are fake so the first step would be to take the cover off and look at the chip and see if it's authentic
Not sure what you're looking for do a Google search for fake TB6600. There's multiple threads on cnczone forum with pics
Mach 3 is best for make small CNC machine!?
For me it is. But other people might have different opinions. All depending on what you're used to or what functions you wanted to use.
The problem is that the original tb6600 have minimum pulse width 2us. You driver is not tb6600 because inside the board ale different chip. Is not original tb6600. And you need change pulse width to more that 20us.
Not many people have a good deal of knowledge about this motor driver like you do. I just chose a easier solution is to buy another type of driver. Dm542 is one example of not too expensive and a lot more reliable.
@Reuel T Good tips, hopefully other people will see your comments and try it. Maybe you should make a video how to solve the problem with TB660s then I can remove my video.
@Reuel T I already have a set of TB6600 drivers and would like to avoid problems. How is the pulse width adjusted? I have plans to make my controller work with an arduino in case I choose to use matching software but I will also have a breakout board on my controller in case I choose to use software that requires a parallel port (Linux CNC). So how would I adjust the pulse width in either case?
I have no idea how to do that with Grbl and Linux CNC. Only mach3 is easy.
how to change this on mach3?
When I saw the 6600 number on the stepper driver, that rang a bell, as there is a huge forum posting on the CNCZONE website about the issues the 6600 drivers have and many solutions to attempt to fix them, so, Minh has been bitten too and solved the problem by nipping it in the bud.....no messing about trying to revamp the 6600 with a quick fix.
For anyone who's interested there is a guy called James Newton who has a website WWW.MASSMIND.ORG and he supplies stepper driver kits for DIY builds using the THB 6064 AH chip.
Kits 2 years ago were going for A$35 each.....he also recommended the DQ542MA stepper driver as a far better one than the 6600 type.
I ordered 4 kits.....4th axis in mind..... and he included for free a pulse generator kit to test stepper motors and run them as motors without a regular stepper driver......that would make a good variable feed motor for a manual mill drive.
I took the bait and thrown away $35. I think at that price, trying to fix those Tb6600 is more expensive than throw them away. I got a few dm542 on ebay for $15 each and also a few more from AliExpress for $11 each. So far so good
@@minhsmancave9448 According to the people on the CNCZONE website, fixing them is impossible due to the bad design of the circuitry but the Chinese sellers still extoll the virtues of them and push them onto unsuspecting buyers......James Newton spent a lot of time attempting to solve the question of a fix but gave it up as a bad deal.
Can make a tutorial on how to build a cnc machine n controller
Look at the main page of my channel, I have 2 playlist that show all steps on how to build CNC machines: plasma cutting table and CNC milling. czcams.com/channels/2WnVQvs841c7HwUIXtaA0Q.htmlplaylists
maybe the problem lies within your supply
Muito bom
Even I am a victim of tb6600
Anh ơi anh giải thích kỹ phần điều chỉnh switch của driver này được k ạ ?
Cách chỉnh switch của các driver đều có ghi trên cạnh của driver. Với DM542, Switch #1, #2, #3 là để chỉnh dòng điện. Switch #4 nếu bật (gạt xuống) là toàn dòng đặt bởi switch #1, #2, #3. Nếu tắt thì sẽ là nửa dòng. Ví dụ stepper motor của bạn dòng tối đa là 2A, thì bạn nên đặt 1.92A, switch #4 gạt xuống. Còn switch #5, #6, #7, #8 là để đặt bao nhiêu bước cho một vòng quay, cái này thì tùy vào máy CNC của bạn.
Agree TB6600, absolute rubbish!
Your looping the power from stepper module to stepper module. Thats your problem. Each stepper module power need to come from the power supply. Do Not Loop The Power From Stepper Module to Stepper Module.
Thank you Martin! This is the first response I have read here that shows knowledge of some of the 'subtleties' of these kind of circuits. It's basic stuff to those who have done some study and practice in electronics, but evidently mysterious to newbies who follow the typically incomplete instructions lying around on the web. Including the (so-called) 'documentation' from many of the hobby and maker suppliers. There are numerous other 'fine points' to learn about stepper motors, drivers and interfaces which need to be covered to get good results. This knowledge is clearly missing here for the most part. Therefore I recommend you all NOT give any weight to pronouncements about this is good/that is bad.
I DO recommend that if folks want to 'do' electronics, they should make a serious study of the subject.
With shielded twisted pair or at least twisted. No daisy chains.