[ssashton]

Hi everyone,

I'm currently building a 3D printer and choose Trinamic TMC2160 drivers offering 60V 4.5A. The Trinamic drivers are the go-to choice for 3D printing due to the Stealth Chop mode that is super smooth. They also have 'Spread Spectrum' mode which offers more torque and RPM but isn't quite as smooth as Stealth Chop.

Attachment 444268
Attachment 444270

To compare I wanted to drive my regular CNC router from the new Trinamic drives and compare with my existing regular CW8060 drivers 80V 6A. These drivers are not that new and I think there are a number of similar designs going around with slightly different part numbers. Inside, the main micro-controller has no part number I've got a similar smaller driver that uses a 56F8013 DSP chip. Though this is a programmable item I Think, so it all depends on the code anyway.

Attachment 444272
Attachment 444274

However, here is the comparison. All moving at 600mm/min. I was testing the Trinamic on 30V from an adjustable power supply which has a fan, with Stealth Chop you can hear the fan ramp up when the motor draws current more loudly than you can hear the stepper!!

> CW8060 1/5 Micro-stepping - Google Video <

> CW8060 1/125 Micro-stepping - Google Video <

> Trinamic 2160 Spread Spectrum - Google Video <

> Trinamic 2160 Stealth Chop - Google Video <


Clearly, the Trinamic drivers are far smoother than even 1/125 micro-stepping. Particularly in Stealth Chop mode, but that does limit power output for the luxury.

So it makes me wonder - are there more modern stepper drivers that offer a similar level of smoothness as this, but with 80V 6A? Have I always been missing smooth stepper drive just because I cheaped out and didn't buy genuine leadshine, or what?

[NIC 77]

with my existing regular CW8060 drivers 80V 6A.

I don't know anything about those drivers, and I'm not an expert on the subject of drivers.

I know from reading on the forums that the TB6600 drives, and ones similar to it, are garbage....there is always someone complaining about them or trying to get them to work. But I don't know anything about the CW8060.

My personal view is to buy either Gecko, Leadshine, and now I am thinking about adding Trinamic to that list. It's just not worth the hassle, or risk, IMO, to get a bad driver.

For my 3d printer project, I have decided on Trinamic for any stepper motors, but may use leadshine AC servo drivers as well. I'm not sure. Also, the Trinamic driver you are using was integrated by "Big Tree" but I might use a Duet 3 board with integrated Trinamic drivers. I don't know how this relates to the overall quality of the system, comparing Big Tree with Duet, when they both use Trinamic at the core.

However, here is the comparison. All moving at 600mm/min. I was testing the Trinamic on 30V from an adjustable power supply which has a fan, with Stealth Chop you can hear the fan ramp up when the motor draws current more loudly than you can hear the stepper!!

That's pretty cool! Nice comparison videos! Hey, you've got rotating nuts! Too cool! How fast can you push this machine? Have you thought about changing up the gearing to make it faster?

Clearly, the Trinamic drivers are far smoother than even 1/125 micro-stepping. Particularly in Stealth Chop mode, but that does limit power output for the luxury.

Smoother or quieter? I am guessing that the movement is better in general.

Stealth Chop mode reduces the torque?

What would be great to see is a comparison where, using the same power supply for each driver, you bump up the speed and acceleration of the machine to see which one starts missing steps first. That could be another good comparison of each driver / mode.

Also checking the temperature of the stepper motors when idle or after a period of movement would be a good comparison.

So it makes me wonder - are there more modern stepper drivers that offer a similar level of smoothness as this, but with 80V 6A? Have I always been missing smooth stepper drive just because I cheaped out and didn't buy genuine leadshine, or what?

Smooth or quiet? My understanding is the Trinamic drivers are the only ones that are so quiet. It might even be proprietary.

The Gecko drives I have would sound like they were playing midi music when I used them. I am guessing Leadshine is in the same group. I think many people started buying leadshine because they are a bit less expensive than Gecko Drives, provide some different options, and have good reviews about quality.

I only have experience with the G201X Gecko drives. And they are noisy, but not in a bad sounding way. It's like midi music, lol. I plan on using the ones I have on another project. I'm very happy with the Gecko drives. I never had any problems with lost torque, or loosing steps, or motor overheating, or anything negative. But they do make "music".

I think Gecko is a fairly small company compared to Trinamic, and they have served the CNC community well for many years. I certainly don't have any complaints about them. And Mariss is a regular on these forums.

But these newer higher voltage and current drivers from Trinamic might be a game changer. Particularly when it comes to driving Nema 23's, and not just Nema17's common in 3d printers and other small machinery, like photocopiers. Also, correct me if I'm wrong, but the Trinamic drives you bought cost $20 on Aliexpress? That's about 1/3 to 1/5 the price of a comparable Leadshine or Gecko product.

Some music played on stepper motors, just for fun (not my video, I just found it):

https://www.youtube.com/watch?v=-yEuYGodNLo

[ssashton]

The CW8060 is not a super cheap item, so I guess it's not total rubbish.

https://www.cnc4you.co.uk/Stepper-Mo...tepping-CW8060

In terms of reliability they have never missed a beat and I've run my machine pretty hard over the years. Trinamic ones seem like a bargain in comparison. Wish they did even higher voltage ones!

I'm sure the CW8060 will drive faster than the Trinamic just because they run 80V and 6A, while the Trinamic is 60V 4.5A max.

I think the sound and smoothness is the same thing since what we are hearing is the steps of the drive signal. The steps of the signal approximate a sine wave and with higher micro-stepping that approximation gets closer, hence smoother sound with higher micro-stepping. Ideally we want a perfectly smooth drive signal and resulting smooth rotation (or as close as it can get). Even with a perfect drive signal there will be torque fluctuation due to the design of stepper motors, but I think the results of the Trinamic drives show we are nowhere near getting the max potential from these systems.

The Duet board does use the (almost identical) Trinamic TMC5160, but this chip and the TMC2160 is not like some of the less powerful chips that is all integrated in one case. It is designed to be used with external power transistors, thereby allowing the designer to scale the systems power (max of 20 amps at 60V). That's why the Duet, although using the same chip, is only rated at 4A 32V. You will really struggle to drive a Nema 23 with only 32V.

Oh, I get about 8000mm/min while still being super reliable. I was running 10,000mm/min but found occasional step losses over time. I don't think it really matters though, since I usually don't cut faster than 4000mm/min. The main advantage is avoiding whip on the long ball screws. Not really necessary on that shorter axis to be honest.

Also something is not perfect with the nuts, I think either the ball nuts are worn or some part of the rotating nut assembly is slightly non-concentric. I need to investigate it.

[peteeng]

Hi Sash - Where do I buy the Trinamic drivers or at least look at their specs and costs? Do they have to be paired to their motors or can any motor be used? Peter

[ssashton]

Hi Sash - Where do I buy the Trinamic drivers or at least look at their specs and costs? Do they have to be paired to their motors or can any motor be used? Peter

I bought them from the Big Tree Tech store on AliExpress. However you can learn most from the Trinamic datasheet for the chip.

Aliexpress seller - Home

TMC2160 Datasheet - https://www.trinamic.com/fileadmin/a...et_Rev1.06.pdf

[NIC 77]

The CW8060 is not a super cheap item, so I guess it's not total rubbish.

https://www.cnc4you.co.uk/Stepper-Mo...tepping-CW8060

In terms of reliability they have never missed a beat and I've run my machine pretty hard over the years. Trinamic ones seem like a bargain in comparison. Wish they did even higher voltage ones!

I'm sure the CW8060 will drive faster than the Trinamic just because they run 80V and 6A, while the Trinamic is 60V 4.5A max.

There's a difference between cheap and inexpensive. LOL.. I'm suggesting the Trinamic drivers may be inexpensive but not cheap. Apologies if I cast any negativity on your CW8060's. I don't know anything about them really. Does that make me a driver snob?

I think part of it comes down to motor choice. You may be able to source a motor that will run better on the Trinamic, even though it is running at a lower voltage.

Also, the max current only applies to low RPM, so at high RPM you can't draw the same current. That's my understanding, and it's a generalization. So for a given Nema 23, the overall performance difference between 4.5 and 6 Amps (max available) might not be that much, or anything, when the weak point in performance is often trying to decelerate the machine when you are already at a higher RPM. Of course, it depends on how the machine is geared and other factors. You know what, that's just a general understanding that I have, correct me if I'm wrong.

I think the results of the Trinamic drives show we are nowhere near getting the max potential from these systems.

I agree it's a potential game changer.

You will really struggle to drive a Nema 23 with only 32V.

That is exactly what I was thinking, and why I was so happy to see the link you posted to the ones you bought. I may end up buying the same ones, which is one reason why I'm so interested in your threads.

I could still use a Duet 2 or Duet 3 board in that case, with an expansion board for the external drivers being run at a higher voltage.

Now I'm not so sure that I need the higher voltage on all my axis.

The Duet 3 board can go up to 32V but they only recommend using 24V input due to back EMF.

But if you look at this graph for a Leadshine Nema 23:

Attachment 444332

You can see that the difference in available torque at a given RPM isn't that much between 24V and 48V. Now there may be some other factors here, like better maximum angular acceleration, or something like that at the higher voltage that I'm not seeing. And that has me wondering.

I'm also wondering about power usage, and if running at the lower voltage might be a more efficient solution for a machine that may be running for several days at a time. Especially for the Z axis motors.

Oh, I get about 8000mm/min while still being super reliable.

That's decent. Your tests were shown at a much slower speed, so I was curious.

The main advantage is avoiding whip on the long ball screws. Not really necessary on that shorter axis to be honest.

Also something is not perfect with the nuts, I think either the ball nuts are worn or some part of the rotating nut assembly is slightly non-concentric. I need to investigate it.

That's still pretty cool that you pulled it off. Many people want to do it but never end up making it work. Is it your own design?

[ssashton]

Apologies if I cast any negativity on your CW8060's.

No worries I didn't take it that way. I just meant they are certainly better than TB6600.

Also, the max current only applies to low RPM, so at high RPM you can't draw the same current.


That is exactly what I was thinking, and why I was so happy to see the link you posted to the ones you bought. I may end up buying the same ones, which is one reason why I'm so interested in your threads.

...

But if you look at this graph for a Leadshine Nema 23:

Attachment 444332

You can see that the difference in available torque at a given RPM isn't that much between 24V and 48V.

Inductance is basically just resistance that increases with frequency making a upward trended curve. You can consider that the motors resistance increases as the speed increases. You probably remeber Ohms law; voltage * current = power. We get that classic triangle of relationships.

Attachment 444620

Take an example of a driver with a maximum output of 24V and electronic current limiting to 5A. At low RPM the motor has low resistance so 24V is easily enough to push the full 5A through the motor until current limiting kicks in.

As the RPM increases the motor has higher resistance. At some point 24V is no longer enough to push 5A through the motor we see torque begin to drop as RPM increases.

If the motor has low inductance (read low resistance at speed), it will not require such high voltage to push the same current though it. In the graph above I think it suggests the driver is reaching the electronic current limiting point with lower or higher voltage.

That's still pretty cool that you pulled it off. Many people want to do it but never end up making it work. Is it your own design?

I can't claim the fame for that, it was designed by a guy on MYCNCUK.com

[NIC 77]

Inductance is basically just resistance that increases with frequency making a upward trended curve. You can consider that the motors resistance increases as the speed increases. You probably remeber Ohms law; voltage * current = power. We get that classic triangle of relationships.

Attachment 444620

Take an example of a driver with a maximum output of 24V and electronic current limiting to 5A. At low RPM the motor has low resistance so 24V is easily enough to push the full 5A through the motor until current limiting kicks in.

As the RPM increases the motor has higher resistance. At some point 24V is no longer enough to push 5A through the motor we see torque begin to drop as RPM increases.

If the motor has low inductance (read low resistance at speed), it will not require such high voltage to push the same current though it. In the graph above I think it suggests the driver is reaching the electronic current limiting point with lower or higher voltage.

Thank you for the explanation. That makes sense.

I think the available torque at 24V would be good enough for my purposes using those low inductance motors.

[rodw]

Italian made Lam Technologies DS1076 will give up to 90 volts 6 amps https://www.lamtechnologies.com/Prod...=EN&idp=DS1076
Their DS1078 will give you 160 volts 8.5 amps

I have some DS1076As I am in the process of installing. These accept AC input up to 65 volts which will give about 90 volts to the drive but there is no need for a rectifier on a toroid power supply.
Probably the best place to buy is at
https://www.mechapro.de/shop/Schritt...tufe::206.html
but I bought mine from the Australian distributor. I don't think they have an agent in the US.

Don't forget to buy the programmer!

[asuratman]

Italian made Lam Technologies DS1076 will give up to 90 volts 6 amps https://www.lamtechnologies.com/Prod...=EN&idp=DS1076
Their DS1078 will give you 160 volts 8.5 amps

Do you think this DS1078 will give higher resolution or better result compare to common chinese cheap stepper driver eg. DM860H ?

[rodw]

Do you think this DS1078 will give higher resolution or better result compare to common chinese cheap stepper driver eg. DM860H ?

I'm certain of it. But I need to get them installed to confirm. I have seen a plasma gantry video running 50 metres/sec on 2 amp nema 23's that uses them. I've got onto some Sanyo Denki NEMA 24's which seem perfect for our machine on paper and they should pull 5.0 m/s^2 acceleration to 19 m/min on a 90 volt power supply! Thats over 0.5 G! Plus I can use the smaller 3 amp DS1073's with them as they are only 2 amps.

[BBMNet]

Hello.

I currently use JMCs 2DM2280 step drivers which among other benefits include the power supply internally. Their operation is rather smooth and quiet when compared to my previous selection, CWDR06CA which also had the benefit of including the power supply internally as well.

While they both can be set by the external switches as in regular drives, the 2DM2280 gives you access to more parameters through an external programmer or an RS232 communication link. In my case I have not needed to tangle with the internal configuration. A total of 22 parameters can be fine tuned.

In any case much of the final behavior of the motors relies on the motors themselves. I have noticed that different brands while advertising similar products when actually installed do not behave the same. Some are quieter than others while some are smoother. Some even vibrate as if no shaft balancing had been done.

I hope this information is useful to someone.

Regards.

[ssashton]

I think it's a shame driver manufacturers do not specify performance of the drive in terms of waveform distortion from an ideal sine. Like audio amps specify the harmonic content as 0.1% for example.

[NIC 77]

I think it's a shame driver manufacturers do not specify performance of the drive in terms of waveform distortion from an ideal sine. Like audio amps specify the harmonic content as 0.1% for example.

Are you still happy with the trinamic drivers you bought from Big Tree Tech? I'm still thinking of buying some myself.

[jfong]

Here’s my test of the TMC2160, different version from yours but the output waveform should be same. It’s a very good stepper driver chip.

https://embeddedtronicsblog.wordpres...06/24/tmc2160/

Yes stealthchop is quieter but limited rpm available. Be good for a belt driven 3d printer but not enough rpm for a lead screw/ballscrew setup. I use spreadcycle.

The most effective Trinamic feature is microplyer. This takes the step input and does 256 microstepping. It really smoothes the output current waveform as seen on the oscilloscope images.

Other driver manufacturers have this similar feature. Higher end leadshines work similar. The output current waveform is very similar to the Trinamic. Smooth and quiet running. Leadshine calls it multistepping. Not all Leadshines have this feature and many “fake” versions don’t work nearly as well.

https://embeddedtronicsblog.wordpres...river-testing/


The newest geckodrive g214 stepper driver also has this type of feature. They call it sub-microstepping. I only have older gecko drives which are louder but do run very well. Very high performance with up to 80volts. Need to order a g214 to test but it is expensive. I have a bunch of older geckos sitting un-used at the moment.


Compare the output waveform to a cheaper TB6600. You can see the difference and why the Trinamic microplyer works incredibly well.

https://embeddedtronicsblog.wordpres...river-testing/

[jfong]

Do you think this DS1078 will give higher resolution or better result compare to common chinese cheap stepper driver eg. DM860H ?

Not necessarily. Many of the industrial drivers are still rather old tech. Many are designed years ago but are still in production. They do usually perform really well. They may not be the quietest or smoothest.

[jfong]

To NIC77

I have a duet clone with the Trinamic TMC2660. It’s still in the package but will do a bench test soon. I think you will like the Trinamic chips for a 3D printer. I currently use Trinamic tmc2208 drivers for a laser, about 2years now.

Tmc2208 output waveform
https://embeddedtronicsblog.wordpres...river-testing/

I did the cohesion3d bench test, qualifications, burn in and smoke test for their new Trinamic TMC2224 laser driver board before they released it.

All my other cnc machines use geckodrives. At last count, I have 19 of them. I like them. May not be the quietest but they are solid and dependable.

The duet is going to be a retrofit for a maker select which has noisy a4988 drivers.

Jim

[NIC 77]

Jim,

Thanks for your input!

Honestly, much of the electronics stuff is above my head. You clearly understand it better than I ever will.

If I use a Duet 2 board, I'll need to go with an expansion board and external drivers with more current capacity, which is why I am interested in the external trinamic drivers also that have more current capacity

If I use a Duet 3 I am limited to 24V, but 6A instead of 2A for the onboard drivers.

I'm not sure, but I'm wondering if the Duet 2 is limited to 5 external drivers expansion capability (can you add two expansion boards to a Duet 2)? That might limit some of the future plans I have like dual pellet extrusion, and I may need to go with a Duet 3.

My printer project is going to use some heavy duty hardware, surplus linear stages, it's overkill, but I shouldn't have to worry about adding some serious weight in the forum of dual pellet extrusion to it later, or turning it into a router if I get sick of 3D printing.

So my stages came with 300W servo motors, I have a thread about it here:

https://www.cnczone.com/forums/servo...-software.html

I was thinking of trying the Leadshine ACS806 AC servo drivers using step and dir inputs from a Duet board, but I don't know if it will work. I was reading yesterday about sinusoidal and trapezoidal AC servo drives, and it was making my head hurt. Any advice or opinions you could add to my thread would be appreciated!

Also looking forward to hearing the results of your bench test on the Duet 2 you have!

[jfong]

NIC77,

Not really sure about the expansion capabilities of the different DUET boards. I’ll find out soon though.

I have a bunch of brushless drivers. Applied Motion, Copley, AMC, Baldor etc.

Some of them can be setup either trapezoidal or Sinusoidal commutation. From what I’ve read , sinusoidal is better so that’s what I choose. Better low speed performance.

You must be building a rather large 3D printer to require or want brushless servos. I was thinking of using brushless too on my printer build. Only because I have a bunch sitting around. I’m not in any hurry to finish it.

Jim.

[NIC 77]

Not really sure about the expansion capabilities of the different DUET boards. I’ll find out soon though.

Please post your findings when you do. I'd be looking forward to reading about it.

I have a bunch of brushless drivers. Applied Motion, Copley, AMC, Baldor etc.

Yes, I was looking at some EBay adds for Applied Motion drivers that specifically said "Sinusoidal" on them, whereas what I've read about the Leadshine ACS806 drivers leads me to believe that they are Trapezoidal, although I am not certain. I just know that it sends a trapezoidal signal when tuning the motors.

http://www.leadshine.com/UploadFile/Down/ACS806sm.pdf

There's lots of trapezoidal looking things in the above manual for the ACS806.

What set me thinking about it were a couple of threads I read on the forums:

https://www.cnczone.com/forums/diy-c...ry-router.html

https://www.cnczone.com/forums/bench...cnc-posts.html

https://www.cnczone.com/forums/bench...4-cnc-cad.html

All of the slides are driven by 4 pole 300 watt brushless Tamagawa servomotor. These require sinusoidal drive or they run very rough.

And it looks like I have the same motors he does.

Some of them can be setup either trapezoidal or Sinusoidal commutation. From what I’ve read , sinusoidal is better so that’s what I choose. Better low speed performance.

That's what I've read as well. The thing is, if I can't find appropriate drivers for a decent price then I may be better off spending the money on something like clearpath SD servos where the drivers are built in. Still kind of pricey, and it's a shame when I already have the brand new, unused Tamagawa Seiki's sitting there in front of me. I can get the Leadshine ACS806 for around $100 USD each. I may just go with some inexpensive Nema 23's. It's all still up in the air.

Also, it goes back to something you said in a previous thread, and I would be running them at 80V instead of 200V. The stages have a rated speed of 1200 mm/s but I would be very happy with 400mm/s.

I am considering running them at a lower voltage. The link below is an old thread some of this was discussed previously. I'm also looking to find a Kv rating for these servos, which is explained by jfong in post #11 in the thread below.

https://www.cnczone.com/forums/servo...33064-cnc.html

That's a quote from my thread here:

https://www.cnczone.com/forums/servo...-software.html

I'm a bit of an idiot when it comes to servo electronics, so as always, any advice is appreciated!

You must be building a rather large 3D printer to require or want brushless servos.

Yes, it's going to be a stupendous machine. Travels of 800mm (gantry) x (400mm to 600mm, I still haven't finalized the table design) x 600mm (Z). Dual lifting column, rising gantry (counterbalanced), moving table design. Also fully enclosed.

For the 3 stages I have already, size 25 NSK linear rails, with 4 bearing blocks and 2 rails on each stage, and ground 15mm OD ballscrews.

I figure that with two pellet extruders, if they each weigh 10 lbs or more, I'd still be more than fine to maintain whatever speed and accelerations the extruders could keep up with, having a very rigid machine.

The rising gantry moving table design keeps the moving parts relatively light for the constant X and Y accelerations of a 3d print that could take literally days to complete.

Plus it gives me a good base to turn the machine into a general purpose router / light duty aluminum mill, should I decide to do that instead, and I have a number of projects and potential mods I might do in the future.

ssashton, sorry for the hijack.