Chinese 3 Axis TB6560 & 300oz @ 24V too slow and not enough power

Hello everbody

First i would like to introduce myself and tell my story.

Fell free to skip the next part if u just want to help :-).

My name is Daniel and i started getting into CNC a while back. My first CNC Mill is the one shown in the first picture :-D. I build it entirely by hand out of Wood and parts from ebay. It did it job more or less (more often less xD) but at least it gave me the opportunity to mill parts for its successor. Since i found CNC-Milling to be extremly usefull and it is a lot of fun i decided to go all in. The result is the monstrosity shown in the second Picture :-). I'm trying to build a CNC-Mill that has it all: a lage working area (600x600mm), precision, (speed) and enough power to maybe mill some aluminum
sometime in the future. The build was long an full off setbacks most of them credited to the lack of precision and speed of the small wooden CNC-Mill. But i finally got it finished. The mechanics work a lot better then i expected. The sleds and the portal are easy to move around and the trapezoidal screws can easy be rotated by hand to move them around. The screws are driven with HTD tooth belts. Do to the fact that i was afraid any motor within my budget would lack the power to move the giant 30 Kg portal around all drives translation have of 2 to 1 to give the hole thing more power at the expanse of Speed. Few days ago i started testing the thing and thats when the problems started...


The electrical Set-Up:

The setup of the electronics is as shown in the third Picture.

I have a 24V 15A Power-supply

An "old" blue Chinese Tb6560 3 Axis 3,5A controller board (used to run the old mill in the past)
The dip Switches are all set to on (max current, fast Decay, no MicroStep)

and 3 Nema23 300oz 3A Motors; 200 Steps for 1 Rotation (Full Name: GB23H282-30-4A Additional Specs: see Picture No. 4)

and an old HP Laptop with MACH3 istalled (also used to run the old mill in the past)


The Problems i encounterd:

1) Speed: The Board is currently set to 1/1 Microsteps /no micro stepping. Which in my setup means 133.33 Steps per mm. Acceleratio is set to 200mm/sec^2 for the speed tests.

At 300 mm/min Motor runs nice and smooth but lacks tourque (i can stop i with my hand, its bareley anough to move the portal)

At 400 mm/min the motor starts turning and then stops shortly after making a buzzing sound and the shaft is jerking a little to the left and right but not turning.

At 500 mm/min the motor goes direktly into jerking and buzzing and turing extremly slowly with no tourque at all


2) Acceleration: as long as the max speed is set at 300mm/min i can turn the accelaration as high as i want to (for example 1000mm/sec^2 does not result in stalling). But i would explain that away with the fact that there is no load attached to the motor so the motor does not experience any resistance when accelerating.

All and all the motor are not only a lot slower then my old set up but also apear to have even less torque then the tiny Nema 17 Motors i used in my old CNC-Mill. My Old setup was running at 500mm/min and an acceleration of 300mm/sec^2 (the old mill also needed 133.33 Steps for 1mm and i used it with 1/8 Microteps, Motors also needed 200 Steps for one complete turn of the shaft). My goal was to build a better CNC Mill not one that is slower and weaker...


Observation and Measurement:

To monitor the Power usage i connected an Multimeter between the Powersupply and the Controllboard. The board itself draws 0.15A with no Motor enabled. I guess most of that runs the small fan and the rest is lost in the 12 and 5V Voltage-regulators. When enabling the one Motor connected to the board 1.5A are drawn from the Power supply. The Current drops down to 0.83A when the motor is run constantly in one direction. If i apply tourque to the shaft of the motor the Current goes up to 2.0A just befor i press down hard enough around the shaft to stop the motor. When the motor is supposed to be running but i stop it with my hand the Current reads 1.5A. All in all it does not draw close to 3A the motor is supposed to draw.


What i tried so far:

I read my way through most of this forum using the search. The "How I fixed my Chinese TB6560 controller (updated)" Thread was very usefull altough very very long :-).

I removed and bridge the Optocouplers as seen in the 5. Picture circled in green. I know it's not the safest thing to do but should it kill the LPT-Port on Laptop it would be sad but i do have other old (and basicly otherwise useless) Laptop to replace this one. The result: No improvement at all. The Laptops LPT is not damaged yet but it made no differnce at all.

I removed the 10k resistors that are supposed to reduce the current running trough the Motor while at a stop as seen in the 6. picture circled in yellow. Resulting changes: None what so ever.

I grounded the heatsink of the TB6560 chips as seen in the 6. picture circled in red. Resulting changes: None what so ever.

I removed the 1000pF capacitors and replaced them with the proper 100pf capacitors. Picture No. 6 circled in pink. Resulting changes: The anoying hissing sound when the motor is not moving is gone. Other than that no changes at all.

I set Mach3 to "Sherline 1/2 Pulsemode". Didn't not improve performance at all.

I varied the Step Pulse length between 0 and 14 µs with also did not improve or affect the performance in any way.

I am still going to try the hole 74HC14 thing to buffer the Clock/Step signal but i had none here and they are supposed to arrive here within the next couple of days. I'm not getting my hopes up...


So my question to all of you is:

Can you please tell me what else i could try?
Is there anyone else with a similar set up (3 Axis Board, 300 or more Oz, 24V Powersupply) who has any insight at all?
Are there any Modifications or Changes in Settings that could Improve the performance?
What do you thing is the Problem? Is it the controll board or the motors? And why?
Should i try another software?
Can the Controllboard not handle more than 600 Steps per Second at 24V? why did it work befor with 12V and Nema17 motors?
Can't the Steppermotors run faster than 180 rpm? Are there faster Motors? Can i make my motors faster?


And please refrain from suggesting i go and buy a gecko controller board unless you think it's the only chance i got because
i really don't have another 300$ to buy one.... :'(


Please Help :drowning::(


Greetings from Germany
Daniel

Daniel,

The reason the steppers are slower and less torque than even the NEMA17's you used, is that according to the specs, these motors need a 74v power supply at 3amps per motor. On it's best day and working flawlessly (which is impossible for most TB6560 boards), the TB6560 simply will not supply anywhere near that current, and the PSU at 24v is way too low. NEMA17's can go up to around 60oz/in, with lower VA requirements, so the current setup you described is closer to ideal for them than to the 300oz/in setup. That's why you can get very good performance and torque out of them. I have a small machine with NEMA17's, and it rocks! You cannot though just re-use the same PSU and controller for much larger steppers, with higher inductance, it just won't work.

If the steppers were able to draw 3A each, the controller would be a smoking mess in no time. If you don't want to go to a G540, you'll at the very least need to bump the PSU up to the max of 36v from memory for the TB6560, and put up with a slight improvement, or go to smaller steppers with a much lower inductance. The calc for working out the required PSU for steppers is: ((the square root of inductance) x 32) = PSUv.

The other issue contributing may also be rigidity in the machine frame itself, and alignment of the rails/drivescrews. It takes very, very, little mis-alignment to create drag or introduce racking, or counter acceleration. And by little I mean almost undetectable. Can you apply pressure to various parts of the machine and have it flex or move? If so, then this will happen under load too! The PSU/VA issue is the first thing you'll need to fix as this a known issue, but rigidity may be something you need to look at and consider in addition.

I know you don't want to hear it, but a G540 really is the way to go, but if the cost is the issue for you, you need to consider how much time, effort and cost you're prepared to spend in solving the issues. As currently you have an incompatible PSU/Steppers/Controller setup.

cheers, Ian

Also, although I have not actually tested this myself, I would expect that a multimeter may not be able to give an accurate motor current draw reading because the current will not be a steady value, rather it will be pulsed.

Hi Ian and doorknob

thank you both very very much. With what you told me thinks are a lot more clear to me. Only having 1/3 of the minimum requiered Voltage for proper operation isn't great news but at least i now know why im having problems with my setup.

The really sad part is that i bought the 3 Motors and the PSU as a set on Ebay. That let me to not questing the adequacy of the PSU. With what you told me it just seems stupid. The combination sold makes no sence at all and i feel really stupid for not questining its adequacy...

Anyway due to bugget limits i'm unwilling to buy a bigger PSU and a controller that could handle the Volatage needed for the motors to run. And since it took me a lot of time to make the nema23 motor-mounts for my new CNC-Mill i'm not willing to give up yet.

Today i found out that by turning the adjustment screw on my 24V PSU so the max it am able to get the Voltage up to 30.8V wich is good news, i think.

The solution for my problems i came up with is: get the 'smallest' NEMA23 Motor avialable. I found this one NEMA 23 Stepper Motor Single Shaft 56mm 4 leads 2HS57-56 | eBay on ebay. Buying 3 would only set me back ~100$ witch is a lot less then a new controller+PSU and i could keep the motor -mouts and meachnikal parts as they are.

With the formular provided by you [thanks, again :-)] the 'small' nema would require ((the square root of 1.1 mH) x 32) = 33,56 V. Are the 30.8V of the PSU enough to properly drive this motor?

Considering the tourque i think 0.5 Nm should still be enough to drive ma CNC-Mill. With my set up i calculated that i should still get ~140 N of force on eache Axis.

Calculated like this: 0.5Nm multiplied by 2 because of the translation from a mall 20 tooth gear on the motorhaft to the 40 tooth gear on the drive-screw makes a tourque of 1Nm at the drive screw. The drive-screw has a diamiter of 14mm. So the force will act upon the moving part with a lever of 7mm. So 1NM divided by 0.007m should result 142N of force. The Portal weighs around 30Kg so at 500mm/sec^2 accelaration it should only take 15N of Force to move that thing. Leaving me with 100+N to spare for drag and what not. Considering the fact that the Portal and sleds are easy moved by hand that should be enough, i think.

And the 'small' Motor only needs 2.5A which the tb6560 board should be able to handle better, i think.

But you have a lot more knowlage and a lot more expierince than me. So what do you think? Will the PSU Voltage be enough? Any gut feelings about the tourque?

Thanks again for the Help rendert so far.
Daniel

double posted the last post. Sorry

Daniel,

The stepper you linked to last is going to be a much better match than the ones you currently have, so even if it's not ideal, the PSU and current the controller can supply will be close enough that you should see a dramatic improvement! Great looking frame you built by the way!

cheers, Ian

Daniel,

If you use 30v, be sure you incorporate some way to sequence the power supplies. Otherwise, the 6560 might self destruct. The simplest way is to use a separate 5v and then always remember to turn the 30v on/off before the 5v.

Quote:

---

Originally Posted by aarggh

Daniel,

The reason the steppers are slower and less torque than even the NEMA17's you used, is that according to the specs, these motors need a 74v power supply at 3amps per motor. On it's best day and working flawlessly (which is impossible for most TB6560 boards), the TB6560 simply will not supply anywhere near that current, and the PSU at 24v is way too low. NEMA17's can go up to around 60oz/in, with lower VA requirements, so the current setup you described is closer to ideal for them than to the 300oz/in setup. That's why you can get very good performance and torque out of them. I have a small machine with NEMA17's, and it rocks! You cannot though just re-use the same PSU and controller for much larger steppers, with higher inductance, it just won't work.

If the steppers were able to draw 3A each, the controller would be a smoking mess in no time. If you don't want to go to a G540, you'll at the very least need to bump the PSU up to the max of 36v from memory for the TB6560, and put up with a slight improvement, or go to smaller steppers with a much lower inductance. The calc for working out the required PSU for steppers is: ((the square root of inductance) x 32) = PSUv.

The other issue contributing may also be rigidity in the machine frame itself, and alignment of the rails/drivescrews. It takes very, very, little mis-alignment to create drag or introduce racking, or counter acceleration. And by little I mean almost undetectable. Can you apply pressure to various parts of the machine and have it flex or move? If so, then this will happen under load too! The PSU/VA issue is the first thing you'll need to fix as this a known issue, but rigidity may be something you need to look at and consider in addition.

I know you don't want to hear it, but a G540 really is the way to go, but if the cost is the issue for you, you need to consider how much time, effort and cost you're prepared to spend in solving the issues. As currently you have an incompatible PSU/Steppers/Controller setup.

cheers, Ian

---

Sorry I'm so late to the party, what a great piece of information here !
Could I ask you for some advise on my intended setup ?
It's 3 nema 23, 4A, 1.8mh inductance, 283 oz/in (unipolar)
Dual Shaft NEMA 23 Stepper Motor 2 83nm 400 oz in 4 0A Unipolar Bipolar | eBay
3 Drivers M542, and PSU based on your formula, should be around 42v but I don't know the necessary amperage.

Could you advise if these parts look compatible or what would you recommend ?

Thanks so much !

If it's a 4 amp motor, give it 4 amps for best performance.

With answers like that I feel like I'm being bullied just because I'm a newbie, lol.

After further investigation, it looks like the power supply needed for driving 3 Nema 23 motors, 4A, 1.8mh inductance, is in the order of 48v 8.5A

I'm sorry you feel that way; these things are not always altogether obvious. The nameplate voltage on a stepper motor, for instance, is only indirectly related to the amount of voltage you're supposed to feed it. A motor labeled "2.5v" might take 48v to run right. But the amperage ratings are real. To calculate the amount of amperage you'll need, figure two motors pulling a full load simultaneously, and add a little for a comfort margin, in case a third motor should kick in. So 8.5 amps should work; 10 amps would be safer.

Thanks for the help :)

I'm going to make a prediction here: You will be back in a few months or weeks or days telling us how the TB6560 driver suddenly stopped working one day. First, it's not designed for higher voltages, as H500 indicated the power up sequence has to be changed to get logic power first, which is difficult to do reliably. Second, spending $100 to downgrade the motors to keep the same driver is backwards. Instead use that $100 to upgrade the drivers. If you can't afford a geckodrive, save some money by buying and building a kit. Not only will you have a better driver than the TB6560, for less money, you will also know how to FIX that driver if you have problems with it in the future. Obviously, I recommend my drivers, which would be under $100 for three and will give you 4 amps at 50 volts or 200 watt capacity. More than enough for those motors to work well... want proof? Here's a local user of my drivers who upgraded his chinese kit:
https://www.youtube.com/watch?v=vtrHp7_FIi0

I'm not saying this to be mean or to bully anyone... you do what you think it right and I'll support your decision. I'm saying this because I've seen it so many times before. TB6560 drivers are NOT intended for this use, they eventually die, and good drivers at the right voltage make all the difference.