eBay TB6560 Stepper Motor Driver Boards

[irving2008]

Dan Brown: I have not done any more testing since after removing the damaged chip.

I have tested that all the other axis are still working using my adjustable bench power supply, but I can only run the board / motors upto 15volts on this power supply.

But at this lower voltage my current rig is very slow, max of 200mm/min, for the very short time that the board was running at 36vdc I was closer to 450mm/min rapid move, not cutting.

I will be back to testing the current sence resistor and dip switches.

When using my scope with my bench PSU I always have to use my laptop on battery, as if the laptop on mains or a PC is used I get very strange readings on the scope software, so thats something else I have to look into.

As per my thread above, on 36v the max step rate for these motors is 900steps/sec, thats 270rpm or about 500mm/min on a 2mm pitch screw.

Sounds like you have some ground loop problems. Make sure stepper PSU and scope/PC are on same power socket so they are earth referenced together. I cant use the same PC as the scope and the driver for the stepper, for the same reason...

[Tom]

When I said zener diode I really meant a transient suppressor diode (transorb), such as these from Vishay. But i think clamping to 50v from a 43v supply might be tricky. A zener will clamp, but wont catch transients. you could use a 47v zener as a clamp with a 48v transorb in parallel to catch spikes, with a series low ohm resistor as a current limit between the supply rails. Farnell sell the transorbs, about 30p a go.

Thanks Irving!

That's just the right amount of info... I REALLY don't want to smoke these drivers, so I feel some internet research, followed by electrickery, coming on....
There's time because I'm not cutting in anger yet - just aligning, and debugging....

Cheers,

[Mad Professor]

Today I have been able to find some free time to do some basic current testing using my bench power supply, and USB Oscilloscope.

Using the oscilloscope across the current sence resistor I get the following results.

Bipolar Parallel
DIP Switches 1-2
ON/ON: 0.469v avg / 0.33ohm = 1.421a.
ON/OFF: 0.313v avg / 0.33ohm = 0.948a.
OFF/ON: 0.234v avg / 0.33ohm = 0.709a.
OFF/OFF: 0.078v avg / 0.33ohm = 0.236a.

[irving2008]

Today I have been able to find some free time to do some basic current testing using my bench power supply, and USB Oscilloscope.

Using the oscilloscope across the current sence resistor I get the following results.

Bipolar Parallel
DIP Switches 1-2
ON/ON: 0.469v avg / 0.33ohm = 1.421a.
ON/OFF: 0.313v avg / 0.33ohm = 0.948a.
OFF/ON: 0.234v avg / 0.33ohm = 0.709a.
OFF/OFF: 0.078v avg / 0.33ohm = 0.236a.

That looks about right... i assume by average you mean RMS, so the peak current will be around 2A. Can you post waveforms as discussed?

[RobMar]

Hi everybody, I'm just another new (and perplexed) user of the TB6560 board. (I'm from Italy, so forgive me for my bad english)
Due to lack and inaccuracy of the documentation available for the board, I've searched around and come by chance to this forum, where I've luckily found answers to some of my doubts and saved some chip-smoking (up to now...) thanks to many intrepid experts that shared their experiences. Kudos to you all.
Now, I've started to test the basic functions of the board (a blue 3-axis HYU68 from Ebay) with 3 NEMA-23 2A motors, 24V supply, just testing the basic signals on the 15-pin port (no parallel connection at this time)
The first problem I've encountered is that the pinout table shown on the manual seems to be wrong, with pins 9 to 15 reversed left to right, like if someone has correctly counted pins 1 to 8 of top row and then continued to count the numbers of second row backwards instead of going back to "new paragraph" as is customary on Cannon-type connectors. I say this because I've found the expected 5V and GND on pin 9 and 10 respectively, instead of 15 and 14 as stated in the manual. (the order of the top row 1-8 appears to be correct)
Did someone else notice this? (and also eventual other mismatches on the 25-pin parallel port?)
Thank you for any suggestion
Roberto

[CAD_Rickosz]

Yes and do not not even once go over 24V with nema23 steppers or
you can put out the frying tonight sign just like my bit of electronics did.

[1113562]

Yes there is an error in the manual for the pin outs to the motors but the board silk screen is correct on my board as supplied.

My board came with a parallel cable for the PC and it seemed to all work after configuring the I/O on Mach3 to the settings that came in the chinglase manual! The problem I found was that the DIP settings for the microsteps was not correct. The 1/8 step was reversed with 1/16, a fairly minor problem but still irritating as I had a scaling error on movement until I figured it out.

I managed to buy the correct 2 mm spacing headers and crimps for the limit and eStop input on these boards if anyone is interested. I have a few spare that I can post at mininal cost.

John

[williamturner1]

do not not even once go over 24V with nema23 steppers or
you can put out the frying tonight sign just like my bit of electronics did.

Hi. Great thread! v interesting to read. Thank you guys for sharing.


My Question: Is it possible to establish the cause of the exploding chip? and how to prevent any future damage?

........1. Back EMF caused it?
..................could it be the board has insufficient protection
..................maybe additional protection should be added - as described by irving2008 in this thread

........2. Something fries when a PSU of >24v is used?

or some other reason?

Surely too higher voltage from the PSU would have damaged the chip there and then?
Does the delay in chip exploding (ie when next turned on) hint at back emf damaging it during motor rundown or at the end of last run and that damage only apparent when you next turned the device on and it exploded?

PS. Anyone in the M25 with a CNC set up that I can come around and admire / ask a few questions about their implementation of driver board vs software and motors? Im a newbie starting out. I'll happily pay you for your time in beer!

[1113562]

I don't really understand this as the actual Toshiba TB6560 chips have an absolute max VmA and VmB voltage rating of 40 volts and the operating rage is up to 34 V. Since it was the chip that blew and not something else on the board then I can see it was a pure over voltage problem providing your PSU was putting out less than 40V but preferrably less than 34 V. The Chinese data sheet for the whole circuit board states a voltage range of 16 to 24V for NEMA 23 spec steppers and 24 to 36V for NEMA 34 steppers. I don't really understand why there is a difference in voltage stated for NEMA 23 and 34 providing the current limit DIP switches are set so as not to exceed the particular Stepper motor current per phase for your wiring configuration (series or parallel). The power supply voltage selected just defines how quick the current limit value is reached for each pulse as it is driving an inductive load.

The when the current pulse is turned off the inductance of the winding tries to maintain current flow so generates a large reverse voltage. The protection diodes are there to clip this reverse voltage and absorb the current from the motor in discharging the winding inductance.

The larger your winding inductance the slower it will be before the current limit is reached when applying a pulse causing reduced speed and torque and also causes a bigger turn off reverse current as the inductance stores energy.

A good fast stepper will have low inductance and series resistance.

What is you stepper motor specifications Resistance, Current rating and Inductance in the wiring configuration you are using?

[williamturner1]

The when the current pulse is turned off the inductance of the winding tries to maintain current flow so generates a large reverse voltage. The protection diodes are there to clip this reverse voltage and absorb the current from the motor in discharging the winding inductance.

So is it that Mr ChinaMan's circuit doesn't provide enough protection be it through poor design or poor component choice