tb6560 driver issues

Hello,

this project is my first time working with steppers and drivers for them, so I ran into a bit of trouble. I bought these tb6560 chips and made my own pcb with them. I am suppose to send control data from an Arduino to it. I want to start slow, but not even the basics seem to work for me...

I leave the motors disconnected and measure the output: it does the stepping procedure correctly. But when I connect the motors, they block (can move them with the hand no matter how hard i try) and make a hummm noise, and the voltage on them (the outputs of the tb6560) are a constant 1 to 2 V, no change, no stepping... It also sucks a lot of current (3A for one motor.)
I tried 2 motors, NEMA 17, one 0.7A 1.8V, one 1.8A 1.7Ohms, and the same thing happens. I'm also afraid i will burn my arduino or something... the board i made is really simple, no protection or anything, the application i want is in itself very simple, I don't need microstepping or all the other advanced stuff the chip can do.
The motors work ok using a L298, again controlled by an Arduino and powered from the USB.

From what I read, the only things I did that might cause problems would be the facts that I connected the sense pins to ground (but others did it and it works for them...), and that I connected the reset and enable pins to 5V (so the startup sequence might not be respected - check the tb6560 datasheet)

Any ideeas of what I am doing wrong? Any solutions?

I have attached the Eagle files of my simple driver thingy.

Thank you

The TB6560 needs resistors on the sense pins, if you connected them to the ground the chip will not limit the current, destroying itself if your power supply is not limited in current.
This trick can work only with motors designed to be driven without current limitation, with a higher coil resistance.

TB6560 are also very delicate, they don't forgive any errors, I fried one by connecting the motor wires in the wrong order, one other by disconnecting the 5V logic supply before the 24V power, also two other for unknown reasons (I suspect bad contacts on the motor connector).

PS : upload your shematic in or PDF or picture format, I don't use eagle.

SO i need those resistors. What values and what power rating? MAth says for 1.8 A i need 0.5/1.8=0.27 ohms...
the pdfs:
[sdrv.ms]
[sdrv.ms]

I found a usefull value table on this forum to choose the resistance and wattage you need :
Or you can do like me, by connecting 1ohms 1/2W resistors in parallel :

if the resistors are in paralel the power ratings add right?
Why do you use diodes? Extra protection right? Because form what i know the TB has some internal ones as well...

Yep, total power = one resistor power * number of resistors.
For the diodes, I used to see diodes on other driver boards, but the TB6560 actually don't need them.
For future boards, I will remove the diodes and use the free space to add a fuse, it will be much more usefull.

understood, thank you.
I will add the resistors on mondaay or tuesday, i am currently having the weekend off.

ok so i added the resistors.
using a 12V source, it only works when I pull up both TQ1 and TQ2 (weak excitation - 20% amps). If i go 50 % or full out, the motor just buzzez and humms and goes crazy. What is the problem now?
I am using 0.33 ohms resistors, for the 1.8 A motors

and the measured voltage with a regular ohmmeter on the motor is +- 0.6 V . But i guess that is because it does the pwm (that is how it does its chopping right)?

0,33Ohms gives you 1,5A
20% of 1.5A = 0,3A
Uphase = 1,7Ohms * 0,3A = 0,51V
so it's OK

When the motor runs at 20% does it have torque or it barely turns ?
If it does not have torque, check the voltage on each wire and see if there is an output stuck at 0V
On my fried drivers it's always an output that is stuck at low state.

it has small torque. The problem was simple, genius me forgot I used a 400 mA power source to test it, so it could not give more amps to the motors...
I haven't tested with a proper source yet but I will.
I want to redo my board. Does anyone know of a good schematic+board on the net that i can just use/ copy, so i am sure i am doing it right? (I just need minnimum configuration, but i can strip the excess...)

Thank you very much

Is it ok if I connect enable and reset pins directly to 5V and thus have no contorl on them?
What would be the implications?

From the datasheet :
After VDD and VMA/B completely stabilizes at the rated voltages, the RESET and ENABLE pins can be
set High. If this sequence is not properly followed, the IC may not operate correctly, or the IC and the
peripheral parts may be damaged.

What firmware are you using? These chips need a longer step pulse than some firmware provides. Teacup works. Marlin needs to be modified. Not sure about the others.

I use EMC2, a software cnc motion controller, it sends the step & dir signals by the parallel port, I set the pulse timing at 6000ns, 5000ns doesn't work

I use arduino

ok so here is my next problem:
i remade the board, the schematic I used is attached. I made it after the schematic here:
[www.cuteminds.com]

the only modofication: reset is connected directly to 5V, and there is no resistor and capacitor to it...
I made the board and tested it, and without the motor attached it did the steps ok (i measured the outputs with a multimeter). When I connected the motor it did that trembling and humming again. Only this time it wouldn't work at all, even at 20% torque. And now the source had enough current for it.
Then i think I blew my chip. I dont remember what i was doing precisely, but now one of the output pins is internally shorted to Vin, and when I give the chip Vin, GND, and 5V, I sucks tons of amperes. (so I'm pretty sure it's burned, just need confirmation)

The thing is I would buy another chip and try it again only if I was sure it is going to work. I've been struggling with this for so long now... and the holidays are here...

Any suggestions, ideas?
Maybe tell me what chip would be better / more stable to work with?
Thanks

I've put a 10kohms resistor and 100nf capacitor on the reset pin to reset the chip at startup. I also encourage you to put a fuse on the power.
As I said before, this chip doesn't forgive any error, so respect the power-on and power-off sequences, double-check connexions before switching the power on and do not connect or disconnect the motor when the chip is enabled.
Once you follow these recommendations it should be reliable. I made some mistakes when I tried and installed the drivers, but now it works fine and I didn't fried an other chip since.

I don't know other chips that can replace the TB6560.

Edited 2 time(s). Last edit at 12/20/2012 03:05PM by Petrus.

oh, I disconnected/ connected the motor with the chip running...
The fuse should be 3 amps right?

but why didn't it work before I burnt it? That is what scares me from trying again...

It was working with 20% current and the weak power supply, no ?
I also fried other chips by connecting or disconnecting a motor when the driver was enabled, it's not specific to the TB6560.
Yep, a 3A fast-blow fuse will be fine.

yes, the weak power suply at 20%. The big professional one... didn't work at all...

i want to try new drivers, ones that don't have these sort of issues, and are easier to work with. How about:
[uk.farnell.com]
or
[uk.farnell.com]
?

alkalin Wrote:
-------------------------------------------------------
> i want to try new drivers, ones that don't have
> these sort of issues, and are easier to work with.
> How about:
> [uk.farnell.com]
> /driver-motor-4-ch-ls-2a-16htssop/dp/2211332
> or
> [uk.farnell.com]
> ptr-t/motor-driver-microstepping-24tssop/dp/205741
> 5
> ?

The DRV8806 isn't a stepper driver, but there are several other suitable TI chips. eg. DRV8825 which Pololu have a breakout board for. I've also noted DRV8811, DRV8821 and DRV8824 as being of interest.

A4982 is used in a few Repraps board, I am using it on sango-bc. It's very similar to A4988 but with slightly different microstepping modes and in a TSOP package, a lot easier to solder by hand. The A4982 seems to have severe availability problems in the UK at least. Digikey seem to have plenty of them though. I made some "stepsticks" with the A4982 for my sango-bc board.

I also noticed Panasonic have a stepper driver, AN44067, which seems quite cheap. It is probably too big to fit in the "standard" Pololu form factor. Otherwise it look interesting, I will grab some on my next order.

I am also amazed by these chips : [uk.farnell.com] a dual FET with 3.7A continuous for £0.09 ! 4 of those plus a suitably programmed PIC and you could make your own stepper driver for $3-4. Writing code to do basic micro stepper sequencing is not too hard (Microchip have an app note on that), but adding all the extra features the Allegro drivers have would take a while. Perhaps if you blow the FETs occasionally you just toss it and use a new one.

Thank you this is very helpful.

how about [uk.farnell.com]?

what are the power ratings of the sense resistors for these allegro chips? Are they the same as the ones you gave already? (if so why dont they specify it in the datasheet?)

Edited 1 time(s). Last edit at 12/30/2012 02:57PM by alkalin.

Power dissipated in resistor (must be less than rating)=(I^2)*R, voltage across resistor (must be <0.5V for the A3982, see the "Internal PWM Current Control" datasheet section)=I*R.

alkalin Wrote:
-------------------------------------------------------
> Thank you this is very helpful.
>
> how about
> [uk.farnell.com]
> b-t/ic-driver-motor-dual-1-5a-soic24/dp/1521717?

You might have trouble with that, it only does half-stepping. I think that the A39xx chips are from an older family of design, pity they don't cut the price much

I would go with A4982 if you can get it (seems to be back in stock at RS here in the UK), or DRV8825 if you want higher current.

> what are the power ratings of the sense resistors
> for these allegro chips? Are they the same as the
> ones you gave already? (if so why dont they
> specify it in the datasheet?)

As rebecca says, the relation is P= I²R. I plugged the values into a simple spreadsheet, to find what values I could use in a 0805 package, since they are cheaper and smaller than 2512 or whatever. A typical 0805 is 0.125W, some are 0.250W.

So for a max current of 1.5A, a 0.05R resistor needs to be rated 1.5^2 * 0.05 = 0.113W which is within spec plus a small margin. You also need to stay within the 0.5V limit rebecca mentioned.

For higher current rating, you need to juggle the figures to find a suitable resistor package. Higher current probably means less options, but you can probably parallel smaller resistors to get a higher power dissipation which might be cheaper than a large package. Of course if you are not troubled by cost or space then you can get a 2512 package going up to 3W.

is the fact that the a3982 only does half stepping the only problem? Because I dont mind that for my application...

alkalin Wrote:
-------------------------------------------------------
> is the fact that the a3982 only does half stepping
> the only problem? Because I dont mind that for my
> application...

I didn't see any other issues. It looks like it should be compatible, except for the stepping mode.

ok i found one issue:
The datasheet says min voltage for motors is 8 V. I might want to run the motors at 5 or 6 V, since this is a robotic application, and I might use batteries for it. The tb6560 had a min of 4.5 V, so it was ok.
Any idea of such a driver?