A possible solution for high current steppers - DRV8711 board?

I came across the specs for the DRV8711 recently and it seems TI sell a breakout
for this driver with external MOSFETs (rated to 4.5A). http://www.ti.com/tool/BOOST-DRV8711

This looks to be a possible solution to a fairly common issue - driving high current
(2A+) steppers from the Arduino.

I've not played with this device myself though - anyone got experience?

[ This is the sort of chip I was looking for a few years ago and couldn't find,
in an ideal world someone like Pololu or Adafruit would catch on to this and
produce a stackable shield for - I'll email them :slight_smile: ]

Very interesting.

…R

So I've built myself a little test board for this chip and some CSD88537 dual MOSFETs
and managed to drive the thing with a small NEMA23 stepper I have. Many of the
register fields were set rather arbitrarily and there's a fair bit of work still to do on making
a decent library for this, but it basically is a good to go and little if any heat, yet to test at
5A which is a tougher test but the MOSFETs are 12.5milliohm so I'm confident.

http://sphinx.mythic-beasts.com/~markt/drv8711-driver.html

Given the need for SPI pins plus step/direction/reset its too pin-greedy as it stands,
but a shield with one or two of these and a standalone 168 to interface to them (via
serial) for setup might be worth considering. They can be configured to drive 2 DC
motors each too.

That looks very neat.
But I am a little confused. I had the impression from the link in your first post that the whole board was available from TI ?

…R

Yes but it wasn't very breadboard friendly and had hard-wired current sense resistor values,
since I'm thinking about a possible shield design I decided to look at PCB layout and before
I knew it I had a breakout board designed :slight_smile:

MarkT:
before I knew it I had a breakout board designed :slight_smile:

Very impressive.
Are you thinking of going into production?

…R

I could release the design as open hardware. I'm also talking to Adafruit about possibilities.

Well I've got a basic library for driving this chip, not in final form by any means
(hard wired to use hardware SPI at the moment), but able to configure the thing.
It does not drive the step / direction pins, that's left to the user
of the library or external hardware even.

Hi markT,
This is a very good idea !
I was actually looking for DRV* component of TI with more microstepping than 1/32 (DRV8825 for instance) when I found the DRV8711.
But I just need 0.6A / phase in order to plug a NEMA8 ...
Do you think you can share your schematic or i need to create mine ?
Have a good day,
Leo

Well I'm on my 3rd iteration of PCB design, waiting for it to be made up - its
a standalone board with ATmega328 and opto-couplers running off the DRV8711's
5V output. I had to be really careful with total current consumption as its limited
to 10mA, initialially had poor choice of opto couplers. The schematic currently is as
shown:

Hello MarkT

I'm not sure to understand your design, don't know if the isolation is between the Mega8 and the DRV8711, or between the connector SV1 and the DRV8711.

Seems that the DRV8711 can be controlled by both the command coming through SV1 and from the Mega8. Certainly for test purpose only, is that correct?

I'm asking because I destroyed an Arduino board by making some tests with that device, and I would know if we can drive the DRV8711 directly by a processor, or is we need some isolation/protection.

Thanks for sharing.

The opto isolation is pretty standard precaution with stepper drivers, don't have to worry about line-borne
interference. The idea is to only use the direct interface to the Atmel chip for programming / testing.

Hi Mark,

Did You make some progress since May?
How much current can You push to the Coil with this mosfets without heat sink?

Regards,
Maverick

I bought an 8-wire NEMA23 that can be configured for 4A in parallel-bipolar.

The MOSFETs do get warm at 4A, but not crazily - and remember these are dual MOSFETs, if you
had 8 discrete ones it would spread the heat better.

and remember these are dual MOSFETs

Do You mean this one: CSD88537 ?

How far did You come with Your development?

Regards,
Mav

Well the second board works fine (10mA total for all the optoisolators was a bit tricky to sort out).
When I have time I'd like to see if a 4 channel Mega shield could be done - might not be enough
room though.

Are you interested in using one?

Obviuosly there is no informations (and interest) for such board on the internet.
I found only very few posts from You on various Forums.

I begin to develop one by myself.

You have seen the DRV8711 development board? http://www.ti.com/tool/boost-drv8711

Yes, I have seen that. I allmost go the same route (with dual MOSFETs).
I even tought about 8 MOSFETs in D²PAK Package.

Finaly I decided to make a Board with 8 MOSFETs (TO220) for better cooling performance.

Regards,
Mav

Aiming for 10A handling then? :slight_smile: