[jerseyguy1996]

I was checking out Texas Instruments offerings in terms of stepper driver IC's and I noticed that they have a booster pack using their DRV8711. It attaches to one of their LaunchPad prototyping platforms to create an 8-52V 4.5Amp max stepper driver. Just wondering if anyone has tried to make their own stepper driver using this. Here is a link to it:

Stepper Motor BoosterPack featuring DRV8711 and CSD88537ND - BOOST-DRV8711 - TI Tool Folder

[jalessi]

Jerseyguy1996,

Interesting find. It would be nice if a four or five device motherboard was available.

Jeff...

[eddyde]

Cool, it looks like the Launch Pad is the TI version of the Arduino, price is cheaper too.

[jerseyguy1996]

Jerseyguy1996,

Interesting find. It would be nice if a four or five device motherboard was available.

Jeff...

The Launchpads are only $10 and the booster packs are $25 so for $35 you could have a stepper driver....in theory. I'm just not sure if it is really that easy.

[jerseyguy1996]

Cool, it looks like the Launch Pad is the TI version of the Arduino, price is cheaper too.

That's correct. I think ST has their own version of the Arduino as well. TI definitely seems to put the most work into developing hardware ideas that are almost ready to use but that can be adapted to whatever your specific use is.

[eddyde]

The Launchpads are only $10 and the booster packs are $25 so for $35 you could have a stepper driver....in theory. I'm just not sure if it is really that easy.

It proabibly would't be that hard, a few hours learning the programing and you'd be good to go. The only thing is these boards are designed for prototyping, not for permenent installation so the question is, would it last as long as an off the shelf drive...

[eddyde]

That's correct. I think ST has their own version of the Arduino as well. TI definitely seems to put the most work into developing hardware ideas that are almost ready to use but that can be adapted to whatever your specific use is.

I bought an Arduino starter kit and found it very easy to learn by the second evening of mucking around with it, I was writing my own code and even got it to vary a tricolor led, according to the weather. I would imagine the TI version would be pretty capable as well.

[jerseyguy1996]

I'm fairly proficient with Arduino. I know that the headers on the Arduino are fairly capable of being connected to one of their shields in a permanent installation. I just don't know enough of the subtleties of motor control to know what I may be missing with these. For instance I seem to remember reading something about having to tune motor controllers to avoid resonance in the motor. I'm not sure what is involved in that or if these would allow for it.

[aarggh]

I'd be very interested in this as well, but have to admit the Arduino crowd have really got things down pat with the breadth of support and projects out there. Looking forward to see how this one pans out.

cheers, Ian

[eddyde]

I'm fairly proficient with Arduino. I know that the headers on the Arduino are fairly capable of being connected to one of their shields in a permanent installation. I just don't know enough of the subtleties of motor control to know what I may be missing with these. For instance I seem to remember reading something about having to tune motor controllers to avoid resonance in the motor. I'm not sure what is involved in that or if these would allow for it.

Ther are several video tutorials on tuning and other motion control subjects, on the Galil website Online Videos | Galil I think you may have to sign up but it's free.

[kreutz]

Interesting, but when I went looking to buy one set, for testing, found that there is only one left at Digikey (US) and 12 left at Mouser (US).
I also saw the thermal image of the DRV8711 board on the TI site and, for practical CNC use, it will need a heat sinking solution for more than 2.0- 2.5 Amps per phase. The two central dual mosfet chips go all the way to more than 100 degrees (C) during their tests.
Even if we were able to make it work, easily and reliably, as intended, there is still the problem of availability of those boards at $25.00 + shipping.

kreutz

[jerseyguy1996]

Interesting, but when I went looking to buy one set, for testing, found that there is only one left at Digikey (US) and 12 left at Mouser (US).
I also saw the thermal image of the DRV8711 board on the TI site and, for practical CNC use, it will need a heat sinking solution for more than 2.0- 2.5 Amps per phase. The two central dual mosfet chips go all the way to more than 100 degrees (C) during their tests.
Even if we were able to make it work, easily and reliably, as intended, there is still the problem of availability of those boards at $25.00 + shipping.

kreutz

Probably wouldn't be too hard to put a heat sink across the back of those fets. As far as availability, they include the gerbers and BOM's so if you are proficient at soldering you may be able to assemble your own for fairly cheap. Plenty of low cost prototyping services out there to get the boards made. I use Oshpark myself for getting boards made.

[kreutz]

Actually, I was looking at it as an opportunity, for the rest of the members, to get a cheap, good middle range (2 to 5 Amp/phase) bipolar drive, since the majority of them don't have the means to successfully assemble and test/troubleshoot an SMT board.

If after testing TI's evaluation board performance I find that this chip is a good candidate for production, then I will think about doing something about it. I still have a full equiped prototype and low volume production electronic's R&D and manufacturing shop at home which includes a professional SMT pick and place unit, with four heads, capable of 13,500 parts/hr. It might even become a "Kickstarter project" .

kreutz

[jerseyguy1996]

Actually, I was looking at it as an opportunity, for the rest of the members, to get a cheap, good middle range (2 to 5 Amp/phase) bipolar drive, since the majority of them don't have the means to successfully assemble and test/troubleshoot an SMT board.

If after testing TI's evaluation board performance I find that this chip is a good candidate for production, then I will think about doing something about it. I still have a full equiped prototype and low volume production electronic's R&D and manufacturing shop at home which includes a professional SMT pick and place unit, with four heads, capable of 13,500 parts/hr. It might even become a "Kickstarter project" .

kreutz

For people that ARE comfortable soldering SMT components, perhaps you could sell just the bare board (hint hint) :-)

[kreutz]

I sold a bunch of K-4 interface boards already populated and gave away, for free, a few of the k-4 bare PCBs (First photo below) for some people to assemble. Not even one of the latter was successful. Even when all of them claimed they had the electronics knowledge and SMT assembly experience. On the other hand, the Mardus-Kreutz unipolar micro-stepper drives (80 Volt- 8 Amps) PCBs were sold bare and were successfully assembled and used by a lot of people around the world. Both designs were posted as open source in this forum a few years ago. That is why I am talking from my own experience about designing for SMT assembly at home. When it is a simple design and a small board (like the M-K board, on the second photo, with only a couple of SMD ICs), the probability of success is higher, a few boards with hundreds of components in total, no that easy... any assembly error, that might not be evident by looking at the pcb, will produce problems not easy to identify without the right instrumentation or special gigs for testing.
kreutz

[kreutz]

I went ahead and ordered the launch pad and booster pack in order to make some tests. I will post the results and probably a video as soon as possible. The booster packs need only to be initialized via SPI before start working (on power up), it would've been a lot easier (and cheaper) if they stored the new configuration inside the chip.

kreutz

[jerseyguy1996]

Awesome! Looking forward to seeing the results! Surprising that the booster pack doesn't come with a boot loader. Regarding the heat dissipation, those dual channel fets have an RdsON of 12.5mOhms. An interesting hack may be to replace them with single channel power fets with a lower RdsON value. Plenty of options out there.

Edit: just noticed that you wrote that it is the booster pack and not the launch pad that needs to be initialized.

[jerseyguy1996]

Just running some quick calculations (and I may be doing this wrong) with 12.5mOhms Rds and 4.5 amps, that gives us a voltage drop across the fet of .05625v. That should give us .253125 watts that need to be dissipated. Those dual channel fets are good up to 2.1 watts provided they have good size copper pad underneath them. Am I working that out correctly? I don't have any formal electronics training.

[kreutz]

The problem is mainly the thermal resistance of the Mosfet encapsulation, it has about 60 - 75 °C/W (1 sq. inches copper pad, depending on the FR4 board thickness). Typical RθJA = 60°C/W on a 1-inch2, 2-oz. Cu pad on a 0.06-inch thick FR4 PCB. The package (SO8) is not a good candidate for attaching a heat-sink. The greatest thermal problem will manifest at low RPMs with high current motors. I am also not confortable with the low voltage margin (they specify 52 Vdc Motor voltage and the Maximum voltage spec for the MosFets is only 60Vdc, at 50Vdc a motor, with a typical leakage inductance, will produce transients over 100Vdc during the recovery time of the antiparallel diodes)

kreutz

[Laubi82]

Hi guys,
Just came across this thread. I am actually working on a design that uses the DRV8711. It also has an Arduino on board that could be used for all kind of cool things.
If there are enough folks interested, it might be a good candidate to get a batch of boards made.

Some details:
- target voltage is 48V (limited by the DRV8711)
- design uses (individual) FETs for better thermals (tested with 5A, no problem)
- possibly getting current up to 8-10A (using 2oz copper PCB)
- Arduino compatible controller to configure and run the DRV8711
- basic STEP/DIR/EN interface
- control over USB possible (UART to Arduino to DRV8711)
- size is approx. 2.6x2.6in
- no additional heat sinks