YetAnotherStepperDriver (a DRV8825 breakout module)

[squalyl]

easier here.

As said in the comments (openmakersdaily) this is a nice board, I could be buying one or two! If you plan a stelpad booster pack, just announce it on the stellarisiti forum, they have a section for projects.

[EBo]

If you could sketch out a Bill Of Materials, I'll poke some of the CNC folks about building out some of these for testing. Maybe a small group of us could pitch in for a 100 or so order to get the price down and test.

As a note, Allegro’s A4988 has a reported design flaw where the motor drive current is asementrical between the rise and the falling pulse. This causes some weird harmonics and may also limit how well you can control the motor near stall speeds. I would have to check with one of the engineers to get the skinny on those details, and see if they have any info on TI's products.

Also note, that TI has the DRV8829 which is a 5A (3.5A RMS) version of the DRV8825, and maybe we can find appropriate heatsinks. The Enzotech MOS-C1 C1100 Forged Copper Heatsinks have been reported to work well for the A4988's, but the DRV8825's will probably take something a bit larger Enzotech MOS-C10 C1100 Forged Copper Heatsinks -- full disclosure I have ordered but not evaluated Enzotech's products, and have never worked with or for them.

EBo --

[katjap]

hi eBO.

I have a new version that i think will work out a lot better. I'll post it up today (eagle files and all.) As far as the 8829 i believe all these DRV88xx's are pin compatable, so you should be able to solder one in place of the other. The new board should be capable of handling more power and has heatsink mounting holes spaced at 20 mm.

~~Katja

[katjap]

Ok, versions 0.3 of YASD is (i think) ready. Although on the same PCB (dp6037) it was altered even more than the switch from 0.1 to 0.2. Ground plane spacing is now 24 mil. The motors are feed by 54 mil traces nearly all the way to the chip. The motor connector had to be switched to a molex KK-156 instead of the screw terminals because the screw terminals were too wide/big to allow for the extra heatsink drill holes that were added. The heatsink holes are spaced 20mm apart and several mm of clearence is allowed between them and the nearest through hole parts. There are now connections for 2 limit switches (although they both feed to the same pin on the CTRL connector) There is no RC filter on them, it is expected that debouncing would be handled in software (but you could also figure that out on whatever you mount the switches to.) A 3v3 LDO was added to provide power for the opto-switches. (change that to whatever you want) and 410 resistors were put to limit current to the emitters (VCC pin on the switch connectors) at 3V3 that would ammount to 8ma, enough to turn a H21A1 on. Pin 1 on the CTRL connector (LSWITCH) is pulled down with a 3k3 resistor. Because of the LDO the board is limited to 30V. If that doesn't work for you simply omit the LDO and solder some other connector to the VOUT and GND pins or edit the PCB ;) The original version of 0.3 had the driver chip much closer to the motor connector resulting in shorter traces (probably better) but then i couldn't fit the holes for the heatsink in. The heatsink was definately the most difficult thing of this whole process, and i can see why many companies would employ people to do nothing but thermal management. I tried to get rid of as many 10/12 mil traces as possible and replace them with 16 mil where possible.



All of the eagle files are available on github at:
https://github.com/katjap/YASD

[katjap]

I saw the decay silkscreen markings were backwards, so that has been fixed in the git.

[katjap]

DRV8825 Samples have arrived from TI.
I changed R3 and R5 to 390 ohm since 410s are rare.
I'll upload a BOM to github later tonight.

[naren]

I am not so sure the DRV8829 is a 'drop-in' replacement for the DRV8825.
The 8825 is a dual bridge setup while the 8829 is a single bridge setup. One would need 2 8829s to drive a bipolar stepper motor.