Forums » Motor Control » jd version of the single sided OSMC....?


Hey JD!

I ran across your stuff away-back-when...before the lawnbot, before the make article ~ some time after you'd posted the triple-8 and then your version of the single sided OSMC board.  When I saw the lawnbot in Make I have to say I thought it was a just reward for all your hard work.  (and I was thrilled that I'd "seen it before it went prime"...though I had nothing to do with it. :)   )

I'm finally getting around to building a high amp motor controller for several scooter/electric buggy projects for my kids.  I found the 2002 dated OSMC single sided controller, but etching efforts have been mixed.  (mostly as it has a number of smaller traces)  I remembered your single sided version and HUZZAH, it's got nice, thick traces with minimal fiddly little stuff.  That said, while I can find pix of it I can't locate the osmc-jd.brd file to save my life.   (the link I found here: http://rediculous.prototyperobotics.com/site/osmc_jd.shtml  to it is quite dead)

Does it still exist?  Can you point me to a copy?

OR...more to the point...if I'm making a 36-48v motor controller capable of handling 30-50 amps, what would *you* recommend..?  The scooters don't need a full h-bridge, I don't believe.  (they rarely go in reverse.  ;-D ) but I've got an offroad go-kart idea (inspired by this guy: http://home.comcast.net/~briano911/gokart.html ) that would need some reversin' ability. Would you still recommend the single sided OSMC?  Your 2x4..?  The venerable triple-8?

Side note: you likely already know this, but mosfets seem to be the biggest eater-of-budget for these things.  After much "surplus electronics" searching and general scavaging I found tayeda electronics has the *best* prices I've seen.  No affiliation, but at $0.88 per irf3205 or $0.44 for an irfz44 I've been a very happy camper!  Oh, and scavaging old UPS systems yields a good supply of high amp n channel mosfets too.




Hi Aaron,

sorry for the late reply..  thanks for keeping up with my site, always good to meet another maker.

As for motor controllers, I do have some designs for higher amperage motors - the 2x4 seems to be the most solid and cheapest to build. I had mixed success with my single-sided OSMC boards. While they seemed to work great most of the time, there were a few instances of me trying to push the Lawnbot too hard and blowing one side of the board... which sucked to de-solder and re-solder parts onto, so I went with other designs.

The 2x4 is my board of choice for home-made, and is a pretty simple design, but also has a place to install some current sensors so you can monitor the amperage going through the board if needed. I found monitoring it extremely helpful to keep from overloading it and blowing up mosfets.

As for scooter motor controllers, I will mention that it is much easier to build a singe direction motor controller as you only need 1 set of switches. In this case, you can basically get a mosfet driver and as many N-channel mosfets as you need in parallel and have a super beefy controller.  If you need help designing a single direction driver, let me know and I can probably whip up a schematic and pcb file.

Once you add the reversing feature, you have to have a full H-bridge with 4 switches and that gets more expensive and complicated - as you have to either use P-channel mosfets which are less efficient or an all N-channel design that must have a voltage doubler to drive the high-side fets above the main battery voltage. The 2x4 uses P-channel mosfets for the high-side, but is intended to operate them either On or Off, while the low side N-channel mosfets are driven through a low-side bridge driver IC with the PWM signal for speed control.

How much amperage are you expecting to push through your controller?

And thanks for the heads up on tayeda... looks like a good supplier, I will have to check them out.


Thank you!

If you've got a handy, simple uni-directional motor controller schematic hanging around, I'd gladly use it.  Seems every time I think I've figured out all the bits that are needed I wind up banging my head against something else.  (oh, diodes for motor voltage spikes are good...whoops, need to take into account gate capacitance on the mosfets, ha...who knew mosfets could actually _spurt flame_ ..?!)   I've got an initial idea to use an attiny85 (a digispark, actually) to provide the pwm signal (taking input from a potentiometer), but controlling the fets has always been a bit of a bugger given the fact that the gate has to be 10v above the drain. 

I've got the p-channel mosfets on order for a 2x4 implimentation - it certianly looks like an easier/cleaner design!  Question though...since all of the mosfets are rated @ 80 amps continuous and you're using 2 per leg...wouldn't that allow for up to ~160 amps continuous on the controller? (this is - of course - assuming that the fets are heatsinked/cooled somehow)  Or am I missing something obvious? 

thanks again!