Motor Driver Suggestions

I'm working on a project where I have a couple motors that need 24v and draw a bit less than 2 amps (they've got limit switches and I'm monitoring feedback pulses with Arduino interrupts to detect stall). I've run the motors on a shield driver kit only rated for 1 amp at 12v without frying the board, but I've been using a dual h-bridge I got from Robokits in India for only $10, and it seemed to work fine for a good while. It uses the TIP 122 and 127 darlington transistors with flyback diodes, rated 5amp, 65v, and they seem like a good choice. I noticed the PC board was really sensitive: holding it in the palm of the hand would trigger the motors. I decided to put some liquid tape on the back side to protect things. Hmm. When I hooked things back up and turned on the power the TIP transistors blew up like firecrackers. I peeled the liquid tape off, took out the blown apart transistors, replaced them, and tried again. Bang, and this is a very impressive bang I might add. At 24v the case is exploding into fragments within fractions of a second - much faster than my 5amp magnetic breaker.
I decided it was maybe better to try a diy board since I had all the components for a discreet h-bridge. I tried this one I found on the internet:

I noticed what seemed like some discrepancies with the traces on the related site, so I laid the schematic out in Eagle to create a board design. After a day of hard work I finished it and carefully tested it with a meter and once again it sounded like the 4'th of July, but I'm getting tired of the fireworks going off in my hands. I cheated and used some readily available 6.5v as my logic rather than 5v, could that have blown it up? Apparently what has been happening is that the Darlingtons have been somehow driving both high and low transistors and self-destructing the bridge, and I'm not smart enough apparently to figure out why that's happening (I know the wiring is correct).
Shipping from India takes too long. Alas if if I had more money I would buy a driver board, but they all seem so pricey, and it doesn't look like it should be as hard as it's been.
I'm now considering a design based on a couple 740x logic chips using the same Darlington TIP 122 and 127 transistors that I've been blowing up (I still have enough but I'm running out!). Maybe the logic gates would be better at avoiding self destruction than my hand wired resistors and transistors.
Now that I've amused you with my saga, can anybody help:

  • suggest a known working design suitable for 24v 5a max
  • give pointers on how to avoid blowing up things and debugging an h-bridge

Perhaps have a look at this thread for an H-bridge driver idea. 5A using BJT-type transistors is not a good idea.

Rugged,

My motors draw 1.6 amp to be a little more precise, and will run with pretty short duty cycles normally so I still 5 amp capacity is plenty. Lately I've been blowing stuff up with just a volt meter hooked up though!

As far as I can tell, the TC4428 in that design you referred me to won't take 24v, and I don't think the MOSFETs will either...

I know wheels come in lots of different sizes, but I'm really surprised I have to invent my own wheel. It would also be nice in my book if I didn't have to supply 12v in addition to the 24 and 5v that I've already got, but that's just because I'm lazy and don't want another power drain since I'm running on batteries.

I guess I could recycle my motor shield (Freeduino motor shield kit from NKC) and use it to drive some MOSFETs. Would the TIP 12X darlingtons work nearly as well in that setup?

Yup, the TC4428 won't handle 24V, but it's not the only game in town. There are lots of higher-voltage MOSFET drivers. The IR2101 comes to mind.

The MOSFETs in the thread I referred you to will easily handle 24V. The FQP47P06 is rated for 60V and the NDP6060L is too.

Can't comment on your motor shield idea...not too familiar with it.

You don't have to reinvent the wheel if you're willing to spend some $$$$. Get a Geckodrive for ~$130.