Re: Lost steps

Jerry,

I have been watching your posts for some time. Sorry to see you
having such problems. You have been getting some good advice but I
think something has been overlooked. Mariss mentioned it in
passing,i.e. the phase outputs of the driver itself!

You are using the camtronics 5A drivers which use MOSFETS for output.
I hope the next part adds your knowledge, but if you already know
this, stay with me:

These transistors are VERY sensitive to static electricity. They are
constructed of many hundreds (or thousands) of little electronic
switches all wired in parallel (so the resistance drop is as little
as possible, this is what makes them good for our uses!). Each little
switch carries part of the workload (the current being switched is
shared among all the workers, as it were).

So heres the rub, I think you have some lazy workers in your output
transistors!
There are a couple of reasons these little guys get lazy. The first
has already been mentioned, static electricity. You may be surprised
to know that the "scoot across the floor and shock someone" trick we
have all played involves some hundreds of volts. Yet, these
transistors are rated only about 80V.

What happens is SOME of the switches(inside) get blown up; now there
are less workers available to do the work, so the others have to pick
up the slack. This means THEY are now working harder than they
should... So some of them eventually quit also. The process
continues, and at some point becomes noticeable to us Big guys, in
almost exactly the way you have been describing.

A few other causes with the same results are mentioned in the
camtronics assembly instructions. First, poorly flowed solder to the
legs of the MOSFETs. This increases the resistance on the path
through the transistor, with the result that those overworked,
underpaid electron carriers say ,ENOUGH already, I'm outta here! (as
above)

These guys also don't like too much heat. (they can be destroyed
during soldering into place) As I mentioned, they may not be hurt
enough to notice at first (during circuit checks) but they are still
injured.

I've saved the hardest to find(and deal with) for last: You CAN have
electrical isolation of the heatsink to the transistor body (when
measured with an ohmmeter) and STILL have all the problems above!
When you check the heatsink to MOSFET continuity,you are only using a
few volts at a tiny current, and this CAN fool you. BEEN THERE,DONE
THAT. There is a world of difference between this and 70V at 5amps!

I had some driver boards where the solder mask(the green film on
PCboards) was the only thing between a heatsink at GND, and 36V at
2A. Checked out ok with the meter, but under power, the electrons
were arcing through the solder mask to the heatsink with EXACTLY the
results you have been describing. (poor board design, but thats
another matter)

Theres some whys, now to the whats:

1.Check ALL of the transistor mount tabs continuity to ground.(the
heatsink) Be SURE they are OPEN circuits. And compare the readings
you get, you're looking for one(s) that are different.
2.Measure EACH leg of the MOSFET to its trace on the board. Again
looking for variations. You may need to re-flow the solder to each
lead (as mentioned in the instructions).
3.Be aware of the parallel construction described above and realize
that these parts do not always fail catastrophically.
4.Mariss is right, its ALMOST NEVER the motors.
5.Jan is right too, still could be a shielding problem.

But my money is on the output transistors being partially blown for
the reasons described in this (long) post.

Thanks for listening. Hope this helps


Ballendo