On Sunday 03 October 2004 12:58 am, John Crighton did deign to grace us with
Have you tried it with the listed fixes fixed?
The fatal flaw is your bang-bang on/reverse cycle.
Since it's only software, do the softstart and deadtime
now, so you quit breaking stuff, and implement the other
suggestions before you fire it up some more.
I'm able to spot a few things I don't like in this one as well, at a
First I think the rectifier diodes across the mosfets are supposed to
be schottkys, of a few amps but that can block full rail to rail
Second is the method I see used to control dead time, slowing the turn
on rate of one side over the other by using different sized gate
resistors. I think you'd want to have one side to switch just as fast
as the other, only delayed in time. What we have here is switching
like: | on one side and / on the other, so one side will be alot less
efficient as well.
Perhaps this was done in a misunderstood attempt at speeding up turn
Those same resistors should be removed from the circuit during turn
off to allow fast discharge of the gate capacitances. Easiest done
with a back diode in this case, possibly better done by only using one
resistor per driver, placed at the emitter of the NPN turn on
transistor. That will also help with gate step induced spurrious turn
on by presenting a lower impedance to the reverse miller charge, it'll
have a straight path to ground through the PNP's emitter. Said PNP
could also be beefed up in size over the turn on NPN transistors. The
more current it can sink the better.
Last, for everyone who said something about "less transistors on the
bottom" etc, I would think you'd want matched drivers for all FETS in
order to preserve any delays induced?
I don't think there's much reason to double up the output mosfets,
just select the right ones for the job, and it will switch that much
faster for the given driver.
What else, would have been nice to look over the rest of it more but
that text made my eyes bleed.
<40Amp> whined and crossposted to 3 groups:>Where can I find 24v upto 40Amp H bridge DC motor driver schematics?
and if I ever catch you cross-posting again, you go in my
permanent killfile. Pick one, and post. If you get ignored,
then you picked the wrong group. Be polite, or you get
Polytechforum.com is a website by engineers for engineers. It is not affiliated with any of manufacturers or vendors discussed here.
All logos and trade names are the property of their respective owners.