> As I read up more on h-bridges, I understand better why you say this
> (about switching time and linear region).  I chose 50khz just thinking
> it's well outside the upper audible range) but that's not a critical
> parameter.  AIUI, if I drop to 20Khz, the drive requirements get much
> simpler.

I've not experienced it myself but my belief is that noise from switching
in the audible range comes from the load, eg motor or coil windings
vibrating, and some loads are better made than others. Occassionally
I've heard the semi ringing if it's not bolted down hard enough to the
heatsink, or like a loose transformer that hums until you give it a clout

Given a fixed transition time, as the number of transitions / sec decreases
then the time / sec spent in the linear region decreases. For example if th=
e
transition time is 1us then at 1MHz ......

> However, I just discovered this seemingly neat device -- VNH5019.
> At US$8 with minimal other support components required, is there
> any reason I should really be trying to engineer my own?  From what
> I see of discrete-component versions, I'm very much in the same price
> range already.

If the cost is about the same I'd go with an IC, someone's already done
the design work

Last year I re-jigged some golf carts. The overseas manufacturer had
put a very poor discrete driver in and the FETs were cooking and failing
unpredictably (although it was often after user abuse - overloading etc).
The PCB was designed for 3 FETs but another 3 (!!!) were added at
the factory to try and spread the stress. In fact the FETs chosen were
so big that only one was almost enough if it was driven properly. An
inexpensive change to the driver got the FET count back to the original
3 and they now run barely above ambient

Joe



-----
No virus found in this message.
Checked by AVG - www.avg.com
Version: 2015.0.5577 / Virus Database: 4257/8853 - Release Date: 01/02/15

--=20
http://www.piclist.com/techref/piclist PIC/SX FAQ & list archive
View/change your membership options at
http://mailman.mit.edu/mailman/listinfo/piclist
.