Steve,
> Firstly, I should stress that you must not confuse me with someone
> who understands all about these MOSFETs :-)
No problem! We are among hams here.
> I have been reading from my old (1987) IR catalogue, application note
> 966. This is describing the HEXFET III family, the first where diode
> dV/dt is specified. I have not had time to read it in full detail,
> but it refers to the diode recovery dV/dt limit applying "..as the
> diode recovers from conduction..." and describes a mechanism that
> occurs during the period that recombination is happening in the
> diode. If the value is too high, the bjt will conduct.
Yes, that make sense.
> I could not find AN-966 on line to give you a link - I can scan the
> article (7 x A4 pages, small print) for you if you wish.
I found an application note that says basically the same. It's AN2626
from STMicroelectronics, available on st.com. So I think that I you
don't need to bother scanning that AN. Thanks anyway!
Ron,
> If the diode is for protection, it doesn't effect the operation until
> the diode/s starts to conduct. Therefor the device has speed
> limitations when it is over driven and different speed limitations
> when it is not over driven.
It's all getting clearer today. The whole low dv/dt limitation is valid
for applications where the diode is made to conduct, then the FET is
driven on for a relatively short time at low current, and then it's
switched off, and all that with an inductive load. This happens in
zero-voltage-switching circuits. In that case, the low dv/dt rating
applies.
In linear amplifier use, normally the drain should never get even close
to ground, but stay at least about 5 or 7V above the source, more likely
10V. So the body diode will never conduct, and the whole issue of diode
recovery dv/dt is a non-issue.
But one needs to be careful to avoid any situation in which a MOSFET's
drain could be pushed negative, for example by resonances of parasitic
circuit elements, the low pass filter, antenna reactance or whatever.
Such a situation could make the MOSFET smoke its way into Neverland.
Thanks to both of you, for helping clear up this matter!
Now the other mystery remains in my mind: How manufacturers can rate a
high current MOSFET to switch in about ten nanoseconds, when the
inductance of the source connection seems to make this impossible!
Manfred.
========================
Visit my hobby homepage!
http://ludens.cl
========================
_______________________________________________
Amps mailing list
Amps@contesting.com
http://lists.contesting.com/mailman/listinfo/amps
|