On Wed, Jul 25, 2018 at 12:38 AM, satbeginner wrote:
the recommended IRF730 has a input capacity of 620pF
The ordered replacement IPA60R280E6 has a typical input capacity of 950pF.
For what it's worth: the broken FET I found in my scope when I got is was a
IRF710 (400V 200pF...)
I am no expert on driving power FETs, however I think the biggest concern is the gate charge rather than Gate/Source capacitance ?
There is a good deal of correlation between the two of course, but they aren't 1:1 related.
In datasheet, on the first page of any MOSFET, the key parameter they specify to brag about how good their FET is (running cool and going fast), is RDson, and gate charge. BTW I got carried away in a previous post... the RDson of the modern replacement I used is 280mohoms not 28mohoms, of course, so "only" an order of magnitude better than the old part, not 2 orders. Still, quite impressive....
Gate charge tells you exactly just how many electrons/energy you need to shove into the gate to turn the FET on. Up to your gate driver to supply this many electrons in as short a time as it possibly can, to make the transition as fast as possible, to waste as little energy as possible and make the FET run as cool as possible. It's more to do with the "rise time". I feel gate/source capacitance is more of a limiting factor when you are trying to drive the FET as fast as possible. Like in the hundreds of kHz. But in the 22XX Tek scopes, the regulator runs at only 40kHz or so.
So I think we are in the case of slow signal but with fast edges, aren't we.
I looked at the datasheet of all the FETs you mentioned, plus the one I bought to fix my 2215 the other day. One can see that there is no linear relationship between input capacitance, and gate charge.
Here goes :
IRF 730 700pF 38nC
IRF 710 170pF 17nC
IRF 820 400pF 19nC
IPA60R280E6 950pF 43nC
IPP60R280P7 760pF 18nC
So for example take the old IRF820 I pulled from 2215. My replacement IPP60R280P7, has twice the input capacitance (Rated 100V higher than the IRF820), BUT..... the gate charge is virtually the same, even a tad lower if anything, at 18nC versus 19nC.
And if you compare it to the IPA60R280E6, though a very similar FET made by the same manufacturer, in the same vein... has more than double the gate charge (but only a bit more input capacitance).
Waiting to the expert to chime in, but as for me, unless I am demonstrated otherwise, I will continue to worry more about (total) gate charge than input capacitance. At least for the old Tek scopes that are driven very slowly at only a few tens of kHz.
So IOW, Leo... the FET you bought, with more than twice the gate charge of the old FET well.... might "work", kinda, but as for efficiency/power dissipation and long term reliability, that would need more investigation... Depends how much headroom was designed in the driver circuitry. I am not competent enough to analyze/quantify the operation of this circuitry.
At least you in your scope, the FET is mounted a heat sink, with some air around it (might have a fan also ??), so you have some margin there... unlike my 2215 which has no heatsink and zero air around it... :-/
Anyway, keep us posted on you progress ! ^^
Vincent Trouilliez
Vince