Quote from: TinselKoala on May 14, 2012, 11:25:31 PM
I want to stress that the bias current I measured STARTED at 90 mA, quite consistently, but always quickly rose to 130-150 mA. I thought this was weird so I tried this with all the IRFPG50s that I have (except ironically the known bad one, I now realize), even two unused ones, and they all showed about the same behaviour, some faster some slower to rise up to the higher current.
So I'm wondering if this is just a characteristic of the particular mosfet (GL has the IRFPF50, I think) or what. How long did GL keep the measurment going? For me, in 5 or ten seconds the current started to rise from 90 heading up.

TK,

I'm using a fairly large heat sink so I did not see that the bias current went so high. The main current did go
up a little as the MOSFET did heat up, but that is normal. What I'm addressing is that there is NOT an
relationship between bias current and main current where you simply ADD the bias current to the main
current as .99 did in his drawing. With the bias current running the main current did drop. But the sum
of the bias current and main current was not a simple addition. I think there is more to it than me have
found out yet.

GL.

Quote from: MileHigh on May 14, 2012, 11:39:48 PM
TK:

It's possible that what you are observing is the heating up of the quasi-junction inside the MOSFET.  Higher temp equals higher current transfer through a MOSFET, correct?

Assume that when the current is bleeding through the zener diode, that a fairly small volume inside the MOSFET represents the actual junction.  So it becomes a kind of pin-point "hot reactor core" and it takes a while for the thermal equilibrium to take place.

MileHigh

Yes, you are right, certainly for the D-S pathway. But in the applications I am most familiar with, the mosfets are either simply not stressed to the point where the circuit causes the body Zener to come into play, or they are stressed so very severely that I use external, replaceable zeners or high-speed diodes for protection... and even that doesn't help sometimes. So in my prior experience the internal body zener either never comes into play or is bypassed by the external ones. Still, I manage to blow mosfets occasionally.... and I don't even consider it an "event" unless I have to pick the pieces out of my teeth or the ceiling or something.

So I'd like to know if GL finds the same current rise over 10-20 seconds using the PF50s that he has.

ETA: I see that GL has answered already. That could be it. I noticed that for the one "bare" mosfet (no HS) I tested briefly, the bias current rose very fast as the die heated.

Quote from: Groundloop on May 14, 2012, 11:47:22 PM
TK,

I'm using a fairly large heat sink so I did not see that the bias current went so high. The main current did go
up a little as the MOSFET did heat up, but that is normal. What I'm addressing is that there is NOT an
relationship between bias current and main current where you simply ADD the bias current to the main
current as .99 did in his drawing. With the bias current running the main current did drop. But the sum
of the bias current and main current was not a simple addition. I think there is more to it than me have
found out yet.

GL.

Yes, I wonder if the currents add like parallel resistances like this: 1/Itotal = 1/Imain + 1/Ibias, or something like that. Or (1/I main - 1/Ibias). I'm too tired to run the numbers right now.

Quote from: TinselKoala on May 14, 2012, 11:48:19 PM
Yes, you are right, certainly for the D-S pathway. But in the applications I am most familiar with, the mosfets are either simply not stressed to the point where the circuit causes the body Zener to come into play, or they are stressed so very severely that I use external, replaceable zeners or high-speed diodes for protection... and even that doesn't help sometimes. So in my prior experience the internal body zener either never comes into play or is bypassed by the external ones. Still, I manage to blow mosfets occasionally.... and I don't even consider it an "event" unless I have to pick the pieces out of my teeth or the ceiling or something.

So I'd like to know if GL finds the same current rise over 10-20 seconds using the PF50s that he has.

ETA: I see that GL has answered already. That could be it. I noticed that for the one "bare" mosfet (no HS) I tested briefly, the bias current rose very fast as the die heated.

TK,

I think the reason that you did see the bias current increasing is that the RdsOn actually went UP in value
as the transistor heats up. That will fit the theoretical analyze. I can set up my test circuit and try it again, but
this time allow the circuit to run for a much longer time and see what I get for the bias current.

The RdsOn going UP in value is often the reason those critters blow. You generate a lot of heat in the transistor
and as it heats up the RdsOn climbs and you are generating more internal heat, etc. So at one point you get
too much heat and you heat sink can no longer cope with it and the transistor blows.

Back in 1/2 an hour or so.................................

GL.

Quote from: Rosemary Ainslie on May 14, 2012, 11:07:35 PM
NO
(snip)

Q. E. D.