Amps
[Top] [All Lists]

Re: [Amps] MOSFET ratings

To: amps@contesting.com
Subject: Re: [Amps] MOSFET ratings
From: Manfred Mornhinweg <manfred@ludens.cl>
Date: Mon, 18 Jul 2011 16:58:51 +0000
List-post: <amps@contesting.com">mailto:amps@contesting.com>
Steve,

> One cross reference for the IXYS transistor you mention is 
> STP4NK80Z. If you look at the ST datasheet, it shows the lower 
> dV/dt as applying to the MOSFET diode

Yes, that's right. But a MOSFET's internal diode is impossible to 
separate from the MOSFET. So if the diode is limited to 5 V/ns slew rate 
or so, then I can't see how the entire MOSFET could have a higher slew rate!

The data sheet of the STP4NK80Z claims 32ns for switching 400V with a 
resistive load. That's 12.5 V/ns. If they mean to switch from 10% to 
90%, it would still be precisely 10 V/ns, but the data sheet gives a 
maximum dv/dt of only 4.5 V/ns.

For the inductive load case, the claimed rise time is even faster, but 
that is probably because of the phase difference between voltage and 
current. With the resistive circuit, both should be in phase, so that we 
don't have this source of doubt. In any case, as I understand it, they 
claim the transistor siwtches at least twice as fast as the maximum rate 
they give!

Do they perhaps mean that the transistor CAN slew at up to 10 V/ns, but 
shouldn't be allowed by the user to slew that fast? As a way of safety 
measure, to keep the internal parasitic elements from becoming part of 
the active circuit?

  and two different circuits
> for test/measurement of the main switching dV/dt and the diode 
> recovery.

Yes, one of those circuits has resistive load, and the other has 
inductive load.

> Does that help?

Not really. I still don't understand how a MOSFET can slew faster than 
its internal diode, or how to interpret the data to make sense of it.

 > I guess you would need to find some more
> general data about the FETs to get detail about what the diode 
> slew rate limit means in practice.

Yes, I tried that a few days ago, and the most interesting paper I found 
so far in this regard is "Power MOSFET Basics", by Vrej Barkhordarian, 
of International Rectifier. That paper explains the existence of not 
only a parasitic diode, but also a parasitic JFET and a parasitic 
bipolar transistor within the structure of a MOSFET, and it explains the 
dv/dt limitation of a MOSFET as coming from two different places: First, 
the basic integrator formed by the total gate resistance (internal plus 
external) with the Miller capacitance, and second, the internal 
parasitic bipolar transistor, which appears as having emitter to source, 
collector to drain, but the base connected to emitter through an 
internal resistance, and to the collector through an internal 
capacitance, and not connected anywhere else.

In this way, the first effect limiting slew rate can be influenced 
simply by harder driving, but the second limiting effect, that of the 
bipolar transistor with its internal RC network, cannot be externally 
influenced, and sets an absolute limit to a MOSFET's dv/dt.

Since the body diode and the other internal parasitic elements all share 
the same structures, I _assume_ that when a manufacturer talks about the 
dv/dt limit of the body diode, he refers to the limitation set by this 
parasitic bipolar transistor and its RC network. But this doesn't 
explain how a MOSFET could slew faster than this basic dv/dt limitation 
dictates.

So, I'm just as lost in space as I was before!

I'm resorting to actually testing several MOSFETs in a practical 
amplifier, and trying to understand from the waveforms and eductaed 
guesses what's happening inside them. But I can't possibly order a 
hundred different MOSFETs and try them all! So I'm working with just a 
few different types, and finding little correlation between the data 
sheets and the actual performance regarding dynamic response.

Any further help is highly welcome!

Manfred.


========================
Visit my hobby homepage!
http://ludens.cl
========================
_______________________________________________
Amps mailing list
Amps@contesting.com
http://lists.contesting.com/mailman/listinfo/amps

<Prev in Thread] Current Thread [Next in Thread>