zapping gates

[John Larkin]

We just tested the gate blowout voltage of a couple of 500-volt,
300-watt power fets:

IXYS IXTH11P50 p-channel 80 volts
IR IRFPS37N50A n-channel 115 volts

John

 

[Richard Henry]

We just tested the gate blowout voltage of a couple of 500-volt,
300-watt power fets:

IXYS IXTH11P50 p-channel 80 volts
IR IRFPS37N50A n-channel 115 volts

Can you provide more information on test setup and proedure?

 

[John Larkin]

Can you provide more information on test setup and proedure?

Well, I connected a power supply between gate and source and turned it
up until the gate blew out... is there another way? Gate was + on the
n-channel fet, - on the p-gadget, which ensured that Vd-s was zero.

So, a failure mode that we thought was possible, isn't. Gotta look
somewhere else.

John

 

[Stanislaw Flatto]

Well, I connected a power supply between gate and source and turned it
up until the gate blew out...

****************************
is there another way?
****************************

YES. It is called "testing" and involves defining what are you testing
and how you control the test parameters so _some_ cosistent info can be
gathered.
Usually you define those things so another tester can repeat your findings.
As reported it is technical gossip;^)


Having fun

Stanislaw
Slack user from Ulladulla.

 

[John Larkin]

****************************
is there another way?
****************************

YES. It is called "testing" and involves defining what are you testing
and how you control the test parameters so _some_ cosistent info can be
gathered.

What else would you want to know? The colors of the test leads? What
we had for lunch?

Usually you define those things so another tester can repeat your findings.

I seriously doubt that the gate breakdown voltage of power fets is
terribly repeatable, and I'm not really interested in testing a couple
hundred $10 fets to destruction and crunching the resulting statistics
to be sure. So if you have any comparable data, I'd love to see it.

I thought it was interesting that both fets died at just about 4x
their rated abs-max gate voltage.

As reported it is technical gossip;^)

Excuse me, I'm just an engineer.

John

 

[Ian Stirling]

What else would you want to know? The colors of the test leads? What
we had for lunch?

their rated abs-max gate voltage.


Excuse me, I'm just an engineer.

Ramp rate, temperature...

Though I'm deeply ignorant of semiconductor physics, I would expect the
electric field across the junction, and the failure threshold to vary
significantly if it's 1v/s, or 1v/ns.

 

[Stanislaw Flatto]

John Larkin wrote:

Excuse me, I'm just an engineer.

John

You are excused.
Now excuse me, but I am just an (electronic) engineer (ret.) with over
25 years in aerospace industry most of them in testing department.
So I read your post and try to make sense of what is written.

Have fun

Stanislaw
Slack user from Ulladulla.

 

[Jim Thompson]

On Fri, 12 May 2006 21:24:54 -0700, John Larkin

[snip]

I thought it was interesting that both fets died at just about 4x
their rated abs-max gate voltage.

Why is that surprising? The gate oxide is just a
very-carefully-controlled thickness of silicon dioxide (aka glass).

Excuse me, I'm just an engineer.

John

You are excused ;-)

...Jim Thompson

 

[John Larkin]

On Fri, 12 May 2006 21:24:54 -0700, John Larkin

[snip]

I thought it was interesting that both fets died at just about 4x
their rated abs-max gate voltage.

Why is that surprising? The gate oxide is just a
very-carefully-controlled thickness of silicon dioxide (aka glass).

If it's very-carefully-controlled, why do they guardband it 4:1?

Maybe because it goes down at high drain voltages?

John

 

[John Larkin]

Ramp rate, temperature...

Rate, as already noted, was a guy turning a knob on a power supply.
And if you read my other posts, I've already suggested the typical
temperatures in San Francisco this time of year.

Though I'm deeply ignorant of semiconductor physics, I would expect the
electric field across the junction, and the failure threshold to vary
significantly if it's 1v/s, or 1v/ns.

1 v/ns is a little fast for my personal wrist. It might be hard to
read the meter at that rate, too.

John

 

[John Larkin]

John Larkin wrote:


You are excused.
Now excuse me, but I am just an (electronic) engineer (ret.) with over
25 years in aerospace industry most of them in testing department.
So I read your post and try to make sense of what is written.

Well, one of us seems to have some doubt about how to measure the gate
breakdown voltage of a mosfet. Interesting.

John

 

[Jim Thompson]

On Fri, 12 May 2006 21:24:54 -0700, John Larkin

[snip]

I thought it was interesting that both fets died at just about 4x
their rated abs-max gate voltage.

Why is that surprising? The gate oxide is just a
very-carefully-controlled thickness of silicon dioxide (aka glass).

If it's very-carefully-controlled, why do they guardband it 4:1?

Hot electrons ;-)

I just checked a process spec... TOX +3%/-5%

Maybe because it goes down at high drain voltages?

John

Plus there are some rate effects.

...Jim Thompson

 

[Richard Henry]

Rate, as already noted, was a guy turning a knob on a power supply.

What kind of power supply?

 

[Rene Tschaggelar]

What else would you want to know? The colors of the test leads? What
we had for lunch?

John,
since you ask : the histogram, do a
hundred and plot the histogram.



Naa. Thanks anyway, I take your measurements
as a hint that the gate voltage could perhaps
survive more than the 15V.

Rene

 

[John Larkin]

What kind of power supply?

I think it was grey.

John

 

[Jim Thompson]

I think it was grey.

John

With vents ?

...Jim Thompson

 

[Richard Henry]

I think it was grey.

How noisy was the knob? How steady was the guy turning the knob?

 

[Mac]

We just tested the gate blowout voltage of a couple of 500-volt,
300-watt power fets:

IXYS IXTH11P50 p-channel 80 volts
IR IRFPS37N50A n-channel 115 volts

John

I don't quite know how this thread went bad. To me, it was a very
interesting post.

Since I've never done anything like this, I was just curious about how one
detects blowout.

Does the current jump up dramatically? Does the part make an audible pop?
I'm sure you had the current limiting set up on the PS, so maybe the
Voltage just drops off suddenly as the current limit kicks in?

--Mac

 

[Ian Stirling]

I don't quite know how this thread went bad. To me, it was a very
interesting post.

Since I've never done anything like this, I was just curious about how one
detects blowout.

Does the current jump up dramatically? Does the part make an audible pop?
I'm sure you had the current limiting set up on the PS, so maybe the
Voltage just drops off suddenly as the current limit kicks in?

The gate goes from being an essentially perfect insulator (well,
gigohms) to a smallish value resistor, as the stored charge in the gate
region becomes enough to break down the gate dielectric.
The energies are small enough that you won't usually get any audible
sign.

 

[Stanislaw Flatto]

Well, one of us seems to have some doubt about how to measure the gate
breakdown voltage of a mosfet. Interesting.

John

OK.
a) Destructive or non-destructive test.
b) Proper control of dv/dt and definition what di/dt is defined as
breakdown.
c) Manufacturer definitions of error levels on number(s) given in specs
sheet.
And so on and on.

All this is done so there is "No Entry!" sign for Mr.Murphy and his helpers.

And it IS a branch in our industry, not cheap, but essential.

Have fun

Stanislaw
Slack user in Ulladulla.