How to test mosfet?

[Nordic]

Hi, I am haveing some problems with a hybrid amp, using some IRF610 mosfets, I suspect one is blown... how can I test it...?

I found some method with diode test on multimeter online, wich you charge up a pin and then take measurement over two other pins and then short all the pins with your finger wich makes the meter reading go high or not... sadly it didn't say wheather if it goes high the fet is dead or not....

One of the two fets I have in does go high, the other not. I have one spare.

Also, is there an easy test for lm317 regulators?

.

I added 3k3 gridstoppers as the smaller version was full of crap with my walwart.

I constructed a battery one.... made one that worked before...
Getting crazy voltages...
getting full 24V after CCS,
20V to heater filaments

erm, there are 2 lm317s in gain configuration each supposed to deliver about 150mA,,,

[richie00boy]

Short gate to source with a paperclip or something, measure resistance across drain-source and source-drain. It should be open circuit both ways but one way will show some conductivity depending on how much voltage your meter applies to determine resistance as it may make the body diode conduct.

If it's less than 100 ohms both ways it's blown. If you have been fiddling with it in your hands without the gate being shorted to source then the gate insulation is probably smashed by static on your hands.

[Nordic]

lol i smashed it with a short from multimeter probe... the low voltage obviously makes for failures that down blow smoke and makes it harder to find the nipper.

[AndrewT]

Hi,
the absolute maximum gate to source voltage of most FETs is between 15V to 20V.
The gate capacitance is the only load (there is no resistance load in a FET gate) any finger or tool on the gate has.
The potential on your finger or tool could be kV. That equals a dead FET.

Look up PASS or ESP for simple FET testing procedure.
An indicated short from D to S or S to D = dead FET

Big Power FETs have big gate capacitance. That capacitance allows a large charge to be absorbed and yet the final voltage can stay below the 20Vgs limit. This allows careless handling and surviving power FETs, small FETS will not let you away with any careless handling.

[Nordic]

ok, made a jig with protoblock to short pin 1 and 3

new one from packet
Negative probe on S positive on D meter reading 1
Negative probe on D positive on S meter reading 695R

I assume this is ok

Will wait for feedback before removeing fets from pcb to test in jig.

[ilimzn]

To measure an Nch MOSFET:
1) Use DVM in diode test position
2) Holding red lead on S
3) Put put black lead on D. Reading should be between 500 and 800, you are measuring the forward voltage drop in the body diode.
4) Put black lead on G. Reading should be 1... (open circuit)
5) Hold black lead on S
6) Put red lead on D. You shuld read 1... (open circuit) as the gate should be retaining negative charge from step 4.
7) Put red lead on G, you should read 1... (open circuit)
8) Put red lead backon D, you should read a low value (close to zero initially, possibly rising very slowly). You are now measuring the D-S voltage drop with the gate positively charged in step 7.

In any case, source to gate and drain to gate should alsways read open circuit. If you get low readings in steps 4 or 7, it's dead for sure (in either case, not a MOSFET... any more ). If you get low readings in step 3 or 6, there is a good chance it's dead but you can test this later on, by doing it again with G and S shorted. If you get anything but the body diode reading in this case, the FET is dead. Also, you can try to measure the drop between shorted D+G and S, you will get the gate treshold voltage for very small Id provided by the DVM - however, most DVMs don't provide a high enough voltage to get over the gate treshold of VMOS, so you are likely to still measure 1... (open circuit), the exception being lateral MOS (will emasure low, about 1V, useful for testing for fakes), and PI-MOS such as the Toshiba Audio MOSFETs which will measure about 1-2V dependant on DVM.

BTW that schematic for a phones amp cannot work right as the required heater current for the 6922 is 300mA, and the current source connected LM317 only gives 150. It appears to me this was originally designed for a ECC81/82/83 12AU/AX/AT7 type tube with center tapped 12V filament, which is divisible into two 6.3V 150mA filaments by grounding the center tap - but then you have a problem with the too low anode voltage.

[Nordic]

Ili there are two of the 150ma sources tied to the heater filament, only one channel shown in schematic...

As i said i build a workig one before....

[Tube_Dude]

Hi Nordic

Before putting the Mos Fet in circuit , make sure that the connection node of the CCS , the anode of the valve and the 220 Ohms resistor , is at ~ half the rail supply voltage.

Only if that voltage is right , connect the Mos Fet in circuit.

[lineup]

How to test mosfet?
===============================================

I guess you missed some of my posts ..
I know you often read them. So must be a rare exception

I have my suggestions.
Circuits I use for testing JFET and MOSFET.
I want as good P R E C I S I O N as I can get .. I am a bit of freak, I guess.

But when it comes to fantastic precision instrument and circuits at our forum
... nobody can beat dear jackinnj.
He has made sure to get the best and build the rest!
See for example:
*******
Heat Sink Tester - jackinnj
DIY Test Equipment for Audio - jackinnj
I happen to know jackinnj use a lot of MOSFET.
He also does MOSFET Matching and offer them for DIYers to buy to a good price!!!!
*******
So, if anybody can tell us how to Test MOSFETs ... it is this man.

Jackiinj HOMEPAGE - http://tech-diy.com/
A visit here may give you some hints and I hope jackinnj read this topic
and contribute some of his knowledge.
************************************


Here I will contribute and post my way of doing FET testing / matching

I post my MOSFET tester circuit.

It is not very difficult to build.
A dual regulated supply with Heatsinks - ADJUSTABLE with potentiometer
... LM317 LM337, will give you possibility to to Tests up to 1.0 - 1.5 Ampere.
This would be enough to get good values.


Lineup Audio Lab
Regards
lineup

[richie00boy]

Originally posted by Nordic
ok, made a jig with protoblock to short pin 1 and 3

new one from packet
Negative probe on S positive on D meter reading 1
Negative probe on D positive on S meter reading 695R

I assume this is ok

Will wait for feedback before removeing fets from pcb to test in jig.

That sounds ok as long as the '1' is just the meter indication of over-range rather than 1 ohm...