MOSFET mystery

[JimBanville]

Re: MOSFET mystery

It's not failing, it's just not high enough voltage to turn it on, and you're just seeing the body diode.

I didn't mean failing as in broken/defective. I just meant failing to do the job in this case. But my question is, why am I only getting a reading from the diode with the probes in opposite orientation to "correct" on the source and drain? And why can't I short the gate and drain to see OL between source and drain, unless the fact that because I'm having to use the probes in an opposite orientation, I keep charging it? Sort of a catch 22 situation.

[eccerr0r]

Re: MOSFET mystery

It's not failing, it's just not high enough voltage to turn it on, and you're just seeing the body diode.

[JimBanville]

Re: MOSFET mystery

If my fluke 115 is failing, is that causing me to look at this backwards?
If you discharge the mosfet and test it "opposite" of correct (probes in opposite orientation), would you get the readings I'm getting?

[eccerr0r]

Re: MOSFET mystery

well that's exactly what it should show so far, just need to charge the gate properly I think.

[JimBanville]

Re: MOSFET mystery

hmm... could you use the transistortester and short G to S and verify that it detects as a diode?

Also the Vt is fairly high here, perhaps your DMM doesn't have enough "oomph" to charge the gate.

I don't know how helpful this will be, but...
if I place the mk168 parts tester leads just on the left and center pins (gate and drain), I get a capacitor with either 2652pF or 1947pF depending on orientation of the leads.
If I place the leads on the left and right outer pins (gate and source) I get a capacitor with either 3274pF or 3190pF depending on orientation of the leads.
If I place the leads on the center and right pins (drain and source) I get a diode with a diode with a Uf of 561mV and a C of 3.76n. I get the same results when flipping the leads, but the "direction" of the diode changes.

[eccerr0r]

Re: MOSFET mystery

use another dmm to measure the voltage in diode check mode.

If you have a 6 volt battery you can use it to charge the gate externally too.

and yes short G and S, then connect to tester to D and S and verify it shows up as a diode.

[JimBanville]

Re: MOSFET mystery

hmm... could you use the transistortester and short G to S and verify that it detects as a diode?

Also the Vt is fairly high here, perhaps your DMM doesn't have enough "oomph" to charge the gate.

My dmm is a fluke 115.

So you're saying connect a wire from g and s, and then connect the device to my tester just as I have it pictured in my previous post and press "test" button?

[eccerr0r]

Re: MOSFET mystery

hmm... could you use the transistortester and short G to S and verify that it detects as a diode?

Also the Vt is fairly high here, perhaps your DMM doesn't have enough "oomph" to charge the gate.

[JimBanville]

Re: MOSFET mystery

i have seen mosfets that meter strange
usually in high voltage invertors.
the internal diode fails open circuit - or sometimes changes it's voltdrop a lot.
i think it cant handle the reverse current and slowly breaks down.
once it has completly failed the mosfet itself doesnt last long.

if i was designing compact invertors for driving discharge tubes i would include larger external diodes!

I don't know why I didn't think of using my parts tester instead of my DMM (duh!), but here are pics. The pinout is as same in the data sheet for the the other brand of 26N40. Tester says it an N enhance mosfet.
But why, when I use my DMM, do I get no reading with negative dmm probe on source and positive probe on drain even after I "charge" the gate with the positive probe? I only get the .5v reading with positive probe on source and negative on drain. And no matter what I do, I can't seem to discharge the fet?

[stj]

Re: MOSFET mystery

i have seen mosfets that meter strange
usually in high voltage invertors.
the internal diode fails open circuit - or sometimes changes it's voltdrop a lot.
i think it cant handle the reverse current and slowly breaks down.
once it has completly failed the mosfet itself doesnt last long.

if i was designing compact invertors for driving discharge tubes i would include larger external diodes!

[JimBanville]

Re: MOSFET mystery

Maybe the mosfets are bad, as in open

Yes. Could be. It's just odd that they didn't short, which is the common fault, and they're measuring like P channel mosfets. Opposite polarity. And I can't discharge it.

[R_J]

Re: MOSFET mystery

My question is why isn't it measuring like a N channel mosfet?

Maybe the mosfets are bad, as in open

[JimBanville]

Re: MOSFET mystery

You are likely measuring the internal diode junction across source and drain

My question is why isn't it measuring like a N channel mosfet?

[R_J]

Re: MOSFET mystery

You are likely measuring the internal diode junction across source and drain

[JimBanville]

Re: MOSFET mystery

Pics?!! A video!

Seriously, I have no way to hold my phone and use both hands to test the mosfets.
Watch any video on "how to test N channel mosfets". That's exactly what I did. Only when I switched polarity of the probes did I get a reading. But what deepens the mystery is that mosfets usually short at the gate and drain pins. I show OPEN between those pins.

[volinakis]

Re: MOSFET mystery

N channel mosfets testing as P type?

Working on a smps.
Using Fluke 115. Red probe in red socket. Black probe in black/common socket. Meter set to diode mode.
Removed a pair of mosfets marked 1L02AB FDP
26N40 from the circuit.
The pair of Mosfets are still screwed to the single heatsink. Mosfet printing facing me.
I did the test where I short the left and center pins. I put black probe on right pin and red probe on center pin. I get OL. While keeping black probe on right pin, I put red probe on left pin and get OL. After a few seconds I move red probe from left pin back to center pin and still get OL.
Now, even after I short the left and center pins, if I switch the probes put the red probe on the right pin and black probe on center pin, I get .49v. The center pin never measures OL even after shorting the left and center pins when measured like this.
This happens with both mostets. But there is no short from the left pin to the center pin (gate to drain).
What's the deal?

[JimBanville]

N channel mosfets testing as P type?

Working on a smps.
Using Fluke 115. Red probe in red socket. Black probe in black/common socket. Meter set to diode mode.
Removed a pair of mosfets marked 1L02AB FDP
26N40 from the circuit.
The pair of Mosfets are still screwed to the single heatsink. Mosfet printing facing me.
I did the test where I short the left and center pins. I put black probe on right pin and red probe on center pin. I get OL. While keeping black probe on right pin, I put red probe on left pin and get OL. After a few seconds I move red probe from left pin back to center pin and still get OL.
Now, even after I short the left and center pins, if I switch the probes put the red probe on the right pin and black probe on center pin, I get .49v. The center pin never measures OL even after shorting the left and center pins when measured like this.
This happens with both mostets. But there is no short from the left pin to the center pin (gate to drain).
What's the deal?