Matching Mosfets

Hello everyone, sorry to bother you with this question the more I read about this more confused I got. I want to macht IRFP244 and 9610 for my aleph2 that I'm going to start building, but I don't know how to get the value of my resistor, I'm using 2.2k resistor with regulated power supply for 5 sec. for both transistors.Is this setup ok? please help me with this. thanks Rommel
 
2.2K is about right for the 9610 @ 15V, but I would use something around 20 ohms for the 244's so you'll be matching at around .5A where they'll be used. You'll need to time the measure closely (5 sec within +- .5 sec) because 1.5W will heat them quickly and skew your measures). And don't try to read it again for at least 20 min so it can return to ambient temp.

I set up everything (supply, resistor and DVM). Clip an aligator clip to the drain and gate, then the other to the source, count to 3(one one thousand two one thousand three one thousand) then read the measure.
 
If you use a 20 ohm resistor for the 244's, make sure that its at least 10 watts.... the resistor will heat up. I match for 30 seconds per mosfet....the meter isn't moving as much after 30 seconds...i don't know how easy it would be to get mosfets accurate at only 5 seconds. On the other hand.... they don't have to be super accurate matched..... close is just fine.

Make sure the mosfets are on a flat surface... i use a glass table....so that they all start at a similar temp, and just like Brian said, if you mess up a reading...wait 20-30 min before you rematch.

Lots of different opinions on matching.... who knows who's right. If Nelson says something....stick with that :) good luck man!

-Matthew K. Olson
 
You need to be measureing the voltage accross drain-gate and source and it should be 3.5 to 4 volts.


If you match the 244 after 30 sec under juice, you've heated them up a bit, and you're assuming that they reached the same final temp, but if the vgs is different, then they're under different amounts of heat, and not always the same temp.

But on the otherhand, they are closer to operating temp. One idea I had for matching was to keep them in boiling water untill testing them so they would be closer to the temp where they'll run, but testing burning hot fets would be a PITA.
 

steenoe

Member
2004-03-23 10:43 am
2,2k is way too much resistance! You need like 10r for the irfp240 and something like 560r for the irf610 (or9610)!! I remember NP said that they just measured long enough to get a reading! I normally "Juice" them for 30 sec's. The resistor needs to be at least 10 watts. And a steady supply at that!

Steen:cool:

BTW 15v for the irfp240 and 10v for the irf9610!!!
Take a closer look at the link DW gave!!
 
steenoe said:
2,2k is way too much resistance! You need like 10r for the irfp240 and something like 560r for the irf610 (or9610)!! I remember NP said that they just measured long enough to get a reading! I normally "Juice" them for 30 sec's. The resistor needs to be at least 10 watts. And a steady supply at that!

Steen:cool:

BTW 15v for the irfp240 and 10v for the irf9610!!!
Take a closer look at the link DW gave!!
2.2k is not good for 9610. that was what Nelson said.
 
Not to say that my way is best, but I just finished matching some IRF9610's for my Aleph-X. I have a detailed web page on my progress so far that you can find here

Scroll down to the section on Matching Components.

Since I'm still a relative beginner, I'm trying to document my project and methods as thoroughly as possible so that others can easily follow it.

Eric
 
other matching idea

Hello everybody,

Today I started matching mosfets for my own little aleph-x, but I wanted to be able to match the components at a higher precission than 10 mV. Unfortunatly I have no acces to highend equipment so my standard multimeter will have to suffice.

The idea is as follows:

Instead of measuring the Vgs of 1 mosfet, I use one fixed mosfet and measure the difference in Vgs between this mosfet and the test mosfet (running parallel).

The result is that you can measure much more precisely using the standard voltmeter (because you can set the precission of the meter to millivolts instead of volts). I'll use a heatsink on the "fixed" mosfet to minimize the diference in temperature. But now my real question: are there any disadvantages of using this method?

See attachment for schematic.

Regards,
Jacco
 

Attachments

  • matching.gif
    matching.gif
    2.6 KB · Views: 1,505
Not a bad idea. Just make sure to let your "control" fet run for a few hours before your test so it can reach equilibrium tem, and make sure there is no AC/Heater vent blowing or anything to cause a draft. Any variation in temp causes a big change in readings.

Also make sure the test fets are all at room equilibrium temp and you keep them in the test the same amout of time because the testing will heat them and change their Vgs.

You're setup should also help make PSU instabilities less of an issue.