I just finished the exercise of matching my output transistors for my Aleph-X (IRFP044) and found a method that helped me match pairs under operating conditions (2 amps each @ 15 volts) without having to worry about the exact temperature of the heatsink they were mounted to.
How you ask?
What I did was to choose 1 transistor as the reference. Then I compared the voltage drop across the source resistors between this reference and all the other transistors (in the circuit below - adjust the 10k pot to get correct current - 4 amps total in my case). That gives two voltage values for each test. Of course you need to make sure you use the same insulators that will be used in the end (including goop). The transistor should be mounted to the heatsink as close as possible to each other, but more importantly, use the exact same location for each test (center of the heatsink is preferable). The source resistors should be matched as close as possible for good results, but keep the same resistor on the reference transistor in case there is some small mis-match.
When you have finished, you will have values for the voltage drop across the reference transistor and each of the other transistors (taken at the same time). You can then match transistors based on the voltage difference between the two values (in the plus or minus direction also). To verify your matched pairs are truly a good match, you can compare the pairs against each other in the same circuit (closely matched source resistor required here)
When I did this, every one of the matched sets was successfully verified (within 0.5%). Using this reference transistor method greatly reduced the effect the heat sink temperature had while matching, however it would be a good idea to do the final verification matching with the heatsink close to the temperature that you are expecting in the finished circuit. You can even check to see that they are matched over a range of temperatures if you would like... mine happened to track evenly from room temperature all the way to 50 degrees C (my transistors were all from the same lot)
Please let me know if I have made any incorrect assumptions.
Thanks,
Jeff
How you ask?
What I did was to choose 1 transistor as the reference. Then I compared the voltage drop across the source resistors between this reference and all the other transistors (in the circuit below - adjust the 10k pot to get correct current - 4 amps total in my case). That gives two voltage values for each test. Of course you need to make sure you use the same insulators that will be used in the end (including goop). The transistor should be mounted to the heatsink as close as possible to each other, but more importantly, use the exact same location for each test (center of the heatsink is preferable). The source resistors should be matched as close as possible for good results, but keep the same resistor on the reference transistor in case there is some small mis-match.
When you have finished, you will have values for the voltage drop across the reference transistor and each of the other transistors (taken at the same time). You can then match transistors based on the voltage difference between the two values (in the plus or minus direction also). To verify your matched pairs are truly a good match, you can compare the pairs against each other in the same circuit (closely matched source resistor required here)
When I did this, every one of the matched sets was successfully verified (within 0.5%). Using this reference transistor method greatly reduced the effect the heat sink temperature had while matching, however it would be a good idea to do the final verification matching with the heatsink close to the temperature that you are expecting in the finished circuit. You can even check to see that they are matched over a range of temperatures if you would like... mine happened to track evenly from room temperature all the way to 50 degrees C (my transistors were all from the same lot)
Please let me know if I have made any incorrect assumptions.
Thanks,
Jeff
An externally hosted image should be here but it was not working when we last tested it.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.