I need to match some parallel output BJTs. From what I gather, the main concern here would be Vbe. How exactly is this measured? I was going to try to use the circuit here: Matching Power and Driver Transistors
Don't you need to keep at least one of the parameters constant to measure the others? The article says never to adjust the pot after the initial setting. Wouldn't this mean that every transistor would have different readings Vbe, Ib, and Ic?
I was thinking the you needed a fixed Ib between transistors to measure the Vbe. Is this right?
Any help would be greatly appreciated.
Thanks
Don't you need to keep at least one of the parameters constant to measure the others? The article says never to adjust the pot after the initial setting. Wouldn't this mean that every transistor would have different readings Vbe, Ib, and Ic?
I was thinking the you needed a fixed Ib between transistors to measure the Vbe. Is this right?
Any help would be greatly appreciated.
Thanks
Hi,
output BJTs are usually not matched as VBE is very consistent, just use sufficient emitter degeneration for current sharing (0.2-0.6 Ohms will do usually). Matching is only practically necessary for mosfets, where VGS_on may vary from 3 to 4 Volts.
Your right, one parameter has to be kept constant and that is usually the bias (collector) current Ic intended for use in the amp.
Have fun, Hannes
output BJTs are usually not matched as VBE is very consistent, just use sufficient emitter degeneration for current sharing (0.2-0.6 Ohms will do usually). Matching is only practically necessary for mosfets, where VGS_on may vary from 3 to 4 Volts.
Your right, one parameter has to be kept constant and that is usually the bias (collector) current Ic intended for use in the amp.
Have fun, Hannes
Hi
This circuit should do well.
It idles the transistor at about 20V, 50mA and you can just hook up a meter to measure the voltage between base and emitter.
Mind out for temperature drift, though. As the transistor warms up, Vbe will decrease. Maybe for each transistor to be tested, use a procedure like:
a) Connect it to test circuit.
b) Apply power.
c) Check Vbe after 5 seconds.
(i.e. same warm-up time for each device)
Cheers - Godfrey
This circuit should do well.
It idles the transistor at about 20V, 50mA and you can just hook up a meter to measure the voltage between base and emitter.
Mind out for temperature drift, though. As the transistor warms up, Vbe will decrease. Maybe for each transistor to be tested, use a procedure like:
a) Connect it to test circuit.
b) Apply power.
c) Check Vbe after 5 seconds.
(i.e. same warm-up time for each device)
Cheers - Godfrey
Attachments
The manufacturer would normally measure Vbe (and Vce(sat)) with constant Ic and Ib, normally at a 10:1 ratio to ensure saturation.
This would be done under pulse conditions to ensure no heating and for test speed.
As stated, Vbe spread is normally very low.
More imprtant is Hfe spread which will be reduced by using degeneration, ie Emitter resistors.
This would be done under pulse conditions to ensure no heating and for test speed.
As stated, Vbe spread is normally very low.
More imprtant is Hfe spread which will be reduced by using degeneration, ie Emitter resistors.
I am replacing outputs in a Crown Power-Tech 2 amp. I understand that the original parts are just Motorola parts matched in some way. I was thinking of following what I read in a different post (http://www.diyaudio.com/forums/solid-state/40138-help-fixing-crown-power-amplifier.html), and adding the extra outputs used in the Power-Tech 3 and replacing the emitter resistors with .27 ohms. So would I be ok just replacing the transistors with On Semi replacements, increasing the emitter resistors, and not worrying about matching them?
If they are from the same batch, don't bother "matching" them because they're already "matched". The Vbe's will be very tight within a lot, but not necessarily between them. Use the diode test function on a DMM (which measures Vbe at some arbitrary current) and see for yourself.
Crown would have parts spanning multiple lots on hand, and would need them pre-graded so techs and assembly operators can get the right ones the first time, without having to think (or stop and measure). They may also have an acceptance spec on Vbe/Hfe, which would put the bias adjust closer to the right spot the first time it's turned on. Most of use a light bulb that will come on if the bias is way high - they can't get away with that at the factory.
Crown would have parts spanning multiple lots on hand, and would need them pre-graded so techs and assembly operators can get the right ones the first time, without having to think (or stop and measure). They may also have an acceptance spec on Vbe/Hfe, which would put the bias adjust closer to the right spot the first time it's turned on. Most of use a light bulb that will come on if the bias is way high - they can't get away with that at the factory.
Hi
You mention that you will be replacing the emitter resistors. That sounds a good idea and I would probably replace all the emitter resistors. Maybe you were planning to anyway but I dont know so I thought I would mention that. You need all the emitter resistors to be same to ensure the same current in each output transistor. 0.27 ohm sounds ok but even higher at 0.5 ohm is also ok. You may need 3 watt resistors but you feel how hot they get in use. - you need to be able to touch them ( carefully ) when the amplifier is at working temperature.
Don
You mention that you will be replacing the emitter resistors. That sounds a good idea and I would probably replace all the emitter resistors. Maybe you were planning to anyway but I dont know so I thought I would mention that. You need all the emitter resistors to be same to ensure the same current in each output transistor. 0.27 ohm sounds ok but even higher at 0.5 ohm is also ok. You may need 3 watt resistors but you feel how hot they get in use. - you need to be able to touch them ( carefully ) when the amplifier is at working temperature.
Don
Thank you, everyone, for the help so far!
I have another question. Would it be necessary to match the drivers?
I have another question. Would it be necessary to match the drivers?
That's where all my oddball (no matching date codes) power transistors end up - drivers. Except in very rare cases where the designer didn't know any better, they never put those in parallel.
- Status
- Not open for further replies.