Transistor matching

System 55 • System 35 • Model 15
Post Reply
Rarecomponentadsr
Posts: 115
Joined: Mon Aug 24, 2015 5:08 pm
Location: Derbyshire England

Transistor matching

Post by Rarecomponentadsr » Mon Aug 16, 2021 1:11 pm

Can someone confirm that the method of matching transistors for most Moog modular systems is by Vbe measurement and not hfe measurement ?
The matching diagram in the back of the Norlin manuals seems to confirm it is by Vbe .

Is there a significant difference between the two measurements …if matching by hfe are the transistors not really ‘ matched ‘ and does it depend on how and what circuits the ‘matched ‘ trans are used in ?

Also I think I read that the Norlin Moogs matching circuits are not reliable ? Is there a tried and tested option for matching Vbe ?

User avatar
analogmonster
Posts: 266
Joined: Wed Nov 24, 2010 9:50 am
Location: Germany
Contact:

Re: Transistor matching

Post by analogmonster » Tue Aug 17, 2021 3:25 am

Yes, the diagram uses Vbe measurement. But I think that Vbe is dependent of hfe anyway, as hfe can be considered as current amplification of the transistor, and the rest of the matching circuit operates as current sink (current source for PNP). So the voltage drop at the emitter of the transistor under test is dependent of a constant (the current sink / source (741 and 2n3904 / 2n3906)) and the variable current output of the transistor under test, which is dependent of hfe and measured by vbe. So both aspects are considered here.

The Norlin circuit works. So it is reliable. I even integrated it in my 904B and 904A clone boards. But to answer your question, any matching circuit which provides a constant measuring part like current sinks or current sources works, as you just want to ensure relative equivalence between the transistors to be matched, no absolute values.

For me hfe is the core competence of a transistor, and if hfe equivalence can be ensured by vbe measurement, I am happy. On the other hand the transistors you buy nowadays have quite close parameters within group buys, that is what I noticed when I matched them. And a ladder filter I built with unmatched transistors did not work really different to that with matched transistors, so I consider this as a more academic question, at least for analogue synthesizers which lives from warm and unsharp behaviour. For other types of circuits this mght be a different story.

Rarecomponentadsr
Posts: 115
Joined: Mon Aug 24, 2015 5:08 pm
Location: Derbyshire England

Re: Transistor matching

Post by Rarecomponentadsr » Tue Aug 17, 2021 7:06 am

Many thanks for your quick and detailed reply.

I believe the question over the Norlin matching circuit relates to a missing 10k resistor on the pnp section ? I think that’s what it is .

I know most test meters have hfe measurements in them.

So….given what you have explained and let’s say in theory your building a 904a clone, which method would you employ to match the transistor pairs ? hfe or Vbe …I would just like to know your personal opinion.

User avatar
analogmonster
Posts: 266
Joined: Wed Nov 24, 2010 9:50 am
Location: Germany
Contact:

Re: Transistor matching

Post by analogmonster » Tue Aug 17, 2021 7:29 am

Yes, I was also wondering about the missing resistor, but the voltage measurements gave also for PNP the range which was notet in the schematics, so this seemed to work. And again, you only need the difference between the transistors under test, not the absolute values.

The question is how hfe is measured in test meters. AFAIK they measure currents also via voltage drop above reference resistors at a given voltage, so the internal test hardware could be similar, but I am not a test meter technician.

I used Vbe for my 904A clone, as described before. See my T904A website for further details.

Add: The Norlin service manual contains indeed several errors. The cv mixer section of the 904b for instance, and others. I recommend simulating a circuit with e.g. LTSpice to see whether the schematic works at all. I did this with all my Moog module clones.

Rarecomponentadsr
Posts: 115
Joined: Mon Aug 24, 2015 5:08 pm
Location: Derbyshire England

Re: Transistor matching

Post by Rarecomponentadsr » Tue Aug 17, 2021 10:00 am

Again, many thanks for detailed answer …that’s answered all my queries
and more besides …I never thought to refer to your informative web site over the query !

Post Reply