Bob Cordell Interview: BJT vs. MOSFET

Jump to Latest
Nelson Pass

Nelson Pass

The one and only
Joined 2001
Paid Member
hitsware said:
>IF you are in class A mode exclusively, THEN there is no optimum >value of voltage drop across the emitter resistor.

Shouldn't it be the same drop as across the load?
Click to expand...

No, John is talking about complementary push-pull operation,
where relatively small Source/Emitter resistances are used to
stabilize the bias current. In this case, the voltage drop across
the resistors is small.

:cool:
 
anatech

anatech

Administrator
Joined 2004
Paid Member
Hi hitsware,
It doesn't work like that. You would increase the losses (heat) dramatically.

The goal of most amplifiers is to make the output impedance of the amplifier as low as reasonably possible with respect to the load. What you are thinking about is power transfer for AC generators and the like. RF transmitters also work on this principle.

Tube amplifiers use load matching for different reasons and it is still not equal to the load. The damping factor would be one, or unity, in that case. Not very useful.

-Chris
 
john curl

john curl

Member
Joined 2003
Fellow designers, you are confusing the issue. These are elementary questions.
The emitter resistor is added to increase thermal stablility. The larger the resistor, the more thermal stability. However, the open loop damping factor might be effected significantly, if the value is 1 ohm or so. My original amp with 1 ohm resistors had a damping factor of 30, BECAUSE it had about 40 dB of negative feedback. More than enough for my purposes at the time.
However, there IS an optimum value of emitter resistor for optimum transistion between class A and class B operation in a class A-B power push-pull bipolar transistor amp. Unfortunately, it is usually a VERY LOW value, such as .05-.22 ohms, and this MAY not be enough resistance to give adequate thermal stability without active thermal tracking.
 
hitsware

hitsware

Member
Joined 2002
miller-music.one/
anatech said:
Hi hitsware,

Nope. You defined it as .......

There was some stuff about emitter resistors in there too. That would create an undetermined source impedance. Your damping in that case would be less than unity.

-Chris
Click to expand...
What I meant was if you say had a 8 Ohm speaker that worked well with say a 8 Ohm series resistance. But I know that is not the convention so is inappropos for this discussion.
 
anatech

anatech

Administrator
Joined 2004
Paid Member
Hi John,
Thanks. I was going to get there once hitsware replied. The question of output impedance of an amplifier vs the emitter resistance completely changes once you wrap it up in a feedback loop.

The entire issue of thermal stability vs emitter resistance is also dependant on the thermal resistance of the heat sink (in case) as well. I like to see a slightly negative temperature co-efficient for bias vs temperature.

Hi hitsware,
What I meant was if you say had a 8 Ohm speaker that worked well with say a 8 Ohm series resistance.
Click to expand...
I would expect that speaker to have a very flat impedance curve for one. The damping factor is one, or unity in that case. You are correct, this is almost never seen.

-Chris
 
john curl

john curl

Member
Joined 2003
Anatech, IF you want a thermal tracking circuit with a slightly negative temperature coefficient, you might try my complementary transistor tracking diode method. I have been using it since 1970 or so, and it tends to track pretty well and go slightly negative when pushed. For some reason, it seems to be slightly better than a single transistor tracking circuit. Perhaps, it has just been my good luck, but it is worth a try.
 
jan.didden

jan.didden

AX tech editor
Joined 2002
linearaudio.net
Paid Member
john curl said:
Fellow designers, you are confusing the issue. These are elementary questions.
The emitter resistor is added to increase thermal stablility. The larger the resistor, the more thermal stability. However, the open loop damping factor might be effected significantly, if the value is 1 ohm or so. My original amp with 1 ohm resistors had a damping factor of 30, BECAUSE it had about 40 dB of negative feedback. More than enough for my purposes at the time.
However, there IS an optimum value of emitter resistor for optimum transistion between class A and class B operation in a class A-B power push-pull bipolar transistor amp. Unfortunately, it is usually a VERY LOW value, such as .05-.22 ohms, and this MAY not be enough resistance to give adequate thermal stability without active thermal tracking.
Click to expand...

John,

You said in two (OK, three) sentences what Self needs two chapters for. Thanks!

Jan Didden
 
bogdan_borko

bogdan_borko

Member
Joined 2005
john curl said:
Anatech, IF you want a thermal tracking circuit with a slightly negative temperature coefficient, you might try my complementary transistor tracking diode method. I have been using it since 1970 or so, and it tends to track pretty well and go slightly negative when pushed. For some reason, it seems to be slightly better than a single transistor tracking circuit. Perhaps, it has just been my good luck, but it is worth a try.
Click to expand...


Wher can we see the schematic?
 
C

Charles Hansen

R.I.P.
Joined 2003
john curl said:
Might be a good idea to try.
Click to expand...

Why John, I *am* surprised!

Error correction is just gain in the feedback loop.

So if you think feedback is good, I say use as much as possible. Use error correction, use complementary feedback pairs, use nested feedback loops, do it all! You could even do one of the tricks used by Jadis -- add *positive* feedback to increase the gain of the circuit so that you can get more *negative* feedback out of the circuit.

But if feedback is not good, then don't use any....
 
You must log in or register to reply here.
Top Bottom