Emitter resistor at output stage transistor.

smithomo

Member
2011-08-24 10:03 am
Hi there,
I'm would like to know what the advantage if I replace small resistance(like 0.22ohm) resistors at emitter pin of output stage power transistor.

Do I get amplifier with lower output impedance?
Is this resistor used for prevent power transistor overheat only?
 
Hi there,
I'm would like to know what the advantage if I replace small resistance(like 0.22ohm) resistors at emitter pin of output stage power transistor.

Do I get amplifier with lower output impedance?
Is this resistor used for prevent power transistor overheat only?

Hi Smithomo
The emitter resistors are usually placed to reduce Class a/b distortion. See article point 6
Distortion In Power Amplifiers :)

Cheers / Chris
 

smithomo

Member
2011-08-24 10:03 am
Hi Smithomo
The emitter resistors are usually placed to reduce Class a/b distortion. See article point 6
Distortion In Power Amplifiers :)

Cheers / Chris
Hi Chris Daly,
That a nice article I have to read right away.
So as I ever seen Krell FPB 300/600 class A schematic, it also have the emitter resistors, so can we state that it is mandatory for all transistor amplifier.?
 
Hi Smithomo
The use of a emitter resistor in a circuit has the tendency to stabilize operating point against changes in temperature and β-value. ,as 65 DegN has pointed out.

Its use generally though - which is your question with all amplifiers is not apparent.
However applies to class a/b amplifiers as a good technique to reduce distortion in output stages. Which is to say transistor bias has many forms.

A very early amplifier from 1967 the quad 303 found great merit in what was termed triples, and used emitter feedback from the use of small value resistors similarly placed, here Doug Self is discussing triples it makes great reading.:)
Audio Power Amplifier Design Handbook - Douglas Self - Google Books





Cheers / Chris
 
Last edited:
The emitter resistors reduce crossover distortion. To a first approximation, each resistor should be equal to the transconductance of one side of the output stage at quiescent current. Then the zero signal transconductance and full signal transconductance (when one side is cutoff) will be equal. Real life is not quite that simple, but that is roughly how it works.
 

smithomo

Member
2011-08-24 10:03 am
The emitter resistors reduce crossover distortion. To a first approximation, each resistor should be equal to the transconductance of one side of the output stage at quiescent current. Then the zero signal transconductance and full signal transconductance (when one side is cutoff) will be equal. Real life is not quite that simple, but that is roughly how it works.

Thank you so much :)
 
I have read more than a few times that the distortion of the output stage reduces as the emitter resistor values reduce.

Taken to the limit (zero resistance for the external emitter resistor) this would indicate that a single pair output stage with just the inherent internal emitter resistance will give less distortion than a stage with added external emitter resistor.

For the minimum crossover distortion the output stage should be set to the optimum ClassAB output bias voltage. This is what Mooly is describing.
 
AndrewT said:
Taken to the limit (zero resistance for the external emitter resistor) this would indicate that a single pair output stage with just the inherent internal emitter resistance will give less distortion than a stage with added external emitter resistor.
A BJT, unless it has internal ballast resistors, will have an effective emitter resistance which varies with current. There may be some limit caused by device bulk resistance but this is likely to vary between samples and it will be small. To a first approximation you have to provide the same resistance at quiescent current. The lower the emitter resistor, the higher the quiescent current. You could end up in Class A, where an external resistor is really there for overload protection only.
 

djk

R.I.P
2001-02-04 4:23 am
USA
Many NAD amplifier lacked emitter resistors, they did, however, have problems with the drivers running away and blowing up.

If you choose not to use emitter resistors on the outputs, I suggest using base resistors on the outputs to avoid the drivers blowing up.

Some Carver amplifiers used emitter resistors as low as 0R05 (50 mill-ohms) !!!
 
Back in the 1970's I modified a Dynaco ST120 by supplying more available current from the power supply(s), doubling the number of output transistors, using a wakefield water cooled heatsink, removed the 5.1 V zener, and removed the emmitter resisters as I speculated that these would have a compression effect on signal.
The amp put out 120 watts per channel @ 4 ohms (vs ~40 watts stock) @ < 0.1% THD 20 ~ 20 Khz.
I don't recall how this affected crossover distortion.
 

smithomo

Member
2011-08-24 10:03 am
My 12 year old designed amp was also no ballast resistor, it was still stable working fine nowadays.
I was currently designing/testing my new amplifier, below is my prototype one which also no ballast resistors.
[IMGDEAD]http://faphost.in/o2m54g[/IMGDEAD]

Dual darington power transistors are used for compensation circuit.
The THD measured by PC oscilloscope is 0.05% @17Vrms 1khz no load.
Circuit topology is about current feedback(feedback only from VAS) then no feedback in current buffer and output stage.

[IMGDEAD]http://faphost.in/jwgtbh[/IMGDEAD]

Unfortunately I have no more accurate analyzer to compare the result between have or no of this resistor.

I still believe that no resister should better in the real world, but I still seen the most hiend amplifier still have this resistors.

Need more opinion from you.. :)
 
Last edited:
A BJT, unless it has internal ballast resistors, will have an effective emitter resistance which varies with current. There may be some limit caused by device bulk resistance but this is likely to vary between samples and it will be small. To a first approximation you have to provide the same resistance at quiescent current. The lower the emitter resistor, the higher the quiescent current. You could end up in Class A, where an external resistor is really there for overload protection only.

I agree. One way to look at it is the emitter resistor reduces the % difference of the variance in Gm as the output stage conducts large and small currents. On the flip side, it also reduces current gain by providing some negative feedback in the form of emitter degeneration. Trading current gain magnitude for more linear gain, reducing the % difference of variance in Gm....they are related.;) You trade low output Z for more linear transconductance. Now if you wanted to lower output Z there are other ways such as HEC but that topic has been thoroughly discussed.:lickface: As the signal passes through crossover, the slope of the distortion components created by the varying output Z in this region is very steep indicating high order frequencies and may be out of range for the global fb loop to compensate for. Besides, who wants to pass all that HF distortion crap through every stage of their amp?:whazzat: Best to comprimise. DF96 posted earlier a way to figure the best comprimise emitter resistor values. Or you can go crazy with local loops and such, complicating things.:D
 
Last edited:
Hi smithomo
Thats nice work and if it works it works, but Bob Pease (dec) I think has more wires, which just indicates such connection method to trial a circuit is quite valid.:)

Cheers / chris
 

Attachments

  • bob.jpg
    bob.jpg
    14.8 KB · Views: 729

smithomo

Member
2011-08-24 10:03 am
Hi smithomo
Thats nice work and if it works it works, but Bob Pease (dec) I think has more wires, which just indicates such connection method to trial a circuit is quite valid.:)

Cheers / chris

Hi Chris Daly,
It is for trial as you said, I just to know how it work in practical. Currently it is not a fully load testing, my test power supply is not enough powerful.

What I may do next is to design the pcb with have blank space for Re then let see how it sounding to my ear.

However some more thing have to do as trial as well such regulation power supply for VAS. :sax: