Is there a rule for the value for emitter resistor

[ostie01]

Hi, I want to buy some non inductive emitter resistor for a power amplifier project. I've seen value from .1 Ohm to 1 Ohm. Is there a rule to choose this value. .22 Ohm and .33 Ohm look like to be a very standard value. But why some use .1 Ohm and others use 1 Ohm. The value I'm looking are for an approx. 200w power amp.

Thanks.

[AndrewT]

Hi,
how many pairs in each channel?
To drive 4 ohm or 8ohm?

[ostie01]

Hi, thanks for the reply:

This amp will drive 8 Ohm load and 8 pairs per channel

[Leolabs]

Maximum output power???

[ostie01]

65 Vdc for approx. 250w RMS

[AndrewT]

Hi,
based on 8pairs into 8ohm then I think 0r39 would work.

A highish value to force some balancing action between the 8pairs.
When all eight are paralleled then the effective emitter reisistor is about 0r05 in each half.

I think you will find that a single pair of emitter resistors are in the range 0r0 to 0r33. When paralleling the outputs you find that the designer increases the values as the pairs multiply.

250W from +-65V supply rails might not happen. Even 200W from PSU rails at that voltage will require a very stiff supply.

[ostie01]

Hi, thank you very much. I can buy those on ebay

http://cgi.ebay.com/ws/eBayISAPI.dll...MEWA%3AIT&rd=1

but they are .47 Ohm, can't find any .39 Ohm

[john curl]

Smaller is better. About .22 would be best for you.

[AndrewT]

Hi,

Smaller is better

provided the output stage remains stable.

[sreten]

Hi,

A good read of Doug Selfs "Power Amplifier Design" is recommended.

Basically it depends on the accuracy and stability of the biasing,
but .22 is a good value for most designs. 0.1 is better but only
with a good sized heatsink and accurate biasing such that thermal
runaway is not an issue.

As AT says generally multiple output pairs the values are typically
increased to end up with the same effective emitter resistance,
though theorectically I can't see why this is done.
(Possibly there are stability issues, that is as you add pairs to
the output stage current gain is increased, which is pegged back
by increasing the emitter resistances, but then again 4 output
pairs are going to need an additional pre-driver stage, which
complicates stability issues no end, compared to no pre-driver.)

With 4 pairs and 0.47 you will end up with an effective resistance
of 0.12, so there is no reason at all why it won't work well, especially
if the pairs are there to dissapate power and swing decent current
into decent impedance loads. If the pairs are there in an attempt
to swing very high currents into low impedance loads then possibly
0.22 would work better.

/sreten.