Multiple chips, one feedback resistor

[xplod1236]

Why can't I use one feedback resistor for multiple chips running in parallel?

[Jocko Homo]

Resistors are really cheap.

Unless you have some Vishay S102s that you want to try.

Not knowing the application.......you may want to isolate the chips to prevent current hogging. Can't do that with one feedback resistor.


Jocko

[xplod1236]

This is for a BPA200 Bridged/Paralleled Circuit using the LM3886s.

My question was, WHY can't I do it with one resistor?

[xplod1236]

Originally posted by Jocko Homo
Resistors are really cheap.

Where can I get cheap .1% metal film resistors?

[jcx]

because of process variations no 2 chip amps will have the same open loop gain or offset voltage

with one feedback R you only have one input voltage to feed to the paralleled chips, I doubt even chips that were adjacent on the same die would match open loop gain to 10%, then add gain and offset tempcos with inevitable heatsinking variations and thermal runaway is highly likely

chip amps are designed to be used with negative feedback, the designers go to no effert to make open loop gain stable or to match between devices, high gain is the only goal

cheap digital multimeters are adequate to select 0.1% matches from a large number of 1% Rs - the measurement doesn't have to be accurate, only repeatable

[leadbelly]

Originally posted by jcx
cheap digital multimeters are adequate to select 0.1% matches from a large number of 1% Rs

How please?

[cowanrg]

Originally posted by leadbelly


How please?

i think he's trying to say that if a meter is accurate to 0.1% (which cheap ones easily will be), you test resistors to the exact value you need. for instance, a 100ohm resistor that tests with a meter will be 100ohms, +-0.1%. you can get cheap 1% or 5% resistors, buy more than needed (like 10 times or so), and sort them out. ive noticed that some dale or vishay resistors that are 1% test very close to 0.001%. i bought 5 times what was necessary and got more than enough of 0.0001% resistors (thats the limit of my meter).

[leadbelly]

Originally posted by cowanrg
ive noticed that some dale or vishay resistors that are 1% test very close to 0.001%. i bought 5 times what was necessary and got more than enough of 0.0001% resistors (thats the limit of my meter).

How are you getting numbers like this? Are you testing with a bridge?

[cowanrg]

Originally posted by leadbelly


How are you getting numbers like this? Are you testing with a bridge?

hum, i dont know what you are asking... its quite simple...

if your meter is within 0.001% accuracy, and it says that its exactly 100 ohms, then that resistor is 100 ohms, +-0.001% accuracy. simple as that. so buy a big batch, test them individually, and use the highest accuracy ones.

[till]

Where can I get cheap .1% metal film resistors?

at www.reichelt.e search for "MPR 100K"
or
http://www.schuro.de/preisl-cmf-55-1.htm
http://www.schuro.de/preisl-491-0_01-1.htm

what is cheap?