PA100 (x2 LM3886) 0.1% Resistors

[Al1B]

i use this schematic ....




as you can see, there are some 0.1% & 1% resistors used ...
can any1 tell me, what's the difference between using low-toleranced resistors or using normal 5% resistors?

thanx already

[nigelwright7557]

In this application the 0.1% resistors are setting the gain through the two amplifiers.
If they were mismatched by much then the two amps would be outputting different voltages and clashing.

[!]

Using less than 0.1% tolerance resistors is less than ideal, but they are used even on the audio input and certainly many people don't use 0.1% there. I usually split the difference, when I buy parts at places like digikey I notice either volume savings or a minimum # purchase requirement to buy 10 or more of things like resistors so I'll buy 10, or 20, etc, and where I want better than 2% (+-1%) tolerance between two resistors I'll just match them myself. Often the "match" won't be within 0.2% but it'll be enough to satisfy.

[Speedskater]

It's not that the exact value is so important.
What is important is that all the resistors have the same value.

[Al1B]

so is there any chance i can use multi-turn variable resistors instead the 0.1% resistors (except the two 3W resistors in output) ?

[abraxalito]

Yes, you can use a multi-turn pot to set the gain accurately. But don't use the pot to give the whole resistance value or stability will be an issue. Ideally you'd have an 18k 1% and a 5k pot in series for setting the gain. You only need one pot - its the relative gains not the absolute which is important.

This schematic has a few flaws - the main one being the matching at LF depends on the tolerances of the 100uF capacitors - better to take their tolerance out of the equation by increasing their value to 1000uF. They don't need to be 50V, 16V is fine. The input resistors R2 & R6 don't affect the gain so have no need to be 0.1% - save money on these and also on the output resistors (R10,R11) which can be 5% types. Better to be on the safe side by increasing the output resistors to 0.22R - it makes the matching of gains less critical.

[panson_hk]

Increasing the e-cap to 1000 uF also avoids its non-linearity (distortion) occurring in the audio band.

[Al1B]

abraxalito thanx for ur complete reply

[Minion]

Just use your digital multimeter and hand match the resistors to each other , if you test enough you will find pairs that are within 0.01% of each other ......

[!]

^ Maybe my math is off but it seems as though to get 0.01% matching, even if you start with 1% resistors instead of 5%, (so there's 2% total spread possible from +-1%), you'd have to test an average of 200 resistors. For all resistors in the circuit. Doesn't seem like a fun way to spend a few evenings to me... and to get this accurate a measurement you can't really rely on just touching multimeter probes to resistor leads, you'll need a clamping contact and to pass some current through them, as in a wheatstone bridge, Wheatstone bridge - Wikipedia, the free encyclopedia