Let’s say the spec calls for a 50uF capacitor with a tolerance or +/- 20%. Could I safely replace it with a 45uF or 55uF capacitor, since the new part is within the tolerance of the old part? Or could there be problems because I’ve moved the center of the tolerance window up or down slightly?
Such a part is only used in noncritical places anyway, like power supply decoupling,
so usually there would be no problem with subs of a value even farther off than that.
And the part is likely to change value considerably more than that during its life.
so usually there would be no problem with subs of a value even farther off than that.
And the part is likely to change value considerably more than that during its life.
Right. But if the specified tolerance for the part in that location is 20%, I should be ok replacing it with a part of a slightly different value as long as it’s within that tolerance? Such as a 47uF electrolytic capacitor for a 50 uF.
Let's do the math by working out the design range. For a 50uF 20% the range is 40-60uF. If you replace it with a 45uF then the tolerance on that needs to be 10% (not 20%) to stay within the design range.
Please never lose sight of this: It is just a guitar amp, not precision NASA lab equipment. None of these values are critical. You will find, these days, that 50uf is not a standard value, the standard is 47uf, which is surely close enough. Leo Fender built amps in the 1950s and 1960s using 20% tolerance resistors. Notes on the schematics said all voltage readings to be within +/-20%. And those old can caps everything used? It was quite common for those to have a rating of +80/-20%. I am not making that up. As others said already, the fact it is a 20% part in the first place tells you it is not critical as to value.