Rectifying bridge.

Sean, sounds like a digital RF amp ur talking about...maybe even a class E? Stil hard to see how 1% suply voltage change alters the output 18%... sounds like these are switching diodes, not rectifying power diodes.

Anyhow you want soft recovery diodes, like HexFreds, fast recovery and turn on will usually add to the noise...

sometimes changing the diodes will really help, sometimes not. Try it.

There is absoloutely no need to "match" diodes - they do not have gain like transistors. All you have to do is get 4 of the same part number. Of course, you have to wire them correctly too. :- )

As far as Hexfred bridges, there may be some, just check around the parts houses and IR's website.

Alternately, you can get very close to the same results by adding a "snubber" circuit around that bridge. Using a higher PIV bridge will only buy you some immunity from wierd transients. A higher AMPERAGE bridge will likely be slightly larger and run cooler. Doing both will up the margin against failure.

The main difference between low PIV and high PIV rectifiers is the way they came out of the wafer when they were made. So in theory the higher PIV ones of a given type designation are "better." In practice, no difference if they are properly rated for the applicaton.

Sean, dunno what a 6A4 is... a tube? But the turn on voltage of a diode in a bridge that is rectifying lots of volts is in essence meaningless. There are some applications where the forward bias voltage of a diode makes a difference, this isn't one of them, imho.

If you're using them to bias something, then it makes a difference, but then you have to ask what's the variation in the part you are trying to bias with the diodes? :- )