under-spec filter caps?

You can change the *capacitance* value, not the *voltage* handling value. There isn't a standard as it depends on the circuit, but you'll often see people use more modern values, like 22uF, than older values like 8uF or 16uF, as they are easier to find, often help reduce hum, and usually improve the bass response.

 

For "spongy" I recommend these:

 

yer gf's IUD? seriously ... what corliss said. you want to keep the voltage designations up or you risk damage. personally, I'd keep the uf values the same unless you want to change the voice of the amp

 

Have never seen lower than 20 where a solid-state rectifier is present, or where the amp is rated at 15W or more.

 

There were two main reasons for low value capacitors in older valve rectified amplifiers, the main being the maximum permissible 'tank' or first filter capacitor for the valve. For example; a GZ34 is 60uF, 5Y3 is 20uF, EZ80 is 50uF, 5U4 is 40uF and 6X4 is a mere 10uF etc. Any slight increase in the maximum value shortened the life of the valve by stripping the cathodes. Bear in mind, electrolytic capacitors had a tollerance of +50 -10% so a 'good' 50uF capacitor could measure as 70uF quite easily when new and fresh.
The second was cost and availability. Back in the 50s the largest commercially viable capacitance was 32uF and 16uF was the most popular. In the 40s 8uF was the norm.
The comming of the silicon rectifier, BY127, BYD33v, 1N4007 and the newer UF4007 meant the limit was the HT windings maximum current capacity. With modern winding techniques, most issues of loading are no longer an issue.

 

Nope - solid state amps don't work the same at all so that's not a thing.

You should still be able to find hot rods all day long in the $400 range.

 

I hotrodded a little solid-state practice amp once but it needed bigger caps, not smaller.

Started with a better speaker but it made almost no difference because the circuit was EQ'd super thin for the crappy speaker. Figured out that the bass was being cut in the feedback loop of the first gain stage. Swapped out a cap in the feedback loop to boost the bass, then it would oscillate uncontrollably since it was now amplifying frequencies below the cutoff frequency of the power supply filters. Increasing the filter caps lowered the cutoff frequency and killed the oscillation.

So the filter caps themselves weren't responsible for changing the sound of the amp. Changing the filter caps was necessary to make the circuit mods work. The circuit mods were necessary to make the speaker upgrade sound better.

 

The amp I modified was an Esteban G-10. The circuit board is labeled Belcat. As I understand it, Belcat makes the same circuit board for many different brands. That's why you see so many "G-10" models.

Replacing the caps is super cheap and easy. The speaker upgrade is the killer. I got a 6.5" Eminence "factory buyout" deal from Parts-Express for $9. They don't have them anymore. I actually bought it for my Champ-clone and had it left-over after upgrading the speaker in the Champ. Fitting it into the tiny amp required removing the grill cloth, cutting a bigger hole in the baffle board, then stapling the grill cloth back on. Also had to move the transformer to clear the magnet. It was a fun project for using up spare parts but I wouldn't really recommend someone going out of their way to build one.