This is not just another gainclone

I probably went a little overboard with this one, but couldn't resist the temptation to use those cute heatsinks. So here's my next gainclone implementation.


  • gc1.jpg
    51.5 KB · Views: 63,256
I was concerned that those heatsinks might ring a bit, but overall appeal was much stronger than common sense and the whole creation is more of a modern sculpture than an amplifier.

The big chunk of acrylic in a center is there to add mass to the structure as well as to damp vibrations created from the outside. The white phenolic block is used to mount PS caps (stripped BG) and Hovlands coupling caps.

The Black Beauty pot is at the back and the shaft tunnel ;) had to be drilled through the acrylic. I like the see through effect, which adds lightness to the structure.

There will be no cover and this is an open concept amp.;)


  • gc2.jpg
    74.9 KB · Views: 58,820
I'm using 100k pot, 2u coupling cap, 10k series resistor and 220K feedback resistor. Positive input is directly shorted to ground.

This amp has no hum whatsoever and the bass is really deep. Smaller value cap makes it probably faster and 1000u BG seems perfect for this application. What I like about such capacitor bank is that it can be basically mounted on IC's pin and the signal path is very short. But it would be interesting to try larger values to see a difference.

I do everything by hand, no milling machine, so manufacturing chassiss parts is not very economical for me yet.;)

Acrylic is very friendly to machine though. The 2.5" block was easily cut on a table saw with Freud laminate blade and the cut is perfect, no chipping. The front and rear panels are rough cut, no sanding, and they look quite nicely this way. The acrylic block was further machined on a router to achieve the angles and uniform matte finish.

The hardest part was to drill the hole for a pot's shaft as the plastic stuck to the drill bit as it got hot, so I had to go slow, changing the size of a bit often.
This amp is very sensitive to proper placement, so a rigid chassiss is required.
If you omit the resistor and parallel cap from positive input to ground (as to Thorsten schematic) the sound becomes somewhat cleaner. I'm also not using 0.22 ohm series resistor at the output, as well no 1u cap between rails (not enough space). So basically you need only 2 resistors to make implement this amp. I didn't play with different types (as it surely make a difference) and I'm using Rikken carbon film types. Probably with this amp they might be beneficial, adding a bit of warmth and body to slightly forward sonic signature of the chip.

The way I build my first one was to make everything compact as possible and reduce possible vibrations. Thats why the chip is directly mounted to 1/2" aluminum rear panel and PS caps have special mounting block to avoid resonances. The first one sounded very well, will see how this one performs. Have to finish wiring and PS first.


Paid Member
2002-08-14 1:47 pm

I still haven't tried this on mine, but please keep in mind opamps achieve their best CMRR when both inputs are driven by the same impedance. It may sound significantly better if you connect the + input to ground through the parallel value of your feedback resistor and whatever is the average value of the input resistor (probably influenced by the volume - not a very good thing). If you omit the capacitor you may just suffer a bit more resistor noise, likely not an issue at these levels.
I understand you like the sound more without the cap, who wouldn't, but the resistor may have to stay. Experiments with opamps have thought me this, may, or may not be true in the gainclone.



Initially I had 220k/0.22u cap on positive input, then shorted it out. Again, when at friends place, removed the short and then put it back. When the input was shorted the amp sounded more transparent and dynamics were better. It would be interesting when more people try it and comment.

Here's another mod, I didn't try yet:

3. Place a 1.5K 2W carbon resistor from the output to the -v supply. I have
mentioned this before but I think its worth mentioning again. It seems to clean
the amp up that last little bit. The output is biased into class A. Not that
the amp becomes a class A amp as its only doing 25ma or so but it seems to be
enough as I guess the output does not switching off at any stage, so you don't
get any switching noise - well thats my theory why it works.



Previously: Kuei Yang Wang
2002-11-10 5:35 pm

Peter Daniel said:
If you omit the resistor and parallel cap from positive input to ground (as to Thorsten schematic) the sound becomes somewhat cleaner. I'm also not using 0.22 ohm series resistor at the output, as well no 1u cap between rails (not enough space).

Please note that I agree with taking out (most?) of the parts you eliminated. I put them in to ensure minimal DC offset and near unconditional stability, given that you never know who is gonna wire the amp how and what sort of speaker cable they'd use.

So all the "left out" are desirable to be left out, however, for beginners who have little experience in hardwiring and keeping current loops short the added parts may make the difference between a working Amp and VHF transmitter... ;-)