This circuit is kind of a mash up of a BA6A front end with a 436ish rear. I was looking for another fairly simple limiter circuit similar to the Altec 436 and Federal 864 that was also cost effective. I've had this circuit built up on the bench for the past week and am pleased with the results so far. The BA6A balance circuit works very well and I've had zero hum issues. The overall compression sound is closer to the 436 which to my ears is a much different animal than the Federal circuit. One is more visceral and the other kind of politely and evenly flattens itself.
Accepting the limitations of the simpler limiters I tried to focus on what I'd likely be using it for the most while experimenting with the threshold voltages and attack and release circuits. Mixdown duties was the call and I tried to optimize things for the 5-10db GR range. The original double time constant values of the BA6A - 0.1//3.3M + 3.3M in series with .27 gave too quick a release and the attack and release sounded badly out of sync. Changing the 0.1uf cap to .57(.1 + .47) gave much better results. This combination also worked reasonably well at very light GR (0 to 3db) which was surprising as these type circuits seem prone to more thumping and transient splatting at low GR settings.
I'm an amateur at this and there's likely some typical mistakes so bear with me. With the circuit as drawn I get the best performance with a threshold voltage (100K pot wiper) of 15-25V. If I try and lower this to the 3-8V range it sounds pretty bad - too much splatter even at 2db dives. At the 15-25V setting I can get light transparent GR or drive it hard and it sings. One thing I'd like to improve is bass distortion. It's not terrible and depending on the source material it can sound great (70s pop) or go blatty on heavy bass material (lots of modern digi stuff). Lower the SC caps? GR stage PS regulation?
I'm open to any critiques and suggestions for general improvement. I'm planning on working out a TC network for at least one good faster release time - for effects use. So far I'm getting the feeling that trying to go fully variable on attack and release may result in too many 'train wrecks' so for now I'm sticking with one good setting for program and one for lots of meter movement. I also think the 100K pot could be replaced with a 50K.
Accepting the limitations of the simpler limiters I tried to focus on what I'd likely be using it for the most while experimenting with the threshold voltages and attack and release circuits. Mixdown duties was the call and I tried to optimize things for the 5-10db GR range. The original double time constant values of the BA6A - 0.1//3.3M + 3.3M in series with .27 gave too quick a release and the attack and release sounded badly out of sync. Changing the 0.1uf cap to .57(.1 + .47) gave much better results. This combination also worked reasonably well at very light GR (0 to 3db) which was surprising as these type circuits seem prone to more thumping and transient splatting at low GR settings.
I'm an amateur at this and there's likely some typical mistakes so bear with me. With the circuit as drawn I get the best performance with a threshold voltage (100K pot wiper) of 15-25V. If I try and lower this to the 3-8V range it sounds pretty bad - too much splatter even at 2db dives. At the 15-25V setting I can get light transparent GR or drive it hard and it sings. One thing I'd like to improve is bass distortion. It's not terrible and depending on the source material it can sound great (70s pop) or go blatty on heavy bass material (lots of modern digi stuff). Lower the SC caps? GR stage PS regulation?
I'm open to any critiques and suggestions for general improvement. I'm planning on working out a TC network for at least one good faster release time - for effects use. So far I'm getting the feeling that trying to go fully variable on attack and release may result in too many 'train wrecks' so for now I'm sticking with one good setting for program and one for lots of meter movement. I also think the 100K pot could be replaced with a 50K.
