The beauty of a decade box: ab763 tweaks

[fdesalvo]

Had a blast with the decade box and amp today. Amp is a single channel ab763 no verb/vib with (2) 12ax7s, (2) 6l6wgb, and a 5u4gb. 440pv. I tweaked a few different areas:

1. V1 plate loads - went up to 120K on both and found that the the amp lost some clarity and sparkle, taking on a more rounded tone that the ts wasn't able to correct. I ended up going back to 100K on both sides of v1.

2. PI plate loads - I moved from 100K/82K to 100K/100K, 100K/120K, and 47K/47K - the smaller the value, the thinner the tone. 100K/100K did sound good, but honestly, not enough of a difference to warrant a change. I tried this with both 12ax7s and a 12at7. Ended up back at the original formula with a 12ax7.

The best mods thus are are reducing the PI filter cap to 16UF and the preamp to 10uf. There's no increase in hum, flab, or other artifacts at volume and the touch sensitivity and bloom/envelope are just outstanding.

[Chuck H]

I've actually had my best results trimming the first and second preamp filter, rather than the PI. But tone and it's many guises regarding WHERE an amp is clipped and WHAT frequencies are accentuated is entirely subjective.

Hugely valid experiment! And whatever gets the tone you're after is fair game! Safe component operation is the only criteria!!! That's at least one reason I work mainly with tubes! They're so versatile and forgiving.

[fdesalvo]

Rock on, Chuck H(eisenberg)!

If I had more than one preamp tube, I might've gone that way haha. So funny how I've spent days researching subjective responses to, "How does it sound if I...?" questions, when in a quick 30 minutes, I was able to hear it for myself. Takes a little motivation to leave the internet comfort zone.

Oh, I forgot to mention that I paralleled a 68k input grid resistor to halve the single input's resistance (oops - didn't consider the input circuitry when moving to one input to save space). After all I read I was expecting very little difference. What I observed was added brightness and some nasty action accompanying strong pick attack - sort of a hardness to the attack. Running my ODs through the front end revealed the same. The amp no longer loved my pedal board. I quickly reversed that! Somewhere Leo Fender was self-face-palming.

[pdf64]

I would feel that the results may be too exposed to being skewed by parasitic capacitive coupling (due to added wiring to and from it) to use a decade box.
I prefer to tack in a trimmer in series with a resistor, eg 47k fixed in series with 220k trimmer for a plate load tweaking.

I agree that pre-amp supply de-coupling arrangements seem to have quite an impact. But I need to do some proper experiments to try and quantify it, it's just a perception thing at the moment, ie the bottom end feels more solid and pronounced with 47uF B+ de-coupling caps on the pre-amp, compared to the normal 16 or 22uF.

The normal input arrangement is that the 2 x 68k input resistors are paralleled by the switching on the jack sockets.
Pete

[daz]

I've done all that and a lot more but without a box, just tacking parts in. And i agree with a lot of your findings even tho it's a marshall style amp. I've made at least 3 turret boards with all new components from years of experimenting on this thing. Crazy, i know. But i've come to realize certain areas just work the way the original circuits were configured. 100k plates for example. Higher values *IMO* just don't sound as good ever. The tricky part with tweaks tho is that some things CAN sound good when it's really just your ear or some other thing that day thats tricking you. After enough times of trying the same tweak or better yet putting it on a switch so you can hear it over and over, you begin to get a very clear picture of something that at forst wasn't remotely clear at all. I've come to learn that after years my amp is mostly where it should in that 90% of the values are absolutly right. V1's 2 stages are a good example. I've had every value and combination of cathode R and cap and no matter what i use, 1.5k/.68uf always works best on the first 2 stages. ALWAYS.

At that is of course valid only to my ears or others with my same likes/dislikes. But many IMO are pretty universal in that i think 99% of people would prefer thenm if they heard the comparison over and over. Right now i have the preamp up to the PI pretty much where i can't change anything because it's all been scrutinized for years till no changes ever seem to be for the better. So if it's not obvious by my latest post about tweaking my master section on back, thats where i've been experimenting the last few weeks. And i have learned a a LOT about the character of NFB especially. This is one area that i have tweaked a good bit, but never really to this degree and never with the A/B testing i;m doing. I feel the preamp is good now but that there is still gold to be found in the PI/PA that i have yet to discover. I find the whole thing a fun game which is why i never stop tweaking. Thats not to say i wouldn't stop if i found what i considered close to perfection. But then what would i do?

[fdesalvo]

I hear that. NFB tweaks made a huge difference in tone and feel. I think the pi tail resistor is the final frontier for tweaking on this amp.