Some thoughts on tweaking a pre4 (pre3 also)

[bikerhifinut]

After a chat with Dr John C yesterday at Tonbridge I have pasted in the following from a previous thread of mine as I feel it could be of interest to those wishing to tweak the gain on a pre4 or pre3 without the addition of the attenuator pot between stages.
It worked for me.

I bought the up to date latest PCB for the pre4 to upgrade the previously fitted pre3 board in my "frankenstein" Pre4 Homebrew lash up in a modded HD3 case, which if memory serves me right was used in the original PC era prototype pre4.

I've never been completely happy with it and after a bit of head scratching I figured that for one reason or another the signal grounding using the original pcb wass less than ideal.
So on inspection when it arrived I was impressed with the spacious layout and also that there were a lot fewer earth points to join up etc as I am sure that was the reason why I was chasing hum loops down for a long time after the initial build.
I took a look at how the attenuation potentiometer was effected and it took a bit of reverse engineering but I realised quickly enough that R2A and R2B were connected by a trace that meant no extra link wires were needed if I wanted to leave the pot out of circuit, which I did.
So I figured a more elegant way of dealing with the extra gain that I dont need would be to adjust the gain in the 2 stage circuit. That was why I was asking about the transformer turns ratio.
Anyway I figured that leaving the second stage unsullied was the best approach rather than faff around with feedback etc. And from experience with another homebrew preamp build I figured that an unbypassed cathode resistor on the first stage could be the answer, the degenerative cathode feedback would reduce the overall gain and quite possibly tidy up the already low distortion a bit. A fag packet calculation and experience from another preamp gave me the value of 7.2k for around 6dB gain and 3.6k for 12dB gain, it's the way Velleman dealt with their ECC82 preamp kit and I'm not ashamed to do a bit of plagiarism, and anyway its discontinued.
So in the absence of a 7.2 k I settled for 7.5, although 6.8 would probably have worked just as well, its not so critical here. Keeping the original 33k anode load too. I also reduced the HT by the use of a 47k resistor at R1 instead of the 1k on the pre3 diagram, I've got a 300V to 310V HT off PSU3 here with our very high mains. Anyway that works a treat bringing the HT at the anode to the intended 170V give or take. The HT to the second Valve via the output transformers was left as per Pre3, i.e. directly wired off the decoupling capacitor and 1k resistor as per original. I used a 100uF F&T on the PCB for the first stage after the 47k dropper.
So switch it on and all measured well. Initial listening test indicated this board could be a lot quieter than the pre3, certainly completely hum free and no discernible hiss. So far so good. I did use good quality standard metal films on the important resistors in the signal Path, matched per side on the DMM. a pair of nice Samwah 470uf 105degC on the second stage cathode.
However when I checked the gain, I found with a 6dB gain on the first stage, the total amplification was basically 0dB i.e Unity gain although there was promise in the sound quality, the level from the Squeeze box Touch wasnt really high enough. Bearing in mind my power amp needs 1V to drive to full power.
So I took out the 7.5k and subbed in a pair of 3.6k for a hoped for 12dB gain on that stage, on measurement afterwards the whole preamp was now showing 6dB gain on a 500Hz and 1Khz input from the signal generator.
Boxed it up and with a pair of US made raytheon 5814 (ECC82 equivalent) in and also tried a pair of NOS ex BBC Brimar 12AU7 for comparison. Both came in very similar.
I like what its doing now, it's got loads of low end thrutch which I felt was missing before and I can now see what all the fuss about transformer output preamps is about.
With this layout you would be ok with 0.1uF between stages, assuming a 1meg gridleak into the second stage, gives a cutoff around 1.5hz i think.
I used a pair of Soniqs 470nF polyprops that I had in the toybox looking for a good home and I'm well happy so far.
Ok it may not be for everyone in here, but it works for me and my needs and its an easy mod that you can reverse easily. I would recommend veropins on the component locations that are different from standard to save damaging the PCB.
Its certainly been interesting for me, and also has brought out the basic qualities of pre3/4 .
You could do a similar job on the pre3 board too.
Cheers.
Andy.