I use Electro-Harmonix Gold - Russian manufacturer selling internationally (just easy to buy here). I did not try any other manufacturers as yet... Heard good feedback on Genalex (more expensive though).
Try to buy the selected ones with matched triodes within each tube - this will ensure better balance (lower distortion, lower DC offset). Although DC offset may be corrected with the pot to some extent, matched triodes are still a preference.
Here is a good online store in Canada (Ontario):
The Tube Store - 12AU7 (ECC82)
They offer "selected" option.
Ahh... just realized... I answer in multiple threads, so things mix up a bit
This one does not require any matching. They only need to be matched for the other design with one tube in input LTP...
Sorry for possible confusion
Ahh... just realized... I answer in multiple threads, so things mix up a bit
This one does not require any matching. They only need to be matched for the other design with one tube in input LTP...
Sorry for possible confusion
All good. I'll be ordering a bunch of tubes so I'll likely get them all matched to avoid confusion anyway.
Jeff, looks very good to me!
Just please check some values with the final schematic in >post #35< - some caps and resistors were adjusted, comparing to earlier schematic.
Worth building. It will be interesting to know your listening impressions. This one has got a clear signature
Also, worth trying it with Lat-Fet OPS from TubSuMo. I did not try this combination as yet, but I surely will...
Cheers,
Valery
Yes, I also realized high 2-nd harmonic doe not scare me any more
It sounds cool. Experimenting with some alternative configurations, including the one with a "light" NFB (~6db loop gain).
Valery
Just found this thread - been considering something similar but using Quad "current dumping" output stage using Sziklai pairs.
On 2nd - it is still distortion. Now, the 12AU7, while cheap and cheerful, is not a low distortion device. The 6SN7 or 5687 are going to be better (by all reports). Should be able to reduce 2nd and 3rd a bit by swapping these out.
Lynn Olson's work suggests SRPP 6SN7 taking output from lower play would be best.
OR
I'd be interested in a wrapping a high NFB loop (ie a decent op-amp. Even the venerable 5532.) around the voltage gain stage - not the whole amplifier. This will keep the speaker OUT of your NFB loop.
As long as you don't clip Op Amps they seem (in my experience) to behave pretty well. Clip them (eg. by generating larger error terms) and they get ugly quickly.
(I'm trying to work out how to build a practical "limiting" error amplifier - needs to limit asymmetrically to minimise odd harmonics, too)
Some "brain food"
I forgot to mention that (if you can find it*) Bruno Putzeys penned an article on the impact of NFB on distortion spectra where he extends Bandaxall's work. Some leaps of faith are needed but it's a thought provoking read.
(*my copy is titled "The F-word" and marked "Linear Audio Volume 1 - Circuit design" but no date or place of publication)
Cool, I will find it
BTW, did you see a couple of my other designs with tube LTP at the input and high loop gain (very low distortion)? A different story, but also a good direction to go, IMHO...
>Low TIM, low distortion hybrid front-end<
>Ultra-high performance, yet rather simple - hybrid and more!<
If I understand the question, the rails should not be hot while the tube is warming up. I tried it and the rail voltage is all over the place for at least 15-20 seconds. If you let the tube warm up for 10 seconds or so the fluctuation will be greatly reduced. Valery's protection circuit is set up to turn on the tube warmer first, then the soft start and then short the soft start for full power. After all that there is a delay on the speaker protection side to let the rails settle at zero. This takes 5 relays, three on a separate board and the two on the protection board. I'm not sure you can do it another way unless you skip the soft start.
The rails won't be hot. This input uses an extra 250 volt supply just for the tubes that is fed by the same transformer as the heaters. It would need a second power delay relay.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.