Comparing a 12ax7 LTP stage with a cascode 12ay7 stage. The later has about 5dB more gain but more distortion. The miller effect will be better but whether that's a help with the NFB is debatable if you need a dominant pole here anyway. I think for 17-20dB NFB you need about 300-400x gain from the input to the grids of the KT88 to get 30-40x gain in the closed loop amp, which is about right for modern inputs + volume control. I have an interest here in building this which is why the comments.
Hit a bit of a snag today, looking for suggestions on a fix. After the amplifier has warmed up the cathodes of the bottom 12AX7 and the 6922 seem to be oscillating at a low frequency, this also shows up in the measurements. I attached one of the measurement plots to show this as well as the power supply schematic and transformer I am using.
Attachments
Sure, it is always interesting to explore other options!Hi, would an hybrid cascode with a BJT on top be something you could be interested into?
Thanks, I’ll be sure to give that another look if I run into LF stability issues again.R16, C20, R41 on my last schematic post may well fix your LF stability issue. The idea is is to a a zero-pole which is zero cancels one of the coupling poles. Too many coupling LF caps results in this problem - you have transformer + 2. Its a simple mod to the NFB.
P.s. I’ve also sent you a private message.
Till so far i see that your product with the pp KT88 has a lot more* distortion then your product with the se 6c33cb. How is this possible?
* the KT 88 has more power but till 10W the 6c33 wins easely. And the SE has even a less good output transformer (lots of copper losses and less bandwith)
btw the ecc83 is good for guitar, not audio
* the KT 88 has more power but till 10W the 6c33 wins easely. And the SE has even a less good output transformer (lots of copper losses and less bandwith)
btw the ecc83 is good for guitar, not audio
Last edited:
- Home
- Amplifiers
- Tubes / Valves
- Cascoded 12AX7 LTP - Stupid good performance?