"Mullard 5-20 KT88 PP blocks!

[SGregory]

dB = 20 log (Vnofb/Vgnfb). Use V rms at a given input and freq. You will also need to tune the feedback cap based on a square wave signal.

Somebody verify my math.

[smbrown]

Okay, so if my gain is 314 and 15.8 it would be 20log(314/15.8) = about 26db?? Seems very high.

[BillEpstein]

Finally got my endbells back from the powder coater and now I can begin stuffing and soldering!

[BillEpstein]

Installed the power tranny, last piece of the puzzle in the plate and just had to see what awaits

[SGregory]

smbrown,
You had a gain of 315 without feedback. That seems high.'

Bill,
Your build looks very good. You may have to rotate your OPT if you find hum an issue. I like that you have the input and LTP far away from the PT and choke.

[nanchangbob]

I finally got through this thread and I was hoping to get a little more information on the sound. Sam how do you like it? You said the gain seems
to be a little high. How does it sound to you?

Bill did you ever finish it yet? Your build looks great.


One question for you guys. I have two James 5k 6255HF 50W output transformers. Would these work for this build or are they to small? I was going to put them in my current amp, but this thread got me thinking.

[Eli Duttman]

I have two James 5k 6255HF 50W output transformers. Would these work for this build or are they to small?

Unfortunately, the James trafos are too small for use with a KT88 Mullard setup. Remember, the O/P trafo is inside a GNFB loop and magnetic headroom is essential. You could use Mullard style circuitry with your "iron" in combination with EL34 family tubes. The yield would be approx. 35 WPC.

[Johan Potgieter]

Originally Posted by smbrown

Okay, so if my gain is 314 and 15.8 it would be 20log(314/15.8) = about 26db?? Seems very high.

I did not check the 314 figure, but regarding 26db of NFB, that is not unduly high. Most of the famous Leak amplifiers used that: the Mullard 5-20 could stably work with up to 30dB.

Point is to check for possible instability at band ends. Perhaps old hat around here, but the easiest way for h.f. is by a square wave, determining where if any there are peaks (those would be way above the audio band at around 60 - 120kHz, but still be important for NFB stability. Remember the loudspeaker can be a very reactive load!) This performance would be totally dependant on the output transformer, so no advice to be given off-the-floor.

Should you feel uncomfortable, bring the NFB down to 20dB; no circuit in front of me now to advise.

[smbrown]

Actually, I ended up removing the negative feedback all together. I also triode wired the EF86. There is still a ton of gain, but the addition of a volume pot solves that. The amp is dead quiet and with out the NFB, sounds more open and less congested, as you might expect. All in all, probably the best PP amp I've heard. I'm still playing with the idea of putting in 300b's but the B+ would have to be reworked some (cathode bias would take up some of the extra).

[SGregory]

smbrown,
Your comment on limiting ther NFB is spot on with my observastions. I use a diffent input tube but have really limited the use of NFB as well. These are good sounding amps.