Returning to the sound of tubes - I've also tried many configurations with the 12AT7 and almost all were disappointing. In a 1961 amplifier book a schematic appeared which was called "ultimo" (the best) consisting of a 12at7 fase splitter and EL84's pp. Unfortunately the sound was absolutely inferior to the ECC83 EL 84 pp I had built before (using the same transformers). As I wanted the 12at7 to sound good I tried increasing the current and yes it sounded better at at least 6ma but still only so-so.
Later on a srpp was tried, also transformer loaded but nothing sounded really good.
Then an ECC85 was tried and that was a different story. I don't understand why it isn't used more. In the eighties I showed it to audiophile friends and this resulted in a direct coupled EL 84 pp amp later published in Electronics World. Hiraga used it as a driver in a 6C33 se amp. I also found an md preamp using the ECC85 on the internet and one more as a 300B driver. However the 12AT7 can be found in numerous schematics.
Comparing the characteristics of the 12AT7 and the ECC 85 it's hard to believe that tubes seemingly so similar can sound that different.
Later on a srpp was tried, also transformer loaded but nothing sounded really good.
Then an ECC85 was tried and that was a different story. I don't understand why it isn't used more. In the eighties I showed it to audiophile friends and this resulted in a direct coupled EL 84 pp amp later published in Electronics World. Hiraga used it as a driver in a 6C33 se amp. I also found an md preamp using the ECC85 on the internet and one more as a 300B driver. However the 12AT7 can be found in numerous schematics.
Comparing the characteristics of the 12AT7 and the ECC 85 it's hard to believe that tubes seemingly so similar can sound that different.
KeesB, I invite you to share your experiences and conclusions.
I'm a bit like you: "what if I tried that" and for a few years now, when I didn't like tubes, I collect all the tube amps I can with different topologies and schematics for "when I 'will have time'.
I've "played" quite a bit with it in recent years, but a change in family life has changed things, I'll come back to it later.
I'm a bit like you: "what if I tried that" and for a few years now, when I didn't like tubes, I collect all the tube amps I can with different topologies and schematics for "when I 'will have time'.
I've "played" quite a bit with it in recent years, but a change in family life has changed things, I'll come back to it later.
no messing around, I've been an after-sales service technician since I was born
and in several areas and this t-shirt would be a hit in the industry 🤣 🤣 😁
Huggygood, apart from tubes already mentioned I've had good results with triode connected PCF 80 (mug2g1 = 47 muT=20); PCF 82 (mug2g1= 32, muT=35); PCF 802 (mug2g1=47, muT=70); ECL 84 (mug2g1=36, muT=65);
4BC8 (mu= 35); 6CG7 (mu=20).
Because few people use them p-tubes have the advantage of being cheap and widely available. I avoid using all 88's , PCF 189 and PC 900.
4BC8 (mu= 35); 6CG7 (mu=20).
Because few people use them p-tubes have the advantage of being cheap and widely available. I avoid using all 88's , PCF 189 and PC 900.
Just a couple of random thoughts. I would make sure that you do not use global FB because as was mentioned earlier it adds a real headache when using tubes with not only different gains but also different capacitances. If you used local FB in the VAS it would help but that adds another set of resistors that have to be "OK" for all.
Perhaps you could use an LED on the cathode and an adjustable active load on the plate. If the CCS was not adjustable you would have to use a current appropriate for the tube with the least current capacity.
Perhaps you could use an LED on the cathode and an adjustable active load on the plate. If the CCS was not adjustable you would have to use a current appropriate for the tube with the least current capacity.
Not surprised that the AX7 didn't do too well as 100K is pretty small load for it but kind of surprised how good the AU7 looked. Of course simulated HD is notoriously unreliable. You could have a switchable plate load pretty easily with one setting for the tubes that like a large plate resistor and another for those who prefer a smaller one.
P.S. Actually the HD for the AX7 was improved most by using a diode with less voltage drop.
Last edited:
