I would put the zener before the 1k. The 1k is the stopper, in my builds I usually cut the G pin and solder the 1/8W resistor between the PCB and the G, for shortest distance (the G pin does not even touch the PCB with all its strays).
@Rod Coleman commented that 1k may be too high, I am now using 680R, Euro21 posted 100R, probably also ok.
I know, this does not really help.
@Rod Coleman commented that 1k may be too high, I am now using 680R, Euro21 posted 100R, probably also ok.
I know, this does not really help.
I usually use -automatically- even 1k grid stoppers for medium-high gm tubes, but high transconductance FETs are another animals.
Sometimes too large gate stopper occurs anomalies.
Try first none, if oscillation created, use it from 100R.
Protecting zeners: generally I connect its directly to FET pins. Its capacitance always modified behaviour of oscillation, so gate stopper value "sizing" depends of the zeners too.
Sometimes too large gate stopper occurs anomalies.
Try first none, if oscillation created, use it from 100R.
Protecting zeners: generally I connect its directly to FET pins. Its capacitance always modified behaviour of oscillation, so gate stopper value "sizing" depends of the zeners too.
Stereo 20 is the ancient UL PP, with gNFB.
Don't assume that other solution what used, if it completely different.
Trioded EL84 has about half, even one third (depends of UL ratio) Ri than in UL.
50uF has 159R impedance at 20Hz, 100uF has 79R.
Put its together with 3.5 .. 5.6k and 1.7k.
The problem I have to face in the Leak Stereo 20 is not exceeding the current capability of the mains transformer. The original circuit has 4x EL84s at 41.5mA each = 166mA, plus 5 of ECC83 across the amp and preamp, which probably doesn't exceed 15mA. So ball park current capability is 180mA by my calculations. Keeping the EL84s as is leaves me with about 15mA to spare for the input stage. Start with 2x 3mA for the mosfet splitter and that leaves around 4-5mA for the input tube. Which is why I asked about running the 6V6 at 4mA. That works fine for a 6J5 or 6P5, but more current isn't going to happen in the Leak. I could use types 37 and 76 but the UX5 sockets are too big, and anyway I really should use a common tube type like the 6J5/6P5 in the Leak. I use quite esoteric tubes mostly but this is a more conventional amp.
I have another PP amp planned, and in that I can use a more massive mains transformer so anything is possible, including using the 26 as driver since that's my go-to driver/input stage in my 2-stage designs with high gain output tubes. And that's where the phase splitter will be on the input of the amp as in the original title of the thread. I put the input and output stages in different chassis to cut down on the weight of each, since i don't like working on anything above about 10kg these days.
I have another PP amp planned, and in that I can use a more massive mains transformer so anything is possible, including using the 26 as driver since that's my go-to driver/input stage in my 2-stage designs with high gain output tubes. And that's where the phase splitter will be on the input of the amp as in the original title of the thread. I put the input and output stages in different chassis to cut down on the weight of each, since i don't like working on anything above about 10kg these days.
Last edited:
I guess it's all about production and quality control. The 6P5 came later when pendotes were first priority.....you might have some better samples but as you say they should be more or less in the same ball park. More options....
I only have the GT type of the 6P5. Looks like late 50ies early 60ies.
If you run the EL84 in triode mode, there is no reason to run it at 40ma+ each. Drop down to 30mA.
I've narrowed the choice of input tube down to either 6P5GY in octal or 37 or 76 in UX5. Pretty much the same tubes, 37 is lower gain.
My question for the mosfet splitter is whether the output impedance of this input stage at up to 12K would drive the mosfet OK. And also what would be the preferred mosfet to use at around 3mA?
Build is progressing. This will be a trial run for my next PP amp.
My question for the mosfet splitter is whether the output impedance of this input stage at up to 12K would drive the mosfet OK. And also what would be the preferred mosfet to use at around 3mA?
Build is progressing. This will be a trial run for my next PP amp.
Hello revintage - can you PM me? I can't seem to PM you.
The reason for UX5 is to use a 37 - mu=9. Or a 76, mu-14. With the octal 6P5 is also mu=14. I don't know any good sounding novals with mu of 14 or lower and that don't require more than around 4mA current. In fact I don't know of that many really good sounding novals, though there are some exceptions.
I found this quote, didn't make a note of who said it....
"12ax7s and 12au7s sound constricted, colored and undynamic when compared to most pre-WW II triodes, including 27, 37, 56 and 76. Most small-envelope dual triodes were designed after the advent of negative feedback. Feedback was almost mandatory to improve those tube's linearity. Early triodes needed to be extremely linear, because the negative feedback technique had not been invented yet. Manufacturers who use 27 or other pre WW-II triodes do not do it to be eccentric, they use those tubes because they are VERY linear devices and DO NOT require music-killing feedback to sound good."
The reason for UX5 is to use a 37 - mu=9. Or a 76, mu-14. With the octal 6P5 is also mu=14. I don't know any good sounding novals with mu of 14 or lower and that don't require more than around 4mA current. In fact I don't know of that many really good sounding novals, though there are some exceptions.
I found this quote, didn't make a note of who said it....
"12ax7s and 12au7s sound constricted, colored and undynamic when compared to most pre-WW II triodes, including 27, 37, 56 and 76. Most small-envelope dual triodes were designed after the advent of negative feedback. Feedback was almost mandatory to improve those tube's linearity. Early triodes needed to be extremely linear, because the negative feedback technique had not been invented yet. Manufacturers who use 27 or other pre WW-II triodes do not do it to be eccentric, they use those tubes because they are VERY linear devices and DO NOT require music-killing feedback to sound good."
I laugh at the "music-killing feedback" part of the quote. That's a cue that the person you quoted have no clue what he/she is talking about. A triode, any triode has internal feedback all day long. Shouldn't that make triodes "music-killing" by design? lmfao. This is cork sniffing at the highest level.
I know nothing of Mosfets - I was hoping you or others would tell me which of these are usable and which are best. I assume N channel?
IRFBC20 - 2.2A, 600v
IRF830 - 5A, 500v
FQNL2N50BTA - 350mA, 500v
FQP1N60 - 1.2A, 600V
AOT1N60 - 1.3A, 600V
LND150 TO-92 - 30mA, 500v
DN2540 - 500mA, 400v
IXTP01N100D - 400mA, 1000V
FQU1N80 - 1A, 800V
other?
Last edited:
People, regardless of age, can learn new things. A year ago I knew next to nothing about MOSFETs and today I use them in every amp I design. I'm a fan of LND150 and encourage simulating the project before building. Simetrix is relatively easy to learn and use. The parasitic capacitances are low in LND150 and it's depletion mode, which makes the biasing a little simpler. Treat it like a triode without the need for a heater supply.