I like the "small cap to ground" idea.
Around 180 pF would be comparable to a typical 12AX7 guitar gain stage ( The Valve Wizard ).Probably not a good idea. A 68k grid stopper is already the dominant source of noise in a 12AX7 input stage - it generates more input noise than the triode itself does!
With maybe twenty to fifty times less input capacitance, you'd need twenty to fifty times more resistance, and that would generate an enormous amount of Johnson noise, and you'd have a very hissy guitar amp.
Perhaps it was too much gain they were trying to tame, rather than stability problems?
For what it's worth, I've experimented with two or three small-signal TV and UHF pentodes in guitar preamps. I didn't have any stability problems with any of them. I did pay attention to layout, though, and used the pentode as the second gain stage, rather than at the input.
I know Merlin Blencowe devotes a whole chapter of his preamp design book to cascode gain stages, but I don't really understand his enthusiasm for them. Particularly when there are still several small-signal pentodes that can be found on the $1 menu at ESRC, often with a free triode thrown into the same glass bottle. 🙂
-Gnobuddy
In my guitar amp with that circuit configuration, i tried pentode(6sj7, ef86), cascade with ecc83, ef86+12ax7(parallel).
However, i like too much the cascode circuit...it sounds very good to my ears.
However, i like too much the cascode circuit...it sounds very good to my ears.
Last edited:
vingenz, can you tell us more? Cascode with what valve (ECC 83?)? And can you describe how the cascode sounded different from the pentodes to you? (Hard to describe sounds, I know, but I am curious what differences you found.)
Thanks!
-Gnobuddy
Thanks!
-Gnobuddy
ehehehehe...hard to describe...
with pentode the amp lost some frequency...are not focused on guitar frequency and so loudspeaker become more critical for the sound. With cascode (i tried only 12ax7) the amplifier have more gain, rounded bass and clean headroom.
with pentode the amp lost some frequency...are not focused on guitar frequency and so loudspeaker become more critical for the sound. With cascode (i tried only 12ax7) the amplifier have more gain, rounded bass and clean headroom.
I have zero experience with cascodes, but, between those two valves, I would suggest the ECC82.
The reason is that the E88CC is an RF triode, and has extremely high transconductance. This can make it "want" to oscillate, i.e., it will be more sensitive to layout than older, lower transconductance, types like the ECC82.
Back to your original problem, have you tried putting a series resistor (say 10k ohm) between the anode of the upper ECC82 and the tone-control circuit?
This is called a "build-out resistor", and it can prevent circuit instability due to difficult loads (in this case, your tone control circuit might be the difficult load).
-Gnobuddy
The typical 68k grid stopper was a leftover from the Fender Hi/Lo input attenuation network arrangement. As suggested here, it does help to guard against local radio station interference, although perhaps with s somewhat excessive high end roll off. A value of 10k will be just as effective, while preserving the high end sparkle of the guitar tone. Alternatively, the residue could be replaced by a ferrite bead.
Sent from my phone with Tapatalk. Please excuse any typpos.
From my recollections of the time Merlin wanted a less microphonic alternative, with a similar sound to a pentode.
I also relay like the idea of using TV pentode-triodes like the 6U8.
Sent from my phone with Tapatalk. Please excuse any typpos.
With roughly 200 pF input capacitance (that includes Miller effect and a voltage gain of 60x), the upper -3 dB frequency is around 12 kHz. That's probably plenty for electric guitar - there doesn't seem to be anything of musical value above 6 kHz from a typical electric guitar. And certainly a typical electric guitar speaker won't even come close to reproducing 12 kHz signals, even if the guitar produced them.
But the 68k input resistor also actually generates more broad-band noise (Johnson noise) than the input valve does. As Merlin Blencowe calculates in his preamp book, that 68k resistor is actually the largest source of hiss in the entire guitar amp. Now, that's just silly!
By the way, this noise problem is true even if that 68k input resistor is a low-noise metal film type. Those people who prefer grotesquely primitive carbon composition resistors will be rewarded with far more hiss.
I've never tried the ferrite bead, simply because I never had one. I usually just use a 10k metal film grid stopper, as AquaTarkus suggested.
-Gnobuddy
Please note that the standard Fender input is TWO 68K parallel; about 34K on my slipstick.
I'm also happier with 100pFd-130pFd than 200pFd at a 12AX7 grid.
35KHz seems fine to me.
Yes, 34K is about 2.6uV hiss over 20KHz, and some 12AX7 may duck under 2uV. This has to be Player's Choice. Few venues really need lowest hiss. OTOH, if you ever have to play at a truck-stop, near a police station, or (gasp) in an AM radio studio right under the transmitter (we used to do that!), RF break-through totally spoils the gig.
I'm also happier with 100pFd-130pFd than 200pFd at a 12AX7 grid.
35KHz seems fine to me.
Yes, 34K is about 2.6uV hiss over 20KHz, and some 12AX7 may duck under 2uV. This has to be Player's Choice. Few venues really need lowest hiss. OTOH, if you ever have to play at a truck-stop, near a police station, or (gasp) in an AM radio studio right under the transmitter (we used to do that!), RF break-through totally spoils the gig.
- Status
- Not open for further replies.