I'm having some issues with an amp I'm building:
When I turn the volume pot past 3 o'clock, some very undesirably screeching occurs (high frequency). What could be the cause of this? Circuit and sound file are attached!
Output tubes are 2x6L6 PP with 470 ohm screen resistors, grid biased at -30v for 38 watts of power.
The 50p cap in the tonestack should still be changed to 500p, I just didn't have one laying around.
https://soundcloud.com/mk5-studios/amp-nosie/s-3io9u
In the clip I first play a note, then I turn up the volume. Second time you hear it, i'm turning it up with the guitar plugged in, third time is without guitar.
When I turn the volume pot past 3 o'clock, some very undesirably screeching occurs (high frequency). What could be the cause of this? Circuit and sound file are attached!
Output tubes are 2x6L6 PP with 470 ohm screen resistors, grid biased at -30v for 38 watts of power.
The 50p cap in the tonestack should still be changed to 500p, I just didn't have one laying around.
https://soundcloud.com/mk5-studios/amp-nosie/s-3io9u
In the clip I first play a note, then I turn up the volume. Second time you hear it, i'm turning it up with the guitar plugged in, third time is without guitar.
Attachments
Last edited:
I don't understand (just woke up , could be my bad!).
How can you add something in series across something? (also english isn't my mother tongue)
Place a 4k7 in series with a 220p cap, solder them together. Place the two components across R8. That will damp the HF response on that valve.
Then Grid stoppers right on the sockets of each triode might help.
Then try 10k "pot stopper" between C9 and R10.
These are mods that have worked for me in the past.
Then try 10k "pot stopper" between C9 and R10.
These are mods that have worked for me in the past.
> What could be the cause of this?
Layout.
The real cause is GAIN. You have a gain of over 5,000 there, plus maybe 10 in the 6L6es, 50,000. If the *least* signal sneaks from 6L6 back to V1, it gets amplified again, and again.
Keep your big tubes and the OT wires very far from V1.
Also V2 output side should be very-far from V1 input. Not so much gain here, but they may be in the same bottle.
Layout.
The real cause is GAIN. You have a gain of over 5,000 there, plus maybe 10 in the 6L6es, 50,000. If the *least* signal sneaks from 6L6 back to V1, it gets amplified again, and again.
Keep your big tubes and the OT wires very far from V1.
Also V2 output side should be very-far from V1 input. Not so much gain here, but they may be in the same bottle.
I do indeed still have to add grid stoppers. I didn't because the Marshall plexi 1959 doesn't use them and that one's also quite high on the gain levels. The OT wires are as far as they can get from V1(12ax7, so together with V2). I'll post updates tomorrow!
+1 on the layout and lead dress.
Even if you were to remove the power section, there is enough gain in the preamp, for squealing to happen on its own. Basically, you have stepped into a minefield.
I strongly recommend taking a look at the designs on AX84.com, which are fully baked, complete with working layouts, and with a whole range of preamp gain choices and power amps.
Sent from my phone. Please excuse any typpos.
I would also add that those cold-biased later stages with the 2k7 and 1k8 cathode resistors are going to suffer from harsh/fizzy distortion. Probably you should swap around the cathode resistor values of the first two stages. Your tone stack really needs to be driven from a cathode follower, and/or you should also have a warm-biased stage, with something like a 620 ohm cathode resistor to give you some good thick crunchy Marshall style distortion.
However, I still recommend checking out the project designs on AX84.com
Sent from my phone. Please excuse any typpos.
However, I still recommend checking out the project designs on AX84.com
Sent from my phone. Please excuse any typpos.
So... Grid stopper at v2 helped, but the problem still occurred, although at higher gain.
Next, I tried adding the "pot stopper" ,10k, which made the amp act really weird. Barely any volume, except at one very specific spot on the gain pot , about 2 O'clock.At which it goes to max volume...
I've replaced the pot with a 1M7 pot, that seems to work for now.
Next, I tried adding the "pot stopper" ,10k, which made the amp act really weird. Barely any volume, except at one very specific spot on the gain pot , about 2 O'clock.At which it goes to max volume...
I've replaced the pot with a 1M7 pot, that seems to work for now.
Your amp is probably oscillating continuously at a high frequency (too high to hear, it's usually at radio frequencies). You may be able to hear or see interference if you take a portable radio or small TV near the amp, which will confirm the amp is unstable and oscillating.
If you're sure the layout is good, RF instability is basically caused by the amp having too much gain at too high a frequency. You have to take measures to reduce the amps gain at frequencies above the guitar range, and that starts with having a grid stopper on every stage. Also, are you using shielded cable on the input? I consider that a necessity, even if some very old amp designs didn't use it.
Beyond grid stoppers on every stage, you might have to add additional capacitors from gain stage anodes to ground. But I suggest starting with the grid stoppers.
-Gnobuddy
I completely agree with having grid stoppers on every gain stage. Standard practice in high gain preamps. Also any longer signal connections should be made with shielded wire (only grounded at one end).
First and second samples clearly sound as HF oscillation, third one is strong buzz which should not be there, all 3 scream poor grounding - poor shielding - poor layout.
Killing gain or filtering highs strongly enough as to make sound muddy attack the symptom and not the main cause.
Adding grid stoppers is fine, because they usually work above the audio spectrum, but you still need to improve layout.
The old fine amps you mention, certainly had tens or hundreds hours in development and tweaking before reaching the commercial form you or I know, with most bugs ironed out.
You either *faithfully* follow an old tried and true layout and grounding or be ready to sweat it 🙂 .
FWIW I started building Guitar amps in 1969 and *all* hummed, buzzed, played radio 🙁 , even when I shielded everything (even filament wires) , tried to use bus or star grounds, etc.
Today I build with plain unshielded wire and have no problems .... but intuitively , say, ground here and not there (an inch away) because I "know" there is a troublesome spot ..... and so on.
As they say, "practice makes perfect".
Killing gain or filtering highs strongly enough as to make sound muddy attack the symptom and not the main cause.
Adding grid stoppers is fine, because they usually work above the audio spectrum, but you still need to improve layout.
The old fine amps you mention, certainly had tens or hundreds hours in development and tweaking before reaching the commercial form you or I know, with most bugs ironed out.
You either *faithfully* follow an old tried and true layout and grounding or be ready to sweat it 🙂 .
FWIW I started building Guitar amps in 1969 and *all* hummed, buzzed, played radio 🙁 , even when I shielded everything (even filament wires) , tried to use bus or star grounds, etc.
Today I build with plain unshielded wire and have no problems .... but intuitively , say, ground here and not there (an inch away) because I "know" there is a troublesome spot ..... and so on.
As they say, "practice makes perfect".
Last edited:
I have owned three Fender guitar amps - a Princeton Reverb, a Superchamp XD (hybrid), and, briefly, a Blues Junior.
In factory stock form, all three were just barely stable:
1) The Princeton will howl if I place a high-gain distortion pedal within a couple of feet of the amp.
2) The Blues Junior went into RF oscillation if the flat cables running from tube sockets to PCB weren't perfectly located.
3) The Superchamp XD is so close to oscillation that it goes bananas if you even replace the stock output transformer with an improved one - Bill M. had to cancel his plan to sell an improved OT for the XD because many of them would go unstable.
Leo was an accountant first and foremost, and if he could save a penny on a grid stopper resistor, he would do it. The company he left behind him seems to often follow the same philosophy.
For guitar, there is nothing useful above 10 kHz, so there is absolutely no good reason to extend the amps frequency response significantly higher than that. All that you get from extra bandwidth is more noise, more interference, and more chance of instability.
12AX7's are relatively easy - with such a high ra and so much Miller capacitance, they don't have that much bandwidth to start with. IMO all pentodes need respect, because they have such small input capacitances. I've used a few little UHF pentodes (6AK5, 6AG5, 6JW8) in preamps - and with those, I took no chances, and added capacitors strategically to control the bandwidth. Of course, I also paid a lot of attention to layout. It seems to have worked, as I had no instability issues I could detect.
-Gnobuddy
In factory stock form, all three were just barely stable:
1) The Princeton will howl if I place a high-gain distortion pedal within a couple of feet of the amp.
2) The Blues Junior went into RF oscillation if the flat cables running from tube sockets to PCB weren't perfectly located.
3) The Superchamp XD is so close to oscillation that it goes bananas if you even replace the stock output transformer with an improved one - Bill M. had to cancel his plan to sell an improved OT for the XD because many of them would go unstable.
Leo was an accountant first and foremost, and if he could save a penny on a grid stopper resistor, he would do it. The company he left behind him seems to often follow the same philosophy.
For guitar, there is nothing useful above 10 kHz, so there is absolutely no good reason to extend the amps frequency response significantly higher than that. All that you get from extra bandwidth is more noise, more interference, and more chance of instability.
12AX7's are relatively easy - with such a high ra and so much Miller capacitance, they don't have that much bandwidth to start with. IMO all pentodes need respect, because they have such small input capacitances. I've used a few little UHF pentodes (6AK5, 6AG5, 6JW8) in preamps - and with those, I took no chances, and added capacitors strategically to control the bandwidth. Of course, I also paid a lot of attention to layout. It seems to have worked, as I had no instability issues I could detect.
-Gnobuddy
- Status
- Not open for further replies.