A screen grid bypass cap is essential in most small-signal pentode circuits. Here, the g4 dropper resistor is usually a huge value (hundreds of kiloohms), and a bypass cap is mandatory.
But, from what I've seen, the bypass cap seems to frequently be omitted in guitar amp output stages. In these, the screen grid dropper resistor has such a low value that it only introduces a tiny bit of negative feedback, and no bypass capacitor is needed or used.
I've seen exceptions, but they seem to be in the minority. The amp I designed and partially described in the "100 buck amp challenge" does actually use a screen grid bypass cap in the output stage. So does the late Fred Nachbaur's "Spunky" amp (schematic attached), which also uses a zener diode to stabilize the screen grid voltage.
-Gnobuddy
Attachments
Many guitar amps exceed the datasheet specifications, sometimes by quite large margins. I think this was driven by cheapness - attempting to get the maximum possible power out of smaller, cheaper valves. Ignorance may also have played a factor, as many early guitar amp manufacturers really didn't understand electronics very well.
Using a dropper resistor to set the screen grid voltage was born of necessity decades ago, when there was no easy or affordable way to supply a nice regulated DC voltage. These days, you could certainly build a solid-state grid voltage regulator, if you wanted to accurately set and hold the screen grid voltage despite variations in screen grid current. Look at Fred Nachbaur's "Spunky" schematic in my previous post - he used a zener diode to set the output valve grid voltage when lightly driven. But if you drive the thing hard enough, screen grid current will rise so much that zener current drops to zero, and the screen grid voltage will drop. (This provides some "sag", or compression, which some guitarists like.)
That will help with the measurement, but do also take the time to make sure you truly "get" the concept of RMS power. With a fixed amplitude, to go from 15 watts to 52 watts would require starting with a waveform with a crazy-low RMS value; it would have to be a weird waveform composed of very narrow, very tall spikes.
Sounds good! Remember to only drive the thing to full power for short periods of time (a few seconds, just long enough to measure). You don't want to cook the screen grid!I'm gonna replace the resistor, take my usual measurements, and I'll let you know how it goes!
-Gnobuddy
no way to regulate the screen voltage? ????
Guitar amp manufacturers were, at least in the early days, very clueless, very cheap. The guitar amp hardly changed from being a modified radio for a long time, even to the point of using that crappy single-cone speaker arrangement, a trend that has survived even to today.
When I went down that (guitar amp building) path, I discovered a lot that was done "wrong" for a good reason, and a lot that could stand improvement.
As confirmed by friends who have been musicians for decades, many "advances" or "discoveries" were completely on accident, much like the 10 Kohm resistor on the cathode of the 12AX7.
Given that "sag" seems to be so desirable, I would hesitate to regulate the screens. I use transistor regulation a lot on hi-fi amps, not so much on guitar amps. One can achieve sag with a dropping resistor in the power supply.
The Conn organs I like to cannibalize and/or modify used VR tubes quite a bit to achieve that very purpose.
Guitar amp manufacturers were, at least in the early days, very clueless, very cheap. The guitar amp hardly changed from being a modified radio for a long time, even to the point of using that crappy single-cone speaker arrangement, a trend that has survived even to today.
When I went down that (guitar amp building) path, I discovered a lot that was done "wrong" for a good reason, and a lot that could stand improvement.
As confirmed by friends who have been musicians for decades, many "advances" or "discoveries" were completely on accident, much like the 10 Kohm resistor on the cathode of the 12AX7.
Given that "sag" seems to be so desirable, I would hesitate to regulate the screens. I use transistor regulation a lot on hi-fi amps, not so much on guitar amps. One can achieve sag with a dropping resistor in the power supply.
Last edited:
Wrong terminology , sorry.
I was actually thinking of a filter cap, 22u or more, to reduce ps ripple getting to the screen.
I know about those voltage regulator tubes, but I did say "no easy or affordable way".
An extra tube means a bigger chassis, additional valve socket, additional heater power, more cost, more weight, and more bulk. I get the same impression, but they had to sell these things to relatively penniless musicians, after all...
This is something I am quite curious about. Elsewhere on DIY Audio I made a post or two discussing the possibility of using a fairly sensitive midrange speaker as a "tweeter" for a guitar amp, possibly including a lowpass filter to roll off frequencies above 5 or 6 kHz.
I already tried using an 8" dual-cone speaker intended for ceiling-mounted public address/paging purposes in a little low-power guitar amp. The resulting midrange and treble response was quite good-sounding, but that particular speaker had a very stiff suspension and a very light cone, so bass response was extremely lacking.
I would love to hear more about what you found "that could stand improvement". I am still on the quest to make a better guitar amp.
I'm not familiar with this, are you talking about extreme cold-biasing, or something else?
For many players, especially blues guitarists, I agree entirely.
Fred Nachbaur was a classically trained musician, and he designed that amp for his wife. It would seem he and she both preferred squeaky-clean guitar tones, and exact reproduction of the guitar's own dynamics.
So it makes sense that he chose to regulate the screens in his design. (He also designed an optional compressor add-on for those who wanted properly engineered "sag"!)
-Gnobuddy
additional heater power?
for a VR tube? I guess I been using mine wrong,I know about those voltage regulator tubes, but I did say "no easy or affordable way".
-Gnobuddy
"I'm not familiar with this, are you talking about extreme cold-biasing, or something else? "
On a Marshall amp, I think the 45, not positive, there was a 12ax7 with 10K on the cathode. This caused the tube to distort readily when driven moderately hard. In researching this, the consensus was that a mistake had been made and that 1K was really what was intended, but somebody screwed up and changed the 1 to a 10 on the schematic and the amp was built accordingly.
Along comes Ken Fischer of Trainwreck fame (I can hear the insults coming already for daring to 'diss the massah). He implements the same 10K resistor on the final gain stage 12ax7 to get that "signature (i can hardly type that with a straight face)" Trainwreck sound.
So, what many would say was CLEARLY a mistake impacted how many guitar amps would be designed and built.
While I'm on the subject of the 12AX7, I don't think I EVER used one. from the very beginning, I found a tube that had the same gain and sounded better and passed more current (thus more transconductance). The 6N2P. I was an early adopter of Russian tubes. I initially tried them because they were cheap. I discovered how good many of them are and, while everyone else was turning up their nose at the supposedly inferior Russian tubes, I was enjoying the sound quality and the savings. NOW, of course, the tune has changed dramatically. The 6P3S-E is very highly regarded as one of the best 6L6GC tubes there are. Several Russian preamp and power tubes are highly sought out now and prices have gone up as a result.
Guitar amp builders have attempted repeatedly to persuade me to "step back from the audiophile precipice" as I took what I learned from building good sounding amps for listening to music and applied some of that to building guitar amps. The first amp I built and sold to a professional guitar player called it the best sounding amp he has ever had. Not saying this to brag but to point out that NOT being close-minded and trying new things and IGNORING what "EVERYBODY knows goes a long way.
So, yes, I found quite a few things that need improvement in guitar amps and had I jumped right in building guitar amps first without every building any for listening to music, I would have made the same mistakes the rest of the crowd is making. That is not to say that guitar amps don't have their design differences and their design challenges, They certainly do. But continuing down the same path of creating trashy sound and calling it good isn't progress. Introducing SOME distortion, okay, got it. Introducing ANY and all types of distortion? IMHO, not a particularly bright idea. and that is what I see implemented. I don't listen to metal or CRAP (RAP) or people trying to be as loud as possible. SO, MY idea of a good amp IS influenced by the type of music I listen to. I won't try to build an amp for metal or any type of music I don't find enjoyable.
On a Marshall amp, I think the 45, not positive, there was a 12ax7 with 10K on the cathode. This caused the tube to distort readily when driven moderately hard. In researching this, the consensus was that a mistake had been made and that 1K was really what was intended, but somebody screwed up and changed the 1 to a 10 on the schematic and the amp was built accordingly.
Along comes Ken Fischer of Trainwreck fame (I can hear the insults coming already for daring to 'diss the massah). He implements the same 10K resistor on the final gain stage 12ax7 to get that "signature (i can hardly type that with a straight face)" Trainwreck sound.
So, what many would say was CLEARLY a mistake impacted how many guitar amps would be designed and built.
While I'm on the subject of the 12AX7, I don't think I EVER used one. from the very beginning, I found a tube that had the same gain and sounded better and passed more current (thus more transconductance). The 6N2P. I was an early adopter of Russian tubes. I initially tried them because they were cheap. I discovered how good many of them are and, while everyone else was turning up their nose at the supposedly inferior Russian tubes, I was enjoying the sound quality and the savings. NOW, of course, the tune has changed dramatically. The 6P3S-E is very highly regarded as one of the best 6L6GC tubes there are. Several Russian preamp and power tubes are highly sought out now and prices have gone up as a result.
Guitar amp builders have attempted repeatedly to persuade me to "step back from the audiophile precipice" as I took what I learned from building good sounding amps for listening to music and applied some of that to building guitar amps. The first amp I built and sold to a professional guitar player called it the best sounding amp he has ever had. Not saying this to brag but to point out that NOT being close-minded and trying new things and IGNORING what "EVERYBODY knows goes a long way.
So, yes, I found quite a few things that need improvement in guitar amps and had I jumped right in building guitar amps first without every building any for listening to music, I would have made the same mistakes the rest of the crowd is making. That is not to say that guitar amps don't have their design differences and their design challenges, They certainly do. But continuing down the same path of creating trashy sound and calling it good isn't progress. Introducing SOME distortion, okay, got it. Introducing ANY and all types of distortion? IMHO, not a particularly bright idea. and that is what I see implemented. I don't listen to metal or CRAP (RAP) or people trying to be as loud as possible. SO, MY idea of a good amp IS influenced by the type of music I listen to. I won't try to build an amp for metal or any type of music I don't find enjoyable.
speakers - right track
You are on the right track there. As a general rule, the more efficient the speaker driver, the more screwed up it's response is. If your goal is trashy sound, fine, don't worry about it. If you want it to sound GOOD, then you should care. Guitar amps tend to use extremely efficient drivers and the response graph looks like the Rocky Mountains. As I studied speakers from the audiophile point of view, I noticed that ALL of them, from the 15-inch to the 6-1/2-inch driver, the response started to go to hell north of 1000 Hz.
I once had some German paper cone drivers, about 7 inches. I listened to them without a tweeter, attempting to use them full range. it sounded HORRIBLE. really nasty. Struth would love them. Added a dome tweeter with a 2.5 KHz crossover or thereabouts, and it sounded normal.
I had often played a guitar through the home stereo system since I'm an amateur and poor guitar player even by amateur standards. Sound quality was good. good "sparkle."
Since those upper harmonics are undoubtedly adding "sparkle" and everybody loves sparkle, why would you choke it off?
I have in all my studying NEVER EVER EVER heard of a guitarist contemplating getting a quality driver for his rig. I went to the manufacturers of guitar speaker drivers and looked at the specs. very efficient. HORRIBLE response. we have plenty of amp power and PA power, so why is this still being done? TRADITION. the same reason guitar amps have been so very slow to evolve beyond that 1930s table top radio that got modified for musical instrument use.
Likewise, very very very very few guitarists ever attempted to use a speaker system that would encompass SMOOTHLY the entire frequency response of the guitar.
If you study psychoacoustics, you will see GOOD reason to capture the entire frequency response of the instrument in question. The research is solid. it's not a new field. I have found psychoacoustics research from back in the 30s .
It's important to remember that the foundation is the FUNDAMENTAL tones. that needs to be as right as you can make it. All that harmonic goodness means nothing if the FUNDAMENTAL is flawed. What that means is your bass response HAS to be solid. Using too-small speakers without good bass response is the surest way to guarantee mediocrity at best, utter failure at worst,
You are on the right track there. As a general rule, the more efficient the speaker driver, the more screwed up it's response is. If your goal is trashy sound, fine, don't worry about it. If you want it to sound GOOD, then you should care. Guitar amps tend to use extremely efficient drivers and the response graph looks like the Rocky Mountains. As I studied speakers from the audiophile point of view, I noticed that ALL of them, from the 15-inch to the 6-1/2-inch driver, the response started to go to hell north of 1000 Hz.
I once had some German paper cone drivers, about 7 inches. I listened to them without a tweeter, attempting to use them full range. it sounded HORRIBLE. really nasty. Struth would love them. Added a dome tweeter with a 2.5 KHz crossover or thereabouts, and it sounded normal.
I had often played a guitar through the home stereo system since I'm an amateur and poor guitar player even by amateur standards. Sound quality was good. good "sparkle."
Since those upper harmonics are undoubtedly adding "sparkle" and everybody loves sparkle, why would you choke it off?
I have in all my studying NEVER EVER EVER heard of a guitarist contemplating getting a quality driver for his rig. I went to the manufacturers of guitar speaker drivers and looked at the specs. very efficient. HORRIBLE response. we have plenty of amp power and PA power, so why is this still being done? TRADITION. the same reason guitar amps have been so very slow to evolve beyond that 1930s table top radio that got modified for musical instrument use.
Likewise, very very very very few guitarists ever attempted to use a speaker system that would encompass SMOOTHLY the entire frequency response of the guitar.
If you study psychoacoustics, you will see GOOD reason to capture the entire frequency response of the instrument in question. The research is solid. it's not a new field. I have found psychoacoustics research from back in the 30s .
It's important to remember that the foundation is the FUNDAMENTAL tones. that needs to be as right as you can make it. All that harmonic goodness means nothing if the FUNDAMENTAL is flawed. What that means is your bass response HAS to be solid. Using too-small speakers without good bass response is the surest way to guarantee mediocrity at best, utter failure at worst,
Last edited:
Oh? Your VR tubes are zero size, zero weight, and zero cost?
Yeah, mentioning heaters was my mistake. It's been a long hard week already, and my brain is fried.
-Gnobuddy
couldn't resist
had to screw with you a little for that one
Oh? Your VR tubes are zero size, zero weight, and zero cost?
Yeah, mentioning heaters was my mistake. It's been a long hard week already, and my brain is fried.
-Gnobuddy
had to screw with you a little for that one
You and I have very different tastes in guitar sound. I have done the same experiment, and experienced very different results: for me, a good flat-response amp and flat, accurate, hi-fi speaker is good-sounding for electro-acoustic guitar, but very bad-sounding for true electric guitar.
Les Paul (the man, not the guitar) was in your camp, though. He preferred squeaky-clean tone, and even went to the trouble of winding his own low-inductance guitar pickups to make sure no "sparkle" was lost.
Me, I love Les Paul's guitar chops, but hate his guitar tone!
I wouldn't, for an acoustic guitar. But for something like a Les Paul, the accurately reproduced sound of the instrument is not very different from the long "ping" of a recently struck tuning-fork. Bland and uninteresting, with far too few harmonics to be a good musical instrument.
If you disagree, just compare the sound of an accurately reproduced electric guitar with the rich timbre of a cello, or saxophone, or viola, or any traditional classical instrument. The electric guitar comes off a very poor second-best, tonally.
So to fix the boring timbre, we introduce deliberate non-linear distortion into the guitar amp to add harmonics, thus improving the timbre. And then we have to filter out the "sparkle" in order to also control the harshness of the high-frequency harmonics, and high-frequency intermodulation products, and now we're now well on our way to the traditional "dirty" guitar amp design...
I'm not arguing with you, of course you're entitled to your own guitar tone preferences. But in this case, our individual experiments won't help each other; we're trying to go in two quite different directions, sonically speaking.
I tried it, and I didn't like the result for electric guitar. What I would like to do is control the (spatial) dispersion of high frequencies better, by using a smaller driver for the high frequencies. Traditional big guitar speakers "beam" treble very badly.
I think you and I will have different opinions once again - I personally loathe the sound of a solid-body electric guitar played through a clean, powerful solid-state amp.
So for me, for electric guitar, it has to be valves, and now "plenty of power" is much harder to come by (costlier, heavier, bulkier, etc). To manage all these things, ten to fifteen watts RMS becomes the practical upper limit. I have an old back injury that's begun to act up this year, and I have no desire to carry any 50-lb guitar amps around!
I do agree with you that we don't need 100 dB @1W speaker sensitivity, even with only 10W - 15W from the amp. There is really no need for absurdly high, ear-damaging SPL levels.
If you want to accurately reproduce the sound of a piano or violin or what-have-you, I agree absolutely.
If I want to create a good electric guitar tone, which means I have to control the unwanted guitar pick noises, fretting noises, and harshness caused by (intentional) nonlinearity in the amp, then I definitely do want to filter out the harsh-sounding upper frequencies...
Y'know, recording engineers frequently high-pass electric guitar before mixing. The low fundamental guitar notes often clash with similar frequencies from the kick drum and bass guitar, so everything "sits better in the mix" if the deep guitar fundamentals are filtered out. Sometimes they filter out everything below 200 or even 300 Hz...you can hear this in lots of commercially successful guitar music from the last two or three decades.
For solo (not in a mix) electric guitar, I prefer to have flat response down to 80 Hz optionally available, but then have electronic means to attenuate the low frequency response if necessary. (Bass cut, either switchable, or via a bass control knob.)
For solo (not in a mix) electro-acoustic guitar, I do want flat response down to 80 Hz, so my Dreadnaught guitars don't sound like little tin cans.
I also play a little bass guitar, and even there, flat response down to the lowest fundamental (around 30 Hz for my 5-string basses with a low B) usually sounds bland, boomy, and hard to hear in the mix. A bass guitar amp that rolls off some of the lowest frequencies produces a more interesting and workable bass guitar tone.
Incidentally, I have almost a life-long history of technical involvement with Hi-Fi, going back to my childhood. It's not that I don't know what's needed for accurate sound reproduction; unfortunately, for me, that Hi-Fi knowledge hasn't, for the most part, translated to improved guitar amp sounds, which don't involve reproducing a good sound, but rather, manufacturing a good sound from an inherently bland-sounding intrument.
With one exception: I find I prefer guitar cabs with sound damping added to the inside. That bit of Hi-Fi speaker design practice does work for me, with electric guitars.
-Gnobuddy
THIS is why I don't bother
at risk of sounding petulant, (I'm not) this is why I don't bother. You are in the same boat as everyone else. you have a suitcase full of preconceived notions and you refuse to throw out the old clothes. the worn out ones. It's like talking to a Russian coworker. He seems to be hearing me, but he doesn't hear a word I say.
I never said i use solid state. I mentioned the home stereo ONLY to point out the improved clarity CLARITY and ONLY the improved clarity one gets from allowing the upper harmonics through. In fact, I did nothing to squelch high frequency response in any of the guitar amps I built and the problems you claim by not doing so simply never materialized.
your own experiment conducted with the 8-inch speakers confirmed the point I was trying to make. While you are friendly about disagreeing with almost everything I say, you disagree nonetheless.
the "boring timbre" comes about by the very distortions you allow. I make mine distort in a controllable fashion the same way the aforementioned Marshall and Trainwreck distort, I simply did it in a more QUALITY-conscious (there, I didn't say hi-fi) fashion, and achieved better results. I allow the distortions that make the sound more "interesting," I minimize those distortions that do not enhance the listening experience. It isn't that we have different "tastes," it's that you don't understand how the different distortions sound and why some should be discouraged. After making several guitar amps without any attempt at choking off the frequency response, i can confidently say there's something wrong with your electronics if yours sound like hell.
in the TYPICAL guitar amp, The bass already gets choked off good by the coupling caps. you have to do that to get the strings to distort equally. And the net result is a weak thready sound. There is a reason I do it differently it sounds better. I have heard the difference you have not.
your bass rig probably suffers from poor design and/or a lot of IM distortion. Again, poor attention to details. I think I couldt write volumes here and you just wouldn't get it. Do you understand WHY a tube amp sounds better than solid state? that fundamental difference is why I chose to accentuate that richness and lushness of sound, the elusive "fat" tone everyone says they want, but then implement a design that GUARANTEES they won't get it.
Again, this is why I don't bother going to great extremes to share what I learned. the vast majority of the people in the world are the reason for that famous zen Buddhist saying, when the zen master keeps filling the cup until the student cries out for him to stop. The Zen Buddhist then makes that remark about emptying the cup first. you are so full of your opinions, you MIS-hear at least half of what I have to say,
The speaker: to prove my point which should be obvious, a guitarist posted how the combination of TWO typical guitar amp drivers sounded so much better than either alone. I looked at their graphs. The one was filling in the holes made by the other. their SUM was much closer to that of a QUALITY flat response driver. Just my opinion, but claiming the crappy poor quality guitar speaker is "part of the sound," indicates a level of insanity rather commonly encountered on the internet. One guy starts the rumor that it's "part of the sound of a guitar amp," implying it's GOOD, and EVERYONE takes up the mantra. conclusions reached with really no evidence at all.
Chopping off the bass response to solve what is probably a resonance peak of some sort is like cutting off the foot for an ingrown toenail. boomy muddy bass means there's a problem that needs to be fixed and the fix isn't EQ. or it shouldn't be
at risk of sounding petulant, (I'm not) this is why I don't bother. You are in the same boat as everyone else. you have a suitcase full of preconceived notions and you refuse to throw out the old clothes. the worn out ones. It's like talking to a Russian coworker. He seems to be hearing me, but he doesn't hear a word I say.
I never said i use solid state. I mentioned the home stereo ONLY to point out the improved clarity CLARITY and ONLY the improved clarity one gets from allowing the upper harmonics through. In fact, I did nothing to squelch high frequency response in any of the guitar amps I built and the problems you claim by not doing so simply never materialized.
your own experiment conducted with the 8-inch speakers confirmed the point I was trying to make. While you are friendly about disagreeing with almost everything I say, you disagree nonetheless.
the "boring timbre" comes about by the very distortions you allow. I make mine distort in a controllable fashion the same way the aforementioned Marshall and Trainwreck distort, I simply did it in a more QUALITY-conscious (there, I didn't say hi-fi) fashion, and achieved better results. I allow the distortions that make the sound more "interesting," I minimize those distortions that do not enhance the listening experience. It isn't that we have different "tastes," it's that you don't understand how the different distortions sound and why some should be discouraged. After making several guitar amps without any attempt at choking off the frequency response, i can confidently say there's something wrong with your electronics if yours sound like hell.
in the TYPICAL guitar amp, The bass already gets choked off good by the coupling caps. you have to do that to get the strings to distort equally. And the net result is a weak thready sound. There is a reason I do it differently it sounds better. I have heard the difference you have not.
your bass rig probably suffers from poor design and/or a lot of IM distortion. Again, poor attention to details. I think I couldt write volumes here and you just wouldn't get it. Do you understand WHY a tube amp sounds better than solid state? that fundamental difference is why I chose to accentuate that richness and lushness of sound, the elusive "fat" tone everyone says they want, but then implement a design that GUARANTEES they won't get it.
Again, this is why I don't bother going to great extremes to share what I learned. the vast majority of the people in the world are the reason for that famous zen Buddhist saying, when the zen master keeps filling the cup until the student cries out for him to stop. The Zen Buddhist then makes that remark about emptying the cup first. you are so full of your opinions, you MIS-hear at least half of what I have to say,
The speaker: to prove my point which should be obvious, a guitarist posted how the combination of TWO typical guitar amp drivers sounded so much better than either alone. I looked at their graphs. The one was filling in the holes made by the other. their SUM was much closer to that of a QUALITY flat response driver. Just my opinion, but claiming the crappy poor quality guitar speaker is "part of the sound," indicates a level of insanity rather commonly encountered on the internet. One guy starts the rumor that it's "part of the sound of a guitar amp," implying it's GOOD, and EVERYONE takes up the mantra. conclusions reached with really no evidence at all.
Chopping off the bass response to solve what is probably a resonance peak of some sort is like cutting off the foot for an ingrown toenail. boomy muddy bass means there's a problem that needs to be fixed and the fix isn't EQ. or it shouldn't be
Last edited:
Some Results- and more confusion
So... on a completely different topic than what has been recently mentioned... Got some results.
I installed the smaller screen resistor in my amplifier (500 ohm) and It raised the power substantially. I then experimented with the bypass cap on the screen. With the same input, I got output powers of
With Bypass cap - 5.4W
Without - 5.1W
Either way, they pass my requirements, so thats awesome. I assume adding the bypass cap gets me a little more power because B+ is bouncing around a bit as the current drawn by the tube varies, and therefore the screen voltage does too. As it drops, the power output does too, slightly.
Because of the feedback, however, the Harmonic distortion is lower without the bypass cap.
I then came across another strange issue: The oscillations returned with high preamp gain. I am reposting my schematic because it will be easy to reference places on it.
I used the scope to start looking for where the oscillation comes from. I started where the "1" is marked on the schematic, and it turned out that the oscillations were detectable there. The frequency is 3.3 Hz, and it is always that, hardly any variation. I also noticed that the oscillations need to be kicked off with some input, such as a strum on the guitar. If the gain pot is turned down enough, the sinusoid gets damped out in a cycle or two, but with the pot turned up, the oscillation is not damped out.
Now, to the reason the oscillations stopped before. It turns out the node marked "2" in the tone circuit was shorted to ground. So I shorted it to ground again with a clip. No oscillations.
You can take the clip off, set off the oscillations, and then put the clip back on, and the oscillations immediately stop. Considering the tone stack follows the second valve, I find it strange that the oscillations are apparent back at point "1".
Now obviously the overall preamp gain is considerably less with node 2 shorted, as the load is significantly lower impedance. But does the load have any effect on capacitance on the grid? I'm trying to figure out why node 1 oscillates. And it isn't that the oscillations are there all the time, but filtered out when node 2 is grounded. They disappear on node 1 when node 2 is grounded.
Do you think I am being bitten by feedback and variations on the B+2? I have a single bypass capacitor for all the preamp tubes. It seems like a controversial question, so I am almost afraid to ask, but do I need to make v1.2 into a cathode follower? Maybe change the decoupling capacitor to reject 3.3Hz?
Honestly, I like the way it sounds with node 2 grounded. The amp has a good tone and is great to play, maybe slightly heavy on the highs, good level of distortion with the preamp cranked, clean when you back off, but I said I would build a tone control, so I would like to get this working, or it won't look good on me.
Thanks,
-Kevin
So... on a completely different topic than what has been recently mentioned... Got some results.
I installed the smaller screen resistor in my amplifier (500 ohm) and It raised the power substantially. I then experimented with the bypass cap on the screen. With the same input, I got output powers of
With Bypass cap - 5.4W
Without - 5.1W
Either way, they pass my requirements, so thats awesome. I assume adding the bypass cap gets me a little more power because B+ is bouncing around a bit as the current drawn by the tube varies, and therefore the screen voltage does too. As it drops, the power output does too, slightly.
Because of the feedback, however, the Harmonic distortion is lower without the bypass cap.
I then came across another strange issue: The oscillations returned with high preamp gain. I am reposting my schematic because it will be easy to reference places on it.
I used the scope to start looking for where the oscillation comes from. I started where the "1" is marked on the schematic, and it turned out that the oscillations were detectable there. The frequency is 3.3 Hz, and it is always that, hardly any variation. I also noticed that the oscillations need to be kicked off with some input, such as a strum on the guitar. If the gain pot is turned down enough, the sinusoid gets damped out in a cycle or two, but with the pot turned up, the oscillation is not damped out.
Now, to the reason the oscillations stopped before. It turns out the node marked "2" in the tone circuit was shorted to ground. So I shorted it to ground again with a clip. No oscillations.
You can take the clip off, set off the oscillations, and then put the clip back on, and the oscillations immediately stop. Considering the tone stack follows the second valve, I find it strange that the oscillations are apparent back at point "1".
Now obviously the overall preamp gain is considerably less with node 2 shorted, as the load is significantly lower impedance. But does the load have any effect on capacitance on the grid? I'm trying to figure out why node 1 oscillates. And it isn't that the oscillations are there all the time, but filtered out when node 2 is grounded. They disappear on node 1 when node 2 is grounded.
Do you think I am being bitten by feedback and variations on the B+2? I have a single bypass capacitor for all the preamp tubes. It seems like a controversial question, so I am almost afraid to ask, but do I need to make v1.2 into a cathode follower? Maybe change the decoupling capacitor to reject 3.3Hz?
Honestly, I like the way it sounds with node 2 grounded. The amp has a good tone and is great to play, maybe slightly heavy on the highs, good level of distortion with the preamp cranked, clean when you back off, but I said I would build a tone control, so I would like to get this working, or it won't look good on me.
Thanks,
-Kevin
Also, clearly I can approximate the accidental short by turning P5 and P6 so they are shorted. Starting with them shorted, and applying a muted strum input for an "impulse". With node 2 grounded, I see on the scope, essentially and impulse. As I start to raise the pots, The time constant for the oscillation increases.
Reading up on this and thinking a little more, I am starting to think it is feedback through the B+. 2 amplifier stages for about 360 degrees of phase shift, feeding back, definitely would cause oscillation. And the level of the feedback is likely not high enough when the gain of the second stage is reduced to cause a problem.
Does this answer sound about right? I am going to try adding another dropping resistor to the B+ and another cap to filter the feedback out and kill the problem.
Reading up on this and thinking a little more, I am starting to think it is feedback through the B+. 2 amplifier stages for about 360 degrees of phase shift, feeding back, definitely would cause oscillation. And the level of the feedback is likely not high enough when the gain of the second stage is reduced to cause a problem.
Does this answer sound about right? I am going to try adding another dropping resistor to the B+ and another cap to filter the feedback out and kill the problem.
Jerry - I am truly sorry that you feel the way you do.
Most people follow the mainstream; a few take the lonely path to a life of curious tinkering. Unfortunately, the next step down that particular path takes one to downtown MartyrVille, where many brilliant and curious people end up, feeling persecuted, unheard, and unappreciated. Once in MartyrVille, it's quite easy to eventually come to the conclusion that everyone else in the world is not only just plain wrong, but also quite stupid.
I wish we were able to communicate; but you are not open to the possibility that other opinions than your own may in fact be equally valid.
I am always glad to hear of other people who experiment rather than take for granted; so I am glad to have crossed paths with you, even though we are unable to communicate.
-Gnobuddy
Yay for you, and I hope you get bonus marks for having fixed your partner's half of your project!
Exactly right. And yet, there is a widely-held (wrong) belief on this very forum that regulating the screen grid voltage will result in lower distortion.
Something is really strange here, there shouldn't be enough loop gain at 3.3 Hz to cause sustained oscillation there.
It's been a long week, and my brain is fried, but if I can think of any useful hypothesis to test, I'll post them here.
-Gnobuddy
Whose consensus was it?
Then consensus of the blind arguing sunlight?
Physics is not run by voting.
That poorly named "cold biased tube" does a lot of very important things and is essential for high gain Marshall sound, from the early "Master Volume" to marshall derivatives (Soldano/Mesa Boogie/ tons more) ; the typical noob mistake is to think that he found an error nobody else noticed and replace that 10k resistor with something between 820 and 2700 ohms, "what other gain stages use" and or bypassing it with a cap "to get extra gain".
Amp loses bite, cut, the typical Marshall snarl and becomes muddy/fuzzy and compressed.
Don´t compare any amp *designed* to heavily distort , often by 20 to 40 dB , to any clean amp that accidentally got distorted by a couple dB .
Oh well.
You want clean flat sound and flat, extended response speakers?
Fine , get a direct box and plug straight into the PA mixer.
are you for real?
the only emotion is feel is frustration in trying to communicate with people who are so busy responding to their own internal stimuli that nothing else gets through.
Again, the issue is you are so full of yourself that real communication is impossible.
you DID ask me where I thought they could stand improvement. you even acknowledged I was correct while contradicting me. YOU ARE INSANE.
your condescending BS is, (not surprisingly) way off the mark. the issue is exactly what I said it is. REAL communication doesn't take place because you think you have all the answers. go look at your responses. not just to me, your responses in general. you have this need to contradict everybody. you have insecurities.
the only emotion is feel is frustration in trying to communicate with people who are so busy responding to their own internal stimuli that nothing else gets through.
Again, the issue is you are so full of yourself that real communication is impossible.
you DID ask me where I thought they could stand improvement. you even acknowledged I was correct while contradicting me. YOU ARE INSANE.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.