Looking for opinions about how this is done.
Since I don't have access to my bench this week and can't troubleshoot my amp yet I have been spending my time thinking about possible second projects. Long story short - my search lead me to this awesome thing.
https://www.youtube.com/watch?v=SGa7UNAQM3k
Seems like the coolest idea ever. Sounds great too.
I am curious about the method used to dial the individual tubes in and out. I found one bit of speculation that the potentiometers are being used to dial in different resistance values for the cathode resistors. I think that this would be awesome and I'm more willing to purchase the parts this time around so I'd be happy to get the "right" OT and PT if I can figure that out.
This seems like it could be the coolest project. Is there perhaps a similar design from another manufacturer that I could look at or a schematic for this amp itself?
Since I don't have access to my bench this week and can't troubleshoot my amp yet I have been spending my time thinking about possible second projects. Long story short - my search lead me to this awesome thing.
https://www.youtube.com/watch?v=SGa7UNAQM3k
Seems like the coolest idea ever. Sounds great too.
I am curious about the method used to dial the individual tubes in and out. I found one bit of speculation that the potentiometers are being used to dial in different resistance values for the cathode resistors. I think that this would be awesome and I'm more willing to purchase the parts this time around so I'd be happy to get the "right" OT and PT if I can figure that out.
This seems like it could be the coolest project. Is there perhaps a similar design from another manufacturer that I could look at or a schematic for this amp itself?
I don't think you could do it with the cathode resistors, as a single-ended amp needs to be centre-biased to work properly. As you increased the cathode resistance to reduce the output of one valve you would get a horrible 'half-wave rectified' kind of distortion. Also, the cathode resistors have to dissipate a bit of power, so pots there would need to be heavy duty types.
Another way to do it, would be to have conventional master volume pots in front of each of the three output stages, which are wired in parallel. The output transformer needs to be rated to cope with the DC currents of all three output valves added together and its primary impedance would have to be a compromise to more or less match all three different output valves.
(Just some initial thoughts, other members may have different suggestions.)
Another way to do it, would be to have conventional master volume pots in front of each of the three output stages, which are wired in parallel. The output transformer needs to be rated to cope with the DC currents of all three output valves added together and its primary impedance would have to be a compromise to more or less match all three different output valves.
(Just some initial thoughts, other members may have different suggestions.)
I'm glad you said that because the cathode resistor volume control idea didn't feel good to me.
I have been digging for more info and there is one account that says that the OT has 3 separate primaries. Not sure if that's the truth or simply uninformed speculation again.
I was wondering if i would be able to find an OT that would work in that event, and it made me wonder if that would be functionally equivalent to wiring up 3 separate OTs, one for each valve, and paralleling the secondary of all of them all to the same speaker.
I have been digging for more info and there is one account that says that the OT has 3 separate primaries. Not sure if that's the truth or simply uninformed speculation again.
I was wondering if i would be able to find an OT that would work in that event, and it made me wonder if that would be functionally equivalent to wiring up 3 separate OTs, one for each valve, and paralleling the secondary of all of them all to the same speaker.
An OT with three separate primaries would be a very unusual custom design. One primary with two extra taps would also need to be a custom build (I expect).
I believe three separate OTs with paralleled outputs would work, just about. If two of the pentodes were idle (their master volumes set to zero) they would have a voltage imposed on their anodes, reflected back across their OT from the speaker voltage, but I don't think that is a problem.
In the video they seem to have a single OT, so my own feeling is that they either have a primary with extra taps to suit the three different valves, or they have juggled around with load lines to find an impedance that is reasonable for all three.
An EL84 usually expects a higher load impedance than a 6L6, but to keep the valves happy it would be best to compromise on a lower impedance rather than a high one. Then, do we also have to consider how well balanced the maximum power from each of the three valves is?
I believe three separate OTs with paralleled outputs would work, just about. If two of the pentodes were idle (their master volumes set to zero) they would have a voltage imposed on their anodes, reflected back across their OT from the speaker voltage, but I don't think that is a problem.
In the video they seem to have a single OT, so my own feeling is that they either have a primary with extra taps to suit the three different valves, or they have juggled around with load lines to find an impedance that is reasonable for all three.
An EL84 usually expects a higher load impedance than a 6L6, but to keep the valves happy it would be best to compromise on a lower impedance rather than a high one. Then, do we also have to consider how well balanced the maximum power from each of the three valves is?
I don't think there is anything special on the OPT (unless someone knows otherwise for sure), imo it just has a single primary winding and three taps on the secondary. Each power tube has its own level control to allow the player to dial in the mix that he/she wants.
Adding details as i find them:
The output of the amp according to Palmer is 5W from the el84, 5W from 6V6 and 8W from 6l6. I guess if you turn all three up it's supposedly about 15W total though. I don't know why 18 = 15.
I also found out that you can use an EL34 in the place of the 6L6, and it also lists kt90 and kt77 as permissable tubes. I'm not familiar with those tubes yet.
I found a project on this forum from a few years ago that paralleled a 6v6 and an el84. The discussion was interesting and the one member who completed his design seemed very happy with it. He claimed that he got best results by getting rid of the bypass caps on the cathode resistors (created oscillations he said) and he didn't do anything special with the plates - he just wired them in parallel and connected them to the OT primary.
The output of the amp according to Palmer is 5W from the el84, 5W from 6V6 and 8W from 6l6. I guess if you turn all three up it's supposedly about 15W total though. I don't know why 18 = 15.
I also found out that you can use an EL34 in the place of the 6L6, and it also lists kt90 and kt77 as permissable tubes. I'm not familiar with those tubes yet.
I found a project on this forum from a few years ago that paralleled a 6v6 and an el84. The discussion was interesting and the one member who completed his design seemed very happy with it. He claimed that he got best results by getting rid of the bypass caps on the cathode resistors (created oscillations he said) and he didn't do anything special with the plates - he just wired them in parallel and connected them to the OT primary.
That's interesting. Removing the bypass caps reduces the gain of the output stages, which would tend to reduce the possibility of oscillation. I wonder if they tried any other ways to stop oscillation (adequate grid stoppers, screen stoppers and maybe even small build-out resistors on the anodes)? Removing the bypass caps also changes the character of the output stage overdrive, as one source of bias shift, due to voltage across the bypass cap increasing during continuous overdrive, is eliminated.
http://www.diyaudio.com/forums/instruments-amps/164563-el84-6v6-parallel-7.html
I found the thread again. I linked to page 7 because that's where the schematic for the output stage was posted (by Cassiel)
I found the thread again. I linked to page 7 because that's where the schematic for the output stage was posted (by Cassiel)
Good thread, especially some of the early posts, suggests it will work fine with one shared primary.
I don't trust an amp "designer" that has a tube rectifier in a class A amp and says it's there for "sag".
Secondly he states that they went with a PCB design because the circuit is so simple. Usually simple circuits are the easiest to implement in point to point wiring. BTW I am not a a person that hates PCB's but I don't think these people know what they are talking about.
Not a huge fan of the amp.
Secondly he states that they went with a PCB design because the circuit is so simple. Usually simple circuits are the easiest to implement in point to point wiring. BTW I am not a a person that hates PCB's but I don't think these people know what they are talking about.
Not a huge fan of the amp.
I wasn't really commenting on the design itself, not having read everything in the thread in detail, just that some of the posts in the thread by tubelab, cerrem and others were informative.
As you say, there is no sag in a class A amp!
As you say, there is no sag in a class A amp!
I was referring to the youtube video in original post and not the linked thread. I do plan on perusing the linked thread.
Oh, I see what you mean.🙂
I forgot the bit about sag in the video.
It did sound quite good though.
I forgot the bit about sag in the video.
It did sound quite good though.
Not bad. Pepsi or Coke really. IMHO you can never get a good feel for amps by watching youtube videos. For me I have to play them with my guitars and usually in a band scenario before I am sold.
Yeah i don't know much about Palmer at all. They could be full of it. I do like what i've heard from that amp though. It's true that you can't trust the sound you're hearing after recording, compressing, and eventually coming out through the crappy cell-phone speaker.
All the same there's still a relative maximum quality that you can hear if you listen to enough amps. The drei sounds good to me. In contrast - no matter how many fender frontman vids i listen to there's never going to be a good-sounding demo of that.
I don't think drei is realistic for me as a build. I would rather attempt zwei if i attempt anything at all. I can see that there's a schematic already worked out for the 6v6/el84 combo. I could try 6L6/el34 instead perhaps.
It seems that these single-ended guitar amps are pretty friendly about these valves in general, and a 4k to 8 ohm OT that can handle the power may be all i would need.
All the same there's still a relative maximum quality that you can hear if you listen to enough amps. The drei sounds good to me. In contrast - no matter how many fender frontman vids i listen to there's never going to be a good-sounding demo of that.
I don't think drei is realistic for me as a build. I would rather attempt zwei if i attempt anything at all. I can see that there's a schematic already worked out for the 6v6/el84 combo. I could try 6L6/el34 instead perhaps.
It seems that these single-ended guitar amps are pretty friendly about these valves in general, and a 4k to 8 ohm OT that can handle the power may be all i would need.
This is technically true, BUT when you drive that Fender Champ with a pedal board having 20 db of gain, the amp is nowhere near class A any more! Tube and solid state rectifiers do sound different in a "class A" guitar amp, especially when overdriven. I built at least a dozen "Turbo Champs" about 15 years ago. Each was different in it's own way, but all contained a tube/silicon rectifier switch. If you are trying to maximize sag, use a 5Y3 tube, but if all 3 tubes are biased up correctly all the time, a 5AR4 is probably the only choice.
I reread the old thread and realized that it was what sparked a lot of multi tube experiments.
I figured out that you can't adjust the tube's level with cathode resistors. The only thing that works is a conventional volume pot (or some variation thereof) for each tube.
You can simply wire all 3 plates in parallel with the correct OPT. I found that about 3.6K (a 300B OPT) was optimum for 3 tubes. You can gain a bit of power by using a higher impedance OPT (5K) for the EL84 and 6V6 while connecting the 6L6 (or EL34 or KT88) to the OPT's UL tap.
The 3 output tubes will interact with each other. Each tube sees the OPT in parallel with the plate resistance of the other two tubes as its load. Each tube eats a bit of power from the other two which is dissipated in it's plate, so you can't run the tubes at max bias. This is why the power output is lower than the sum of each tubes expected power output. Removing the cathode bypass caps increases each tubes output impedance reducing this effect. So does a minor bias tweak (slightly less current).
I have made the 3 output tube circuit work in this manner, in fact I had 6 tubes running at once in a test amp. I used a 6L6GC, an EL34, and a 6V6GT on each side of a P-P amp. I kept adding stuff and tweaking to get a different tone, and my "test amp" took up half the work bench and had about 20 knobs. It could go from one to 3 tubes in SE, and 2 to 6 tubes in P-P. It was far too complex to use in a real playing situation, so computer control is being added.
I have recently been "retired" after 41 years in the same workplace, and have moved out of Florida. I won't have a real workspace for a while (a year or more) so all this stuff is boxed up right now, but it will be back!
I find that the sounds that this amp produced are rather good through headphones. However I think you could produce this range of sounds through a single output tube with an amp that has a typical tone stack, ample gain and a master volume control, and an adjustment to the guitar's controls and playing style. It would be interesting to hear the same music played through the amp while ONLY the output stage controls were adjusted.
- Status
- Not open for further replies.