Push-pull amp using 6BM8/ECL82 tubes

[c.stoffel]

Hey guys, first post here!

I'm building this amp mixing elements of various somewhat similar amps like a Fender 6G2,  a Fender 5F10, a Gibson GA-8T, using a 12AX7 for preamp and 2 x RCA 6BM8 for a cathodyne PI, tremolo and a push-pull power amp. I'm using larger grid stoppers on the power tubes, and a 1M grid stopper on the PI input as well, like Merlin suggested in his website.

Well, first there was this annoying noise (kinda fizzy) when the notes were decaying, and the hum was too loud.  It was ruining the clean tones so I started testing the components, experimenting with different grounding schemes, different filtering, and nothing seemed to help. Even removing V1 didn't change the noise. Then I just swapped the power tubes and the noise changed (for worst). In the end it was one of the 6bm8s that was deffective, despite being NOS. Well, I guess next time I'll remember to check this first... lol

The amp is running fine now, the tone is very nice for playing blues/rock jams, which was indeed my idea for this amp. But as I've never used 6BM8s before, I was wondering if the experienced guys could help me checking if my circuit looks good? I've seen larger screen resistors (2k2, 4k7) in other designs using 6BM8s here at the forum (56T 6BM8, Little Wing), but I thought as I'm using a 1k resistor instead of a choke, I could make them smaller? Do my voltages look ok? Should I aim for almost 100% dissipation at idle for the pentodes, since this amp is cathode-biased?

I'm not sure about the value of the PI input cap, any opinions? 500pF is used by the GA-8T, and seems to give some more "clarity" compared to the 22nF used by Fender.
The cathode bypass cap at the 2nd triode deserves some experimentation too, it seems.

Well, any opinions will be appreciated! Thank you guys!

Best regards,
Christian

[tubenit]

My opinion is the 6BM8 tubes give a reasonable amount of bass tone.  I have found my best results with lower value cathode caps and lower value coupling caps.

I have listed values that I like which I feel like gives me a reasonably transparent and articulate tone.

I definitely would lose the 220uf on V1.  All that is going to do is give you a muddy tone, IMO.

Voltages look good to me.

AND thank you for sharing an innovative idea!  It is appreciated!

with respect, Tubenit

[c.stoffel]

Hey Tubenit, thanks for your help!

I will try lower values for cathode and coupling caps as you suggested! I'm really liking these 6bm8 tubes, great tone at lower power! I've built a Marshall 1987 clone and, while it sounds amazing, I quickly realized I REALLY needed something else to play at more sane volume levels.

So my power section looks good? Any opinions on the screen grid resistor values?

Thank you again!

Best regards,
Christian

[tubenit]

Any opinions on the screen grid resistor values

I typically have liked somewhat larger screen values like 1.5k- 2.2k range. They tend to have a more compressed tone on the cathode biased amps (to my ears) and I like that.

However, I've been trying some smaller values in this last yr or so and like those reasonably well also.

I think it depends on the tone/sound that you want.

With respect, Tubenit

[jbefumo]

Just off the workbench:

Recording quality doesn't do it justice.  Really pleased with the tone. 

[c.stoffel]

Just off the workbench:

https://youtu.be/wWvXW9zm8QU

Recording quality doesn't do it justice.  Really pleased with the tone.

Sounds good! I have never heard this bootstrapped cathodyne arrangement used in a guitar amplifier before. How would you describe the tone compared to a traditional cathodyne? Does it have a lot of gain?

Thanks for sharing.

[jbefumo]

Well, filched the bootstrapped cathodyne and the cascode front end from Merlin's site, so pretty hard to go wrong. 


I've never actually owned anything with a traditional cathodyne before, but comparing it to vids I've heard of the Univox U45B, I think it's smoother at higher levels.  Harder to compare it to Fenders that used that arrangement because of the different output tubes. Someone commented that it reminded him of a Vox tone, with a bit of "Fender rude," which I take as a complement.  I was about to dismiss the comparison, though, because I thought I had more or less started with a clean sheet, but then, looking at relevant schematics, I could see how the cascode first stage could be contributing to the tone created by the EF86 in early AC15s, and the cathodyne, of course, like some of those early Fenders, so yeah, why not? Standing on the shoulders of giants and all that.


It does have a lot of gain, and much of that I attribute to the lack of a tone stack. (Truth be told, I just didn't have the right parts on hand to implement any of the tone controls I really like.) I may try another version using 2 6L6s in Class A (18.5W), with this front-end and phase inverter, and one of the Blonde Fender tone stacks that use that 4-terminal treble pot.  I already used that in the Normal channel of my blackfaced Super Reverb and like it a lot.


My plan is to build that into a standard Tweed Champ box with an 8" alnico speaker -- probably one of Weber's.


Joe

Just off the workbench:

https://youtu.be/wWvXW9zm8QU

Recording quality doesn't do it justice.  Really pleased with the tone.

Sounds good! I have never heard this bootstrapped cathodyne arrangement used in a guitar amplifier before. How would you describe the tone compared to a traditional cathodyne? Does it have a lot of gain?

Thanks for sharing.

[jbefumo]

Oh yeah -- I did run a simulation comparing the bootstrapped setup to a LTP, and not unexpectedly, it had quite a bit more gain across the audio spectrum.

Just off the workbench:

https://youtu.be/wWvXW9zm8QU

Recording quality doesn't do it justice.  Really pleased with the tone.

Sounds good! I have never heard this bootstrapped cathodyne arrangement used in a guitar amplifier before. How would you describe the tone compared to a traditional cathodyne? Does it have a lot of gain?

Thanks for sharing.

[pompeiisneaks]

Wow very cool, I just recently decided to build an amp very similar to this, I'm using an ef86 in the preamp.  Idea was a very low wattage p-p amp.  Ok be sharing the details soon enough, when I get mine working.  This will really help!

Phil

[jbefumo]

I'm a big fan of the EF86, but was hesitant to use it in so small an so enclosed a combo.  I find the touch sensitivity of the cascode to be very similar. Another consideration was that space was really tight in this chassis, and I was really counting every component.  Would love to eventually side-by-side them in the same circuit.  At the moment I'm using a 12AU7, but will be ordering an E88CC to replace it shortly.


Joe

Wow very cool, I just recently decided to build an amp very similar to this, I'm using an ef86 in the preamp.  Idea was a very low wattage p-p amp.  Ok be sharing the details soon enough, when I get mine working.  This will really help!

Phil

[jojokeo]

Well, filched the bootstrapped cathodyne and the cascode front end from Merlin's site, so pretty hard to go wrong. 

The coscode front end has been widely implemented in the Firefly amp also. A 12ax7 ups the gain considerably.

[jbefumo]

Thanks for the pointer!


Actually, I'm thinking of ways to REDUCE gain!

Well, filched the bootstrapped cathodyne and the cascode front end from Merlin's site, so pretty hard to go wrong. 

The coscode front end has been widely implemented in the Firefly amp also. A 12ax7 ups the gain considerably.

[jojokeo]

Sorry, I messed up attaching the wrong file. That was the original 'fly without the coscode extra gain set-up. If yours is too much gain then I'd suggest popping in a 12au? Also try lessening or removing the cathode bypass cap, increasing the resistor to 1K or more.

[jbefumo]

It's an AU7 now -- will probably just reduce the plate resistor and tweak the cathode a bit.  I do like the touch sensitivity of that front end though.

Sorry, I messed up attaching the wrong file. That was the original 'fly without the coscode extra gain set-up. If yours is too much gain then I'd suggest popping in a 12au? Also try lessening or removing the cathode bypass cap, increasing the resistor to 1K or more.

[kagliostro]

You can also try splitting the plate resistor and feeding the following stage taking the signal at the junction

Franco

[jojokeo]

It's an AU7 now -- will probably just reduce the plate resistor and tweak the cathode a bit.  I do like the touch sensitivity of that front end though.

Sorry, I messed up attaching the wrong file. That was the original 'fly without the coscode extra gain set-up. If yours is too much gain then I'd suggest popping in a 12au? Also try lessening or removing the cathode bypass cap, increasing the resistor to 1K or more.

Your idea will help some but your were asking for high gain to begin with in using the coscode set-up in the first place as that's what it's all about. A nice benefit of the 12au7 is it should sound nice and chimey?

[kagliostro]

Splitting the plate resistor you preserve the point of work and feed the following stage with the desired level of signal

This didn't affect the actual touch sensitivity of the amp

Franco

[jojokeo]

joe,

I did some re-reading last night on the design aspects. The lower the load resistor (and associated bias point w/ corresponding cathode resistor along with it's bypass cap setting the low freq point) the lower the gain. The example was using the 6DJ8 tube as you were first using it since it was designed to be used in a cascode application and that's on your schematic. Now that you've changed the tube maybe you've also changed these things too? Without proper info you're supplying it's hard to help any further. There's also three different types of cascode circuits to utilize or available to try as I'm aware. But if you stick with what's on your schematic try lowering the load resistor and re-adjust the bias and you'll get to the point of the best you're going to get using the cascode configuration. Otherwise, gut the preamp portion and re-make in the standard grounded cathode application of your choice.

It's been more than several years since I've played around with this and at the time wanted to incorporate it into a new build at some point. After revisiting the info it's got me contemplating using it again it some manner and possibly in a new pedal too? We shall see? 

[jbefumo]

Yup -- actually, I did design it for the 6DJ8, but am just using the 12AU7 until I order the 6DJ8 later week.  Putting off any further playing until I get the final tube in and see what it sounds like.  Thanks! 

joe,

I did some re-reading last night on the design aspects. The lower the load resistor (and associated bias point w/ corresponding cathode resistor along with it's bypass cap setting the low freq point) the lower the gain. This was using the 6DJ8 tube as you were first using since it was designed to used in a cascode application and that's on your schematic. Now that you've changed the tube maybe you've also changed these things too? Without proper info you're supplying it's hard to help any further. There's also three different types of cascode circuits to utilize or available to try as I'm aware. But if you stick with what's on your schematic try lowering the load resistor and re-adjust the bias and you'll get to the point of the best you're going to get using the cascode configuration. Otherwise, gut the preamp portion and re-make in the standard grounded cathode application of your choice.

It's been more than several years since I've played around with this and at the time wanted to incorporate it into a new build at some point. After revisiting the info it's got me contemplating using it again it some manner and possibly in a new pedal too? We shall see? 

[jbefumo]

Funny thing is that I generally favor lower gain, high headroom amps, then always build in too much gain the first pass.  Reminiscent of building a big-block for an old '68 Firebird convertible -- did all this research, knew that the way to go for a street car was smaller carb, mild low-end cam, and moderate compression, but when I got to the parts store with money in hand .... when for 12:1 compression, hot cam, mammoth Holly double-pumper -- monster engine, but never could manage to idle .... but I digress ...

joe,

I did some re-reading last night on the design aspects. The lower the load resistor (and associated bias point w/ corresponding cathode resistor along with it's bypass cap setting the low freq point) the lower the gain. The example was using the 6DJ8 tube as you were first using it since it was designed to be used in a cascode application and that's on your schematic. Now that you've changed the tube maybe you've also changed these things too? Without proper info you're supplying it's hard to help any further. There's also three different types of cascode circuits to utilize or available to try as I'm aware. But if you stick with what's on your schematic try lowering the load resistor and re-adjust the bias and you'll get to the point of the best you're going to get using the cascode configuration. Otherwise, gut the preamp portion and re-make in the standard grounded cathode application of your choice.

It's been more than several years since I've played around with this and at the time wanted to incorporate it into a new build at some point. After revisiting the info it's got me contemplating using it again it some manner and possibly in a new pedal too? We shall see? 

[jojokeo]

Funny Joe, I can totally relate. Hard to remember the name exactly but back in the day was a local Super Shop or Super Sport Shop or something like that? I had everything you could want. This car originally had the Craiger rims w/ the tire of the day BF Goodrich T/A radials before putting on the Centerlines. I had the spun aluminum buffed out to a high shine - better than chrome. All the neighborhood cats will attest to that too. The damn things always new right after I polished them cause right after they'd always have to leave their marks all over them...little bastards! It also had the turbo mufflers, headers, elect. ignition, manifold, carb, super cool cassette stereo & 3-way speakers, power amp with equalizer (was all great at the time) etc. etc. etc...my toy & love machine that took me through high school and beyond. I went through 3 or 4 paint jobs on it. The license plate said GIT PLYR and I can't tell you how much fun and trouble I had in this car. But when I say trouble it wasn't due to breakdowns. It never let me down or stranded and in fact I was always picking up & taking friends to pick up their cars from that stuff, whether Chevy's to VWs. My girlfriends' parents always knew when I was coming and going it would rattle windows so sometimes I had to park down the street 

[kagliostro]

the ecc88 has a gain of 33 which is higher than the ecc82 who has a gain of 17, so, if the problem is/will be the excess in gain

why don't use a tone control ? As we know they suck signal

Franco

p.s.: I apologize c.stoffel, seems we have hijacked your thread

[jbefumo]

I could easily be mistaken, but I think there's more at play than just the mu factor, which I believe is just the maximum achievable gain. 


Didn't use a tone control because I was barely able to fit the circuit into that minuscule space as it was. Even my usual BMP tone control built right onto a push-pull pot wouldn't fit. I know there are other simpler tone controls, but without the bypass you're locked in, and if it's only going to buy me a simple treble bleed, might as well just use the tone control on the guitar. If I build another it will be in a Tweed Champ or Princeton cabinet/chassis, and I plan to use a Blonde Fender type tone stack.


As it stands I still like the tone of the amp, and did want it to have more gain on tap than my usual.  I think just going to a 33k plate resistor will do the trick, along with corresponding change to cathode, but don't want to mess with it until I get the tube for which the circuit was designed -- just waiting for next Social Security check ;^)

the ecc88 has a gain of 33 which is higher than the ecc82 who has a gain of 17, so, if the problem is/will be the excess in gain

why don't use a tone control ? As we know they suck signal

Franco

p.s.: I apologize c.stoffel, seems we have hijacked your thread

[jojokeo]

Forgot to mention something along the same "loading" lines...if you were to change out the 1M Vol pot to 50K, 100K, 250K, etc - you'll get a lot more of a loading effect on the circuit without changing anything else therefore lowering your gain to whatever you like...simple.

[jbefumo]

Just acquired an ECC88, installed and reduced the plate resistor to 33k.  Sounds incredible. Note Bene: for other old fogies who refuse to admit they need glasses, that dotted line from pin 9 that appears to be the expected filament center tap ... it isn't - it's a shield. That tube wants 6.3V across pins 4 & 5 ...

[jojokeo]

Sweet! Sounds like you're in your happy place now?

[jbefumo]

Definitely pleased with this design. Looking forward to building on in a tweed Champ chassis and cabinet.  The larger space will allow tone controls and a less congested board layout. I'm totally sold on that tube combination, though.

Sweet! Sounds like you're in your happy place now?