What's Your Favorite 6550 for a JCM800?

[FourT6and2]

What screen voltages are EL34's rated for???

JJ EL34: 450V
JJ EL34 II: 450V
JJ E34L: 450V
JJ KT77: 600V
Gold Lion KT77: 650V
JJ6CA7: 500V
Tung Sol EL34B: 425V
EH EL34: 425V
Philips EL34: 500V
RFT EL34: 500V
PSVANE EL34C: 500V
Mullard Reissue: data sheet does not specify
Svetlana New Sensor: data sheet does not specify
SED Winged C: 500V
TAD EL34-STR Red: 500V
TAD EL34-STR Black: 500V
Shuguang EL34B: Data sheet does not specify, but other sources say 500V
Ruby EL34B-STR: 450V
Ruby EL34-BHT: 560V

That's all I could find quickly. And it's important to note that you should technically check both screen and plate voltage to make sure a tube is up to snuff. For example, while one of those tubes can handle 500V on the screens, the plate can't. And the opposite is also true in some cases.

The 6550 thing was mostly Marshall installed the heaviest duty tubes to ship to the USA because they were having some many EL34's damaged in transit.

It wasn't Marshall. It was the U.S. importer/distributor. They didn't want to deal with the warranty issues associated with EL34. The U.S. importers and/or distributors were also responsible for various other anomalies over the years like the polarity switch and death cap, additional speaker outputs (4 in total), and yes, the 6550s. For Canadian markets, the amps had metal toggle switches instead of the plastic rockers, additional fuse boards inside the chassis, and lower voltage power transformers.

I'd say mod the JCM800 with a large bias range with a 50K pot to where it can run any tube type. EL34, KT77, 6L6, 6550, KT88. That's what I did with my LG.HG Cameron and it worked out great, I have a thread on it in the tech section when I did it.

It's not so much the bias range resistor you need to worry about. It's the phase inverter output grid leaks (bias splitters). Each tube type has a specified maximum total resistance it can handle on its grid. This number is the sum total of these resistors, the bias resistors, and the grid stopper resistors. This is why Marshalls that had 6550s have 150K or even 82K in the PI instead of 220K (EL34). It's also why the very first bias voltage dropping resistor is different and not just the bias range resistor.

Changing the PI grid leaks changes the low frequency response of the amp. And if you change those, you need to also change the PI coupling caps to compensate in order to keep the same bass response as the EL34's 220K/0.022uF combo.

Overall I'd say any Beam forming element tetrode like the EH6ca7 or 6L6 and i think a good Kinkless tetrode KT77 would give you a very similar experience like the 6550 IMHO. But that said I've alwasy loved the 6550's in my 79 Marshall compared to EL34's, I modded it to run EL34's and tried a set of real SED EL34's and I went right back to my 6550's in that amp so it is an acquired taste.

I think the outrageous cost of good 6550s is prohibitive. Modding the amp back to its original EL34 design and running KT77 or 6CA7 is probably a good idea. But I've had great luck with some new production EL34 lately. That said... this amp is only putting out ~450V on the screens. So even 6550s rated at 450V will probably do just fine, considering most people run EL34s rated at less without even knowing...

 

[harddriver]

JJ EL34: 450V
JJ EL34 II: 450V
JJ E34L: 450V
JJ KT77: 600V
Gold Lion KT77: 650V
JJ6CA7: 500V
Tung Sol EL34B: 425V
EH EL34: 425V
Philips EL34: 500V
RFT EL34: 500V
PSVANE EL34C: 500V
Mullard Reissue: data sheet does not specify
Svetlana New Sensor: data sheet does not specify
SED Winged C: 500V
TAD EL34-STR Red: 500V
TAD EL34-STR Black: 500V
Shuguang EL34B: Data sheet does not specify, but other sources say 500V
Ruby EL34B-STR: 450V
Ruby EL34-BHT: 560V

That's all I could find quickly. And it's important to note that you should technically check both screen and plate voltage to make sure a tube is up to snuff. For example, while one of those tubes can handle 500V on the screens, the plate can't. And the opposite is also true in some cases.


It wasn't Marshall. It was the U.S. importer/distributor. They didn't want to deal with the warranty issues associated with EL34. The U.S. importers and/or distributors were also responsible for various other anomalies over the years like the polarity switch and death cap, additional speaker outputs (4 in total), and yes, the 6550s. For Canadian markets, the amps had metal toggle switches instead of the plastic rockers, additional fuse boards inside the chassis, and lower voltage power transformers.


It's not so much the bias range resistor you need to worry about. It's the phase inverter output grid leaks (bias splitters). Each tube type has a specified maximum total resistance it can handle on its grid. This number is the sum total of these resistors, the bias resistors, and the grid stopper resistors. This is why Marshalls that had 6550s have 150K or even 82K in the PI instead of 220K (EL34). It's also why the very first bias voltage dropping resistor is different and not just the bias range resistor.

Changing the PI grid leaks changes the low frequency response of the amp. And if you change those, you need to also change the PI coupling caps to compensate in order to keep the same bass response as the EL34's 220K/0.022uF combo.


I think the outrageous cost of good 6550s is prohibitive. Modding the amp back to its original EL34 design and running KT77 or 6CA7 is probably a good idea. But I've had great luck with some new production EL34 lately. That said... this amp is only putting out ~450V on the screens. So even 6550s rated at 450V will probably do just fine, considering most people run EL34s rated at less without even knowing...

Yeah I think it was the Unicord US distributor who started the 6550's. I have the original owners manual for my 79 2203 and it has Unicord all over it. I used the have the Big Book of Marshall years ago for all that information.

I am aware about the PI bias resistor and PI caps you have to strike a happy medium there if you use the 50K bias pot. I went with 150K's and either .1uf or .047uf I'd have to crack the amp back open, I think .1uf's and 110K would be a 1:1 equivalent to the EL34 220K/.022uf specs. Even though Marshall used 150K/.022uf in the bias splitters and it proved out to work. Years ago Larry (Novsibir) recommended 82K's specifically for technical reasons about 6550's in one of my high gain 2203 4 gain stage builds and so back then I went with the 82K bias splitters and started at .1uf PI caps since that is a close frequency equivalent to 220K/.022uf PI specs for EL34's and I think I ended up going to .047uf caps to shave a little more low end out and make the amp really tight on the chugs and that amp was set up for 6550/KT88's specifically.

If you want to hear 6550's I think with 450 screen voltage you'd probably be safe enough. Remember those Ruby Shugaung 6550A had a checkered past about fluctuating bias to wards the hot end and they weren't stable when I bought a couple of quads, I had to shit can at least two of them out of a quad and really didn't trust them after that and I won't buy any more of those. I think if I had to buy another quad which I should I'd look at the new sensor Sovtek 6550's, I ran those in a couple of amps and they were pretty decent all around.

 

[midnightlaundry]

Fryette says it's the Impedance mismatch of the Output Transformer that changes the response of different Power Tubes and not the Tubes themselves..

That said, in my Trainwreck build I've tried EL34s, EL34L's which are pretty much the same. KT77's have a bit more topend sizzle and seemed more compressed. 6CA7's gave a big bass response that is welcome in that amp.. I'm running those at this time..

 

[Greazygeo]

You can expect better mids than from any Wizard.

Bullshit…had enough of them to know. Turds.

 

[Beyond Black]

Bullshit…had enough of them to know. Turds.

Yeah? Then why’d you buy more than one?

 

[Racerxrated]

Bullshit…had enough of them to know. Turds.

You must've missed the 82/83 year 2203/4s. The higher PV versions are simply killer; and one 83 2203 sounded so good I sold the 2010 MC100. The MC is a great amp but that particular 2203, stayed right with it....just a killer 2203.

 

[FourT6and2]

You must've missed the 82/83 year 2203/4s. The higher PV versions are simply killer; and one 83 2203 sounded so good I sold the 2010 MC100. The MC is a great amp but that particular 2203, stayed right with it....just a killer 2203.

What constitutes higher plate voltage in these amps? The only other actual 2203/JCM800 from those years I've personally seen was a late '82 running at 460V (PT: T4145-B) Mine is a January '82 with 1981 transformers running at 450V (PT: T4145-A)

 

[Racerxrated]

What constitutes higher plate voltage in these amps? The only other actual 2203/JCM800 from those years I've personally seen was a late '82 running at 460V (PT: T4145-B) Mine is a January '82 with 1981 transformers running at 450V (PT: T4145-A)

I've had 3 or 4 2203s/4s from 82/83 that the loaded PV is 485, 490v. I think this started by mid 82 and lasted all of 83. Early 82s are similar to the number you have. My present 2203 is a last yr JMP, and it's at 465v. Ted had one of them, a white head shell 82 2204 with a Monomyth mod at one point....unloaded pv was 515v if I remember right.

 

[FourT6and2]

I've had 3 or 4 2203s/4s from 82/83 that the loaded PV is 485, 490v. I think this started by mid 82 and lasted all of 83. Early 82s are similar to the number you have. My present 2203 is a last yr JMP, and it's at 465v. Ted had one of them, a white head shell 82 2204 with a Monomyth mod at one point....unloaded pv was 515v if I remember right.

What are the transformer codes on the ones you have/had with 480V or higher? T4145 A, B, C, D, etc.? The vertical input 2204s have Drake iron and 2203s have Dagnall, so that's a bit of a different story.

 

[Anje]

I will state the obvious, but would be good to list the input power voltage the amps sees + voltage selector tap / transformer primary wiring used + point of measure & tube config when comparing power transformer output voltages.