Can I use 6550s in my JTM45 clone ?

[gtrplayr1976]

Hi folks, I have just finished a JTM45 clone ,and I want to use 6550s. I have Classic Tone xfmrs. What do I need to look for to be sure they are safe to use ? I don't want to burn up a xfmr. I have them installed ,and biased, but want to be sure.

Thanks !

[The Dude]

I can't see why not. Filament current draw will be nearly the same as EL34's. Tube current will be based on bias. You may want to list what OT you are using, although impedance from EL34 to 6650 should be similar enough. Does it sound good?

[big_teee]

On the JCM800 they used a few different resistor values with the EL34s, vs the 6550s.
See Schematic, it shows resistors for both 34s, and 6550s.
http://mhuss.com/MyJCM/JCM800_2204.pdf
It would probably be similar with a JTM45?
T

[The Dude]

Yes, it would be helpful to have a schematic of what you built rather than just the layout.

[drewl]

Marshalls shipped to the US were sent without output tubes where 6550's were installed.

In the early 70's they changed the bias feeds to 150k for them but was often not needed if there was enough bias range.

I have a '73 50 watt MKII which probably had 6550's new, but already had EL34's in it when I got it in the late '80's.

I have since changed the 150k's to 220k but it has enough range for either output tube.

That sucker screams!

[gtrplayr1976]

I can't see why not. Filament current draw will be nearly the same as EL34's. Tube current will be based on bias. You may want to list what OT you are using, although impedance from EL34 to 6650 should be similar enough. Does it sound good?

It sounds great. Plenty of bias range. The OT is a Classic Tone also.

[gtrplayr1976]

On the JCM800 they used a few different resistor values with the EL34s, vs the 6550s.
See Schematic, it shows resistors for both 34s, and 6550s.
http://mhuss.com/MyJCM/JCM800_2204.pdf
It would probably be similar with a JTM45?
T

I really didn't go by one specific schematic or layout ,but 90% of the layout I attached. I added a PPIMV.

[gtrplayr1976]

Yes, it would be helpful to have a schematic of what you built rather than just the layout.

The attached schematic shows some of the differences between the actual schematic ,and what I have done. Stuff that makes the most difference I think.

[mhuss]

The bias changes may or may not be needed depending on your circuit values.

The 150k/220k change has to do with the maximum grid resistance spec of the two tubes. Clearly, 6550s (and other beam plate types) will work with 220k grid resistors, but I'm thinking there was a reason for that spec.

And actually, the 6550 'maximum grid 1 resistance' for fixed bias is 50k(!), which no one is using. So I wonder.

[gtrplayr1976]

The bias changes may or may not be needed depending on your circuit values.

The 150k/220k change has to do with the maximum grid resistance spec of the two tubes. Clearly, 6550s (and other beam plate types) will work with 220k grid resistors, but I'm thinking there was a reason for that spec.

And actually, the 6550 'maximum grid 1 resistance' for fixed bias is 50k(!), which no one is using. So I wonder.

The 220k resistors were replaced with a dual 250k pot with 2m2 resistors to produce the PPIMV.

[potatofarmer]

And actually, the 6550 'maximum grid 1 resistance' for fixed bias is 50k(!), which no one is using. So I wonder.

...because nearly every guitar amp has ignored that spec in favor of not having to design a better PI?

Thermal runaway, pfft; sounds like user error.

[dai h.]

why do they seem to survive with 150k grid load Rs in Marshalls? Was it anything to do with the robustness of particular 6550s (General Electric were the ones Marshall's US distributor installed for US and Japan models--they seemed to have a reputation for reliability)?


(couple factory datasheets for reference which show 50k max for g1 R) :

http://www.radiostation.ru/tubes/6550A.pdf

http://www.nj7p.org/Tubes/PDFs/Frank/hb3/6550.PDF

http://www.newsensor.com/pdf/tungsol/6550-tungsol.pdf

[mhuss]

I have a U.S. JCM 800 that still has the original GE 6550s in it, and they're still going strong. I don't know why this spec is so widely ignored, but it doesn't seem to affect the tubes that were being made when the spec was published.

I note the EL34/6CA7 have a 700kohm rating for this spec.

[pdf64]

I think that the grid leak limiting value may become critical if the tube is being operated close to its limiting plate dissipation; the higher temperatures resulting in increased control grid secondary emission, which may result in thermal runaway if that grid current, passing through the grid leak resistor, causes the control grid to become more positive than intended.
Hence the grid leak limit is lower in fixed bias than cathode bias, which would tend to self regulate in that scenario.

In a Marshall, 6550 shouldn't get near their plate dissipation limit.

Just to note that their static dissipation shouldn't be anywhere near 70%, but rather should take the intended load / likely capabilities of the power supply into account, eg similar static current draw to EL34.

[gtrplayr1976]

I think that the grid leak limiting value may become critical if the tube is being operated close to its limiting plate dissipation; the higher temperatures resulting in increased control grid secondary emission, which may result in thermal runaway if that grid current, passing through the grid leak resistor, causes the control grid to become more positive than intended.
Hence the grid leak limit is lower in fixed bias than cathode bias, which would tend to self regulate in that scenario.

In a Marshall, 6550 shouldn't get near their plate dissipation limit.

Just to note that their static dissipation shouldn't be anywhere near 70%, but rather should take the intended load / likely capabilities of the power supply into account, eg similar static current draw to EL34.

I think I have it at about 60% maybe a little higher.