Recommend output tubes for tube PA conversion (50H-B26 to ???)

[glebert]

I've got an old Elk Vocal HiFi 50 tube PA head I've been thinking about converting into into some sort of awful guitar amp by cascading some channels and other stupid stuff. No schematics available, but the channel circuits looked pretty similar to the Fender tube PA schematics. The big difference is the output tubes, as it had 50H-B26 tubes, which are unobtanium at this point. I'll have to change sockets and get a filament transformer since the 50H-26B used mains voltage directly for heater supply. It seems like most people go to a 6L6 type of tube to replace these tubes. There was a different version of this amp (Vocal HiFi 80) that used EL-34 and there was someone who said it sounded amazing for guitar even unmodded. No idea on if it used the same OT or not though, and I wasn't planning on replacing the OT if I didn't have to.

Here is a datasheet for the 50H-B26. but I don't know that the translations are accurate(?) https://frank.pocnet.net/sheets/084/5/50H-B26.pdf

Thoughts?

[Enzo]

Do you have a schematic?

The fact it has a 50v heater suggests the other tubes heaters are wired in series with it. You mention the heater is wired to the mains. But that also suggests you may not have a power transformer at all, is that right? If so it is what we call a widow maker - a hot chassis. Also, if they ar series heaters, you likely need to replace ALL the tubes.

[The Dude]

Here's a thread that may be useful.

https://music-electronics-forum.com/...estion?t=28543

[glebert]

Do you have a schematic?

The fact it has a 50v heater suggests the other tubes heaters are wired in series with it. You mention the heater is wired to the mains. But that also suggests you may not have a power transformer at all, is that right? If so it is what we call a widow maker - a hot chassis. Also, if they ar series heaters, you likely need to replace ALL the tubes.

Sorry, thought I had said there was no schematic that I have found for either this one or the Vocal HiFi 80. Definitely has a PT. The channels work without the power tubes installed (taking an output at the effects send) so no series heaters. I'm guessing that they could save some cost on the transformer by not having to do the high current heater leads. The 6.3V tap on the PT is labelled as 0.27A.

[glebert]

Pics

[glebert]

Not so blurry pic

[g1]

I worked on an Acetone 301 recently, it had a separate 50V tap to run that tube heater if I recall correctly. It was a single ended power amp.
Also saw a schematic where they used a crazy amount of them in parallel to drive a speaker with no OT.

[glebert]

I worked on an Acetone 301 recently, it had a separate 50V tap to run that tube heater if I recall correctly. It was a single ended power amp.
Also saw a schematic where they used a crazy amount of them in parallel to drive a speaker with no OT.

Now THAT is weird.

[J M Fahey]

So it looks like power tubes are replaceable by 6L6 , with some effort.

I´d do it as per the other thread suggestions, rewiring to new octal sockets as neded, connecting screen to proper voltage (do not forget screen resistors) and calling it a day.
As of filaments, I´d add an extra 6.3V filament transformer, properly rated.
Small size anyway, it will fit "somewhere".

[g1]

Seeing as these 50H-B26 run at lower voltages than common in guitar amps, 6V6 may also be an option.

[Helmholtz]

From the datasheet original tubes call for a Zaa of 3.2k per pair. So if that's the OT's primary impedance it would fit best with EL34s (typical Zaa: 3.4k).

6V6s need higher load impedance, like 8k.

[glebert]

From the datasheet original tubes call for a Zaa of 3.2k per pair. So if that's the OT's primary impedance it would fit best with EL34s (typical Zaa: 3.4k).

6V6s need higher load impedance, like 8k.

I'm guessing it would be helpful to know what speaker load is going to be used? There is no marking on the mixer what the assumed load would be. 4/8 ohm might be the logical assumption but with wacky amps like this who knows? Maybe I should measure the voltage ratio I get from the OT and infer the impedance ratio?

[Helmholtz]

Maybe I should measure the voltage ratio I get from the OT and infer the impedance ratio?

Definitely yes.

[J M Fahey]

From the datasheet original tubes call for a Zaa of 3.2k per pair. So if that's the OT's primary impedance it would fit best with EL34s (typical Zaa: 3.4k).

6V6s need higher load impedance, like 8k.

Not that simple.
Both cases you mention suggest those impedances for very specific +V voltages and desired power outputs, which are NOT what´s present here.

Proper Zaa is not a mantra to be repeated by heart no matter what but a design choice, based on actual conditions.

I bet original 3k2 comes from such considerations, and if supply Voltage and expected Power out are not changed, they are good even with different tubes, as long as the new ones can stand current voltage and pass needed current.

Both conditions are easily met by 6L6 replacing weaker 50H-B26 .

Will OP get, say, 50W as easily pulled out of a 6L6 pair?

No, he will get same power out as before , but newer stronger 6L6 will easily supply that, under very little stress.
personally I would consider them "50H-B26 on steroids" , go figure.

FWIW EL34 will *also* work fine there, doing the original tubes job.

Zaa is not a magical number by any means, nor does "apply everywhere, no matter what" ,it comes straight from available voltage swing divided by available current swing, pure Ohm´s Law by definition.
Nitpicking: "multiplied by 4" because I just defined Za , the optimum load impedance for one tube, but convention/tradition dictates that transformers are labelled by plate-to-plate impedance.

But in any case: not an immovable "tube property" but something which comes from actual circuit design.

So much so, that even for the same tubes, power and supply, you can choose different load impedances, different load lines, based on personal preference; design is always a compromise.

[Helmholtz]

Not that simple.

Simple or not, just saying that EL34s would be a good match and 6V6s not. 6L6s will probably work fine as well, especially if B+ is below 400V.

Optimal plate load is a compromise between max. clean output - and staying within plate dissipation as well as screen dissipation limits for given supply voltages.
So it's not as easy as applying Ohm's law.
Datasheet recommendations can be expected to take care of that all. I consider them a good reference for specified conditions.
Often different Zaa values are specified for different operating conditions.

Supply voltages are not known here.

Lower supply voltages allow for some flexibility regarding plate load.
High supply voltages often demand for higher plate impedance to stay within plate dissipation limits.

Generally lower plate impedance increases plate dissipation at medium output and higher plate impedance increases screen dissipation especially at full output.


Even at a plate voltage as low as 300V I expect excessive plate dissipation with 6V6s and a Zaa of 3.2k. Doubling speaker impedance would help.