GU-50 PP triode project

[nerdorama]

I'm well on the way to having this one built. I mocked up a single channel on a breadboard and it sounded quite good. It was cathode biased for simplicity but made almost 20W at clipping. The final amp is fixed bias and hopefully a bit more. Should be class A up to around 16+W at least. The highlights in the top picture are reflections.

 

[maxhifi]

Wow, this is out of the ordinary! Care to share the schematic?

 

[nerdorama]

I have schematics but they aren't final. Here's what I have so far. The power transformer is a Tektronix 141 from an old scope. It has many windings and voltages so is very versatile. The output transformers are from an Eico ST-70 which is about 6500 ohms primary and very husky. They can handle much more wattage than this amp will require. On my mono breadboard the no feedback FR was -3dB at about 90kHz. That was with cathode bias. This project will use fixed bias with a bias/balance PC board from the Fisher forum courtesy of @ncwalz .

 

[maxhifi]

Really interesting. I had to look at the phase inverter a few times to understand it. Do you have the cage style GU50 sockets or just the bases?

 

[gadget73]

I thought a CCS in a long tail pair was typically done using a single CCS device supplying both cathodes? Is there some advantage to coupling the cathodes with a capacitor instead?

 

[maxhifi]

Yes and supplying negative voltage to the cathodes also doesn't make sense, as this would make the tubes biased to completely on. The choke seems to also be involved in the phase inversion, but it's not a conventional circuit.

 

[nerdorama]

The double CCS and cathode cap coupling is an idea from a PP GU-50 builder on AudioAsylum from several years ago. He had a hard time with bias stability on the 6E5P tubes so he put a CCS on each cathode and cap coupled the cathodes for the LTP action. The plate chokes are just something I had and decided to use them. The CCS's will keep the plate currents well balanced and the chokes allow the use of a much lower supply voltage. There are several things in this amp that are experiments for me. More wiring progress today.

 

[maxhifi]

The double CCS and cathode cap coupling is an idea from a PP GU-50 builder on AudioAsylum from several years ago. He had a hard time with bias stability on the 6E5P tubes so he put a CCS on each cathode and cap coupled the cathodes for the LTP action. The plate chokes are just something I had and decided to use them. The CCS's will keep the plate currents well balanced and the chokes allow the use of a much lower supply voltage. There are several things in this amp that are experiments for me. More wiring progress today.

View attachment 3733009View attachment 3733010

That color and those sockets makes it almost look like a Russian made piece of gear. Cool project, looking forward to the result.

 

[gadget73]

Plate chokes won't have anything to do with inversion per se, but if they are of the proper values you get comparatively low DC resistance and fairly high AC impedance, so it can give more voltage swing for a given plate voltage since its not losing a bunch in a resistor. The negative supply does work with a CCS to give more potential voltage swing too. It can be done with a resistor as well if you wanted to, allows for a much bigger tail resistor without giving up a ton of voltage across the tube. That inverter should have serious voltage output potential for comparatively low supply voltage.

I thought the whole appeal with the CCS is that you didn't have to pay any mind to matching, it just made it work regardless of tube particulars. Somewhere I've seen tests done using a CCS to force completely different triodes to give balanced output.

 

[maxhifi]

Plate chokes won't have anything to do with inversion per se, but if they are of the proper values you get comparatively low DC resistance and fairly high AC impedance, so it can give more voltage swing for a given plate voltage since its not losing a bunch in a resistor. The negative supply does work with a CCS to give more potential voltage swing too. It can be done with a resistor as well if you wanted to, allows for a much bigger tail resistor without giving up a ton of voltage across the tube. That inverter should have serious voltage output potential for comparatively low supply voltage.

I thought the whole appeal with the CCS is that you didn't have to pay any mind to matching, it just made it work regardless of tube particulars. Somewhere I've seen tests done using a CCS to force completely different triodes to give balanced output.

Then plate chokes are on a common core, they aren't individual from what I can see. So it's more like a push-pull auto transformer.

 

[nerdorama]

A single CCS in the tail would ensure equal AC outputs tube with the plate choke I thought it better to have balanced DC as well. I'll make measurements and post them once I have it running.

 

[Lavane]

Where did you find the retainers for those tubes? Personally I would
try some high heat paint on them like you find at auto parts stores.
As long as it can hold up to the temp of the tubes.

 

[maxhifi]

Where did you find the retainers for those tubes? Personally I would
try some high heat paint on them like you find at auto parts stores.
As long as it can hold up to the temp of the tubes.

Those are the correct sockets for GU50s. They're common USSR surplus items.

 

[gadget73]

The "hat" is just a handle right? Not actually any electrical connection?
**edit** just looked up the datasheet. Wow, 40 watts plate dissipation.

 

[maxhifi]

The "hat" is just a handle right? Not actually any electrical connection?
**edit** just looked up the datasheet. Wow, 40 watts plate dissipation.

The "hat" is a knob that is used to install and remove the tube into that holder. Without the knob, there's no easy way to get out out. You actually have to unscrew the knob to close the holder, and a lot of used GU50s are missing it for this reason. I have a single GU50 and socket/cage that I got as a souvenir on a trip to Europe, years ago. Never done anything with it because it's just the one.

 

[gadget73]

interesting, I'd never actually looked at the sockets before. Just figured it was a socket, not a whole thing with a retainer and a lid and all that. Def wouldn't have to worry about those getting knocked out of position

 

[Lavane]

The "hat" is a knob that is used to install and remove the tube into that holder. Without the knob, there's no easy way to get out out. You actually have to unscrew the knob to close the holder, and a lot of used GU50s are missing it for this reason. I have a single GU50 and socket/cage that I got as a souvenir on a trip to Europe, years ago. Never done anything with it because it's just the one.

Thats what I meant. The meatal part around the tube. So you have to unscrew the top of the tube to close the hat? Then screw the knob back on to remove the tube? I see cheap Chinese kits using this type tube without the meal cover.

 

[maxhifi]

interesting, I'd never actually looked at the sockets before. Just figured it was a socket, not a whole thing with a retainer and a lid and all that. Def wouldn't have to worry about those getting knocked out of position

The way the tube aligns, is it has a key molded into the glass, which fits into a slot in the retainer. There's no foolproof alignment with the retainerless Chinese sockets, where the tube could be inserted in the wrong orientation. It's a very robustly designed tube, RF output for mobile equipment.

Thats what I meant. The meatal part around the tube. So you have to unscrew the top of the tube to close the hat? Then screw the knob back on to remove the tube? I see cheap Chinese kits using this type tube without the meal cover.

The little black knob unscrews, leaving a threaded hole in the metal cap. Then you can close the retainer cover and it's held closed with a latch.

I've thought about doing something with GU-50s, but as interesting as they are, it's hard to make an argument apart from novelty for using them when the KT-88 and EL34 exist.

 

[nerdorama]

it's hard to make an argument apart from novelty for using them when the KT-88 and EL34 exist.

Not really. They are very rugged and robust and sound great. I have KT88 and EL34 and these are contenders. Very economical too relatively speaking. In the sims I've done the GU50 shows much lower THD in PP triode mode that either of them and also 300b's. I'm anxious to see how this amp measures.

 

[maxhifi]

Not really. They are very rugged and robust and sound great. I have KT88 and EL34 and these are contenders. Very economical too relatively speaking. In the sims I've done the GU50 shows much lower THD in PP triode mode that either of them and also 300b's. I'm anxious to see how this amp measures.

I'm interested how it turns out.