Trying to ID some RCA octal tubes - probably 6L6GC but could be 8417?

[buddy_coles]

Hi All,

I acquired a homebrew amp a while back and I am gradually trying to piece together parts of the mystery - I don't know if there were design errors, or if it was designed for something other than guitar...  Anyway, I am trying to get a firm ID on the output tubes so I can set up the bias and speaker impedance correctly.  I believe these are 6L6GCs, but with no ID on the tubes it is hard to be 100% sure.  My confusion is because I have 640VDC of plate voltage which seems ridiculously high!  The screen voltage is regulated to 400V (well, that's the value scratched in to the chassis next to the adjustment pot) by means of a neon 0A2 tube and 1/2 a 12AX7 and a 6L6GC in a neat regulator circuit.

The only ID I can see is the RCA logo in red on the bases of the tubes, with one having 'DO' and the other having 'Made in USA BW'
I've attached a couple of photos for info.

If anyone can positively ID these I would be really grateful - a virtual beer will be all yours! 
Many thanks for any advice you can give,

Sid

PS I'll post a circuit diagram and a bit more info soon - got a Mouser delivery coming today so I'm going to tinker with the circuit later on...

[astronomicum]

The 640V would not rule out 6L6GCs. Univox used voltages that high. I took a look at one I have and it is much shorter with no top getter. The support rings on your tubes are much thicker and there are pads on the top ring that would help dampen vibration (both different than the example I have). Could be a ruggedized version of the 6L6 (5881?).

[astronomicum]

The 640V would not rule out 6L6GCs. Univox used voltages that high. I took a look at an RCA 6L6GC I have and it is much shorter with no top getter. The support rings on your tubes are much thicker and there are pads on the top ring that would help dampen vibration (both different than the example I have). Could be a ruggedized version of the 6L6 (5881?).

[buddy_coles]

Thanks astronomicum!
I'll check out some Univox schematics to see how similar they look to my circuit.  I believe some Sovtek amps in the 70s also used very high B+ for the 6L6 plates so it may yet be OK.  I just want to be sure of what I have, in order to set them up without frying anything (especially myself!).
I've attached the schematic for reference - I sketched it out as a starting point before doing anything to the amp so I knew what was what - I've since replaced the electrolytics and fired it up, and it makes the right kind of noises but there's almost no gain from the pre-stages.  I'm going to tinker, starting at the input and working my way forwards, to try to get some Fender-esque tones out of it.  I'm intrigued to hear how the pentode sounds once I drive it a bit...

[sluckey]

That bias circuit is dangerous. I suggest putting a resistor between the bias pot and ground to prevent turning the pot down to zero volts.

[buddy_coles]

That bias circuit is dangerous. I suggest putting a resistor between the bias pot and ground to prevent turning the pot down to zero volts.

Hi Sluckey,
Agreed - it's on my mod list to add a 10K fixed resistor between the 10K pot and ground to give a bit of protection from user error :)  I'll probably also add a fail-safe 100K resistor between the hot side of the pot and the wiper, in case of wiper failure.  I'm guessing whoever built this never intended for anyone else to adjust it...!

[PRR]

The way the NFB is injected under the pentode is not going to work as planned.

All 8417 are the same height (6L6 types vary). Check against the datasheet.

[astronomicum]

it makes the right kind of noises but there's almost no gain from the pre-stages.

Check plate voltages for V1. Are V1B, V2A and V2B all being fed by the same 39K Plate Load Resistor?

[buddy_coles]

The way the NFB is injected under the pentode is not going to work as planned.

All 8417 are the same height (6L6 types vary). Check against the datasheet.

Thanks PRR - the datasheet I found here https://frank.pocnet.net/sheets/168/8/8417.pdf gives a maximum height for the glass and the complete tube.  I'm afraid I have no experience here, so would all 8417s have dimensions equal to the maximum values shown?  Or is there a RCA datasheet that I didn't find?

Now, the two tubes I have are different in height by about 1/4", but both within the prescribed range for 8417s.  BUT I just looked at the 6L6GC datasheet https://frank.pocnet.net/sheets/127/6/6L6GC.pdf (it's the Tung-Sol one but hopefully representative of all 6L6GCs) and one of the tubes is taller than the maximum height...

I think I need to post these to my buddy who has all the kit for testing valves and creating the various graphs, and see if they are indeed a matched pair - they are in the positions they were in when I got the amp but that's no reason to assume they are where they should be!  There's another 6L6GC (still has the writing on the glass) used in the regulator circuit.  I'd assumed all three were all the same but now I'm not so sure 

Re. the negative feedback arrangement, that's something I'll look into a bit later in the project - I'll post elsewhere for circuit queries.  I very much appreciate the tip, but for now I just want to make sure I can correctly bias the output tubes and match the speaker impedance by identifying the tubes.  I have no doubt I'll be back many, many times with questions about the circuit!!!

Thanks again, Sid

[buddy_coles]

Check plate voltages for V1. Are V1B, V2A and V2B all being fed by the same 39K Plate Load Resistor?

Yes, that is correct - not a nice arrangement and I will simply remove V2 for now.  I have a feeling this amp was not designed for guitar; probably more like hi-fi (the Baxandall tone stack is not very intuitive) or possibly bass, if an active bass was used...  I measured just 27VDC at the plate of V1A at idle!!!

I'll post separately about the rest of the circuit - for now I'm really just trying to work out what I have in terms of output tubes 

Thanks again for your help though - very much appreciated, and I'll post again here when I start another thread regarding bits of the circuit.  There's a lot of stuff I think I'll need to ask about...

[PRR]

So put in a pair of 100 Ohm cathode resistors, install one power tube and meter it, turn bias to MAX negative, and cautiously power-up. Sneak it up to 4V (40mA) on cathode resistor. Note grid voltage. Repeat on other power tube.

The Ik Vg1 Vg2 numbers should point to 6L6 or 8417.

I would think "all" 8417 would be within 1/8" of the "MAX height" spec. Why would it vary? However 6L6 has a much longer history and many odd variants. Sorry that this is no help for you.

[2deaf]

I believe some Sovtek amps in the 70s also used very high B+ for the 6L6 plates so it may yet be OK.

Nyet.  Sovteks didn't appear until the 90s.  530V was a pretty typical plate voltage.  Nobody knows what those tubes were in those amps, but they were some tough boys.  They said 5881 on them, but they weren't the same as our 5881's.

[buddy_coles]

Thanks 2deaf, my mistake re. the Sovtek era!  And also re. the voltage...  However, I had a look at some of the Univox circuits and although they use a voltage doubler rather than a very high voltage secondary, the voltages look similar to what I am seeing (640VDC of B+ on the OT, 300-400VDC on the screens) so I feel a bit less worried now. 

Thanks to all who have posted on this thread - I really appreciate your input and advice.

[HotBluePlates]

... I am trying to get a firm ID on the output tubes so I can set up the bias ... correctly.  I believe these are 6L6GCs, but with no ID on the tubes it is hard to be 100% sure.  ...

Right away, I say no-way these tubes are 8417.  I've got an amp that uses those, and the plate-shape & screen spacing is distinctive.  See here & here.

At a glance the plate-style looks like RCA 6L6GC.  The tall-bottle also reminds me of Sylvania 6L6GCs.  See what you can sort out about which tube element goes to which pin, specifically whether Pin 4 is connected to Pin 1 and to the tube's screen.  A 7027 is very much like a 6L6GC but has that extra connection to Pin 1.  See below for photos of 7027s that seem similar to your tubes (also here and here):



I agree with others that you want to sneak up on the bias to be certain of the tube type.

[buddy_coles]

Thanks HotBluePlates - I think the mystery is solved!! Pins 1 and 4 are connected, as are pins 5 and 6. So I will assume 7027 for now. Really appreciate your reply and the very useful links.
For now I have set a very cool bias, and having tweaked the preamp circuit here and there I am getting some reasonably pleasing noises out of the amp.
Thanks again, Sid