Are my voltages whacked?

[DrewTube]

I just finished building my new Eros.  All the resistances check out (to a surprising and comforting degree of precision).  So I put in the tubes (stock EH) and fired it up for the voltage checks.  All the heaters came online, all 12 LED's are lit, but my voltages on the C4S/Servo boards are a little off and I'm wondering if this is cause for concern. I also noticed that if I swap the EF86's, the OB voltage moves with the tube. I happen to have a spare 6922 and swapping that tube did not change anything by a significant degree.

IA -    Right = 223.5 - Left = 224.1
Kreg - Right = .008   - Left = .008
breg - Right = 67.6   - Left = 79.2
OA -   Right = 208    - Left = 186.4
bA -   Right = 0       - Left = 0
OB -   Right = 63.3   - Left = 76.5

[Doc B.]

Is K reg really .008V? Should be more like .8V, close to 1V anyway.  I would suggest letting the tubes burn in for several hours, maybe overnight, and checking the voltage again.

[DrewTube]

Yeah, that seemed weird to me, too...so I checked it on different ranges and two different meters...it's definitely 8mV.  And it's the same for both channels.  I'll let the tubes burn in a bit and will report back.

[DrewTube]

Is K reg really .008V? Should be more like .8V, close to 1V anyway.  I would suggest letting the tubes burn in for several hours, maybe overnight, and checking the voltage again.

OK...after letting the tubes burn in for ~6 hours, I am getting some different readings.  Any thoughts or reaction to these values?  I'll continue burning it in for a few more days and see if things continue to change or stabilize.

I also noticed that the transformer gets uncomfortably hot during operation (uncomfortable after touching for 1-2 seconds, but not hot enough to burn).  Is that normal?


IA -    Right = 223.2 - Left = 224.0
Kreg - Right = .008   - Left = .008
breg - Right = 92.4   - Left = 72.1
OA -   Right = 169.7 - Left = 197.6
bA -   Right = 0       - Left = 0
OB -   Right = 90.3   - Left = 68.7

[DrewTube]

OK...so after another ~5 hours of burn-in, I turned on my power amp and took a listen.  The basic circuit seems to be working, since I'm getting music, and it sounds good!  But, the imaging is way off-center (pulled to about 3 o'clock on the side with the proper voltage, other side has much less gain).  When I swap the EF86's (using the stock EH), the voltages and the off-center imaging follow the tube and switch to the other side.

Anyway, I bought a pair of NOS Telefunken EF86's, so I put those in...and oddly, it went whacky.  The B+ voltage still reads correctly (~225 VDC), but the voltage at the plates is about half what it should be (~50 VDC vs. ~100 VDC).  When I spin a disc through it, I get a very quiet, heavily distorted warbling sound out of the speakers...you can tell it's the song I'm playing, but it's horrible.

http://www.upscaleaudio.com/Made-In-Europe-EF866267_p_97.html

Any advice or ideas?

I'm also still curious about how hot the exterior side of the transformer should be getting.  Thanks all!

[Doc B.]

We are happy to send a replacement tube, just call Eileen - 360-697-1936 or email her at queen at bottlehead dot com. Not sure why the other tubes are running at such a low plate voltage, as the servo circuit is designed to adjust for different tubes so the new ones should hit the proper voltages. Perhaps they need some burn in as well. And yes, the power transformer will be very warm to the touch after a while.

[Grainger49]

Your problems seem to be changed when you change the tubes.  And I'm with Doc, the servo circuit ought to bring the voltage up where it should be.

Another poster told about the secret tube, the USSR Surplus Military 6J32P.  They can vary, some might not be good enough.  But they can be had for $5-6 each delivered if you buy a few.  And at the price I have bought quite a few.  The last seller threw in 2 extras, I bought 4 got 6.  

Paully's transformer was quite warm but I didn't put my hand on it.  

[Paul Joppa]

Late entry here, I've been tied up a lot lately.

I don't think the servo circuit is capable of pulling the cathode down to 0.008 volts. Something is probably wrong on the board that hosts the servo, or the cathode is otherwise grounded and bypasses the servo altogether.

Measure voltages at the 27K 1W resistor ("R4" on the B end) and the 174 ohm resistor (R3, nearby) for starters. Make sure of the orientation and leads of the 2N2222A, installed in the "431" holes - and check to be sure the transistor can is not shorting to any of the PC board pads, and there are no solder bridges accidentally shorting those pads. Check orientation of the 2700uF/4v capacitor as well. This is most of the servo circuit. Also inspect the EF86 sockets pin 3 to be sure it's not shorted to an adjacent pin.

[DrewTube]

[...]
Measure voltages at the 27K 1W resistor ("R4" on the B end) and the 174 ohm resistor (R3, nearby) for starters.
[...]

OK, I think I've solved the mystery.  It looks the R3 resistors I got are 174 Kohms, not 174 ohms!  I didn't realize because when I tested with my DMM, it read "174"...I didn't notice the "K!"  I guess that would explain the 8mV instead of 1V and would also explain the failure of the circuit to adapt to my NOS Telefunkens, right?  I'll contact Eileen for a new pair of resistors.

Thanks Paul J., Doc, and Grainger.

[DrewTube]

It looks like the R3 resistors I got are 174 Kohms, not 174 ohms!

Resistors are en route.  Paul, Doc, et al., - what are the chances this might have damaged any of the tubes or other components?  Anything specific I should check for?  Shucks...now I'm paranoid.

[Paul Joppa]

When you first posted the 174K problem, I looked up the 2N2222A data sheet. I believe the 27K resistor (and the C4S feeding the tube) will limit the base current into the transistor to an acceptable value and should cause no harm.

If I'm wrong, I'll suggest to Eileen that she send a full set of parts to replace the board - but at this point I expect the correct resistor to solve the problem, and I don't see any mechanism that is likely to harm any other components.

This made me think a little and correct a prejudice I've had for a long time. We sneer at transistors and call them "three-legged fuses" because they are excessively sensitive to voltage surges or spikes. But in fact they are pretty tolerant of current surges or spikes. Tubes, however, are quite tolerant of voltage spikes but can be quickly damaged by excessive current. It's just a matter of maintaining an appropriate perspective.

[DrewTube]

Thanks, Paul.  I received the new 174 ohm resistors and put them in.  All the resistances and voltages now check out well within specs - kreg is nailed at 0.98 volts, OA = ~160 vDC, and OB = ~96 vDC on both channels.  After 5-6 hours with the stock EH tubes, all the voltages were still stable and it sounds very good.

I popped in my NOS EF86 Telefunkens and now they work perfectly!  With the NOS tubes, this thing sounds amazing...knocks the socks off my Seduction.  I'm looking forward to hearing what this sounds like after some more burn-in.

I did notice that OA and OB are well within spec with the Telefunkens, but the kreg is ~1.3 vDC...I assume that's OK and not unusual with a non-stock tube?

Thanks for all the help and another killer preamp design.

[Paul Joppa]

...
I did notice that OA and OB are well within spec with the Telefunkens, but the kreg is ~1.3 vDC...I assume that's OK and not unusual with a non-stock tube? ...

You got it. That's the servo bias in action - kreg is the cathode voltage (bias) of the EF86, which the servo circuit adjusts to obtain the right plate voltage. As you can readily imagine, if you use a fixed voltage there (which we do in the big Repro amps) you have to be darn picky about getting an input tube that works right in the circuit.