Are we talking about the same tubes and pins? I'm referring to pin 7 on the 12AX7s immediately before the output tubes. I get 19.5VDC and 17.7VDC on those pins. I do see 68VDC on pin 8 of those same tubes (the schematic can be a little hard to read).
I'm trying to figure out why I see high grid voltages on pin 7 of those tubes and pins 2 and 7 of the cathode follower immediately after the phono section.
You might also look at the connection to the tape record jack on channel 1 (pin 8 of the cathode follower). Somehow, that doesn't look right to me.
The grid (pin 7) and cathode (pin 8) voltages of the 12AX7 will be very close, within a volt or so of each other.
The elevated grid DC voltage is due to the self-bias network for the 12AX7, and is correct.
All cathodyne phase splitters will have an elevated grid voltage unless the cathode resistor goes to
a negative voltage instead of ground.
The 6EU7 cathode follower after the phono stage is also self-biased in the same way, and should have
about 48VDC on its cathode.
The elevated grid DC voltage is due to the self-bias network for the 12AX7, and is correct.
All cathodyne phase splitters will have an elevated grid voltage unless the cathode resistor goes to
a negative voltage instead of ground.
The 6EU7 cathode follower after the phono stage is also self-biased in the same way, and should have
about 48VDC on its cathode.
Last edited:
Thanks for the info. I figured that there was some reason that these (and only these) grids showed elevated voltages.
However, the disparity is far greater than 1V. For the cathode follower, I see pin 2: 27VDC; pin 3: 50 VDC; pin 7: 34VDC; pin 8: 75VDC. For the phase inverters, I see pin 7: 17.7VDC; pin 8: 59VDC on channel 1 and pin 7: 19.5VDC; pin 8: 58.2VDC on channel 2.
And a clarification: when I wrote "I do see 68VDC on pin 8" in the previous post, I meant that I see 68V on the schematic (rather than 58V). The measured values are as above (59VDC and 58.2VDC).
However, the disparity is far greater than 1V. For the cathode follower, I see pin 2: 27VDC; pin 3: 50 VDC; pin 7: 34VDC; pin 8: 75VDC. For the phase inverters, I see pin 7: 17.7VDC; pin 8: 59VDC on channel 1 and pin 7: 19.5VDC; pin 8: 58.2VDC on channel 2.
And a clarification: when I wrote "I do see 68VDC on pin 8" in the previous post, I meant that I see 68V on the schematic (rather than 58V). The measured values are as above (59VDC and 58.2VDC).
Last edited:
Strange voltage readings like that can be caused by the loading that the voltmeter makes on high impedance circuits, like those around 12AX7's. What kind of volmeter are you using?
If you have two voltmeters, you can confirm whether or not this is the case by fixing one voltmeter from cathode to ground, and observing changes when connecting and disconnecting the other voltmeter to grid.
It's common to measure voltages around valves at the anode and cathodes only, and to *infer* grid voltage from those.
YOS,
Chris
If you have two voltmeters, you can confirm whether or not this is the case by fixing one voltmeter from cathode to ground, and observing changes when connecting and disconnecting the other voltmeter to grid.
It's common to measure voltages around valves at the anode and cathodes only, and to *infer* grid voltage from those.
YOS,
Chris
An auto-ranging DVM like yours is typically 10M Ohms load on the tested circuit, so, not terrible and not the sole cause of strange readings. But, those are definitely strange readings, and not to be believed, or worried about.
If your anode and cathode readings (both less susceptible to meter loading errors) look reasonable, call it and move on.
Type 12AX7 valves have such a low cathode work factor that a well made one will run pretty much forever, although some can develop 1/f noise issues that matter for phono stages. I certainly wouldn't change any without good reason. And I wouldn't believe a random valve tester (very sensitive to inaccurate calibration at these low grid bias voltages) over performance in-circuit.
YOS,
Chris
If your anode and cathode readings (both less susceptible to meter loading errors) look reasonable, call it and move on.
Type 12AX7 valves have such a low cathode work factor that a well made one will run pretty much forever, although some can develop 1/f noise issues that matter for phono stages. I certainly wouldn't change any without good reason. And I wouldn't believe a random valve tester (very sensitive to inaccurate calibration at these low grid bias voltages) over performance in-circuit.
YOS,
Chris
Sounds like the DVM is loading too much when measuring the grid voltage to ground.
There's a 1M resistor there.
Thanks for the pointers.
Anode and cathode readings:
The plate voltage on pin 1 of the CF is a little puzzling. Pin 1 and pin 6 are fed by the same 282V source. Pin 6 connects straight to the source; there's a 47K (actual 51K) resistor between pin 1 and the source. The schematic says that pin 1 should see 180V; that seems a huge drop across that resistor. Since pin 3 is close to spec (50V vs. 48V), is it reasonable to speculate that might just be another error in the schematic?
It's also not clear if the 9-10V difference between spec and observed at the cathodes in the PI is anything to worry about. I'm kind of inclined to think that since both tubes exhibit similar behavior, that may be acceptable.
The auto-ranging feature on the meter was mentioned. That can be turned off and a specific scale (Ω, KΩ, MΩ) selected. Would it change anything to do that?
Anode and cathode readings:
Code:
spec actual
Cathode follower:
Pin 1: 180 217
Pin 3: 48 50
Pin 6: 282 282
Pin 8: 71 75
Phase inverter:
Pin 6: 304 306
Pin 8: 68 58
Pin 6: 304 306
Pin 8: 68 59The plate voltage on pin 1 of the CF is a little puzzling. Pin 1 and pin 6 are fed by the same 282V source. Pin 6 connects straight to the source; there's a 47K (actual 51K) resistor between pin 1 and the source. The schematic says that pin 1 should see 180V; that seems a huge drop across that resistor. Since pin 3 is close to spec (50V vs. 48V), is it reasonable to speculate that might just be another error in the schematic?
It's also not clear if the 9-10V difference between spec and observed at the cathodes in the PI is anything to worry about. I'm kind of inclined to think that since both tubes exhibit similar behavior, that may be acceptable.
The auto-ranging feature on the meter was mentioned. That can be turned off and a specific scale (Ω, KΩ, MΩ) selected. Would it change anything to do that?
You can not measure these grids directly!! Not with any meter in the house.
Cathode follower, Cathodyne, Fender Long-tail. If self-biased, touching the grid with any material meter throws the grid voltage WAY down. There is a lot of leverage here.
Here is a Sherwood (maybe not the right one) with "no" grid-loading (normal operation), a 10Meg load (you poking a VTVM at it), and an unlikely high 500Meg grid load (nearly electrometer territory). The plain VTVM pulls 49V down to 15V. This can be computed forward or back to find the no-load condition but it is a brain buster. You have to go to 500Meg to get a mild 5% error, and such meters have troubles. (We can do it inside sealed microphone heads, but dirt and moisture never sleep.)
On these stages, just check the cathode voltage! If that is nearly right (almost never critical) then the grid will be right when you are not poking it. So don't bother to poke it.
Attachments
The cathode followers are strange because one channel is actually a phase splitter. Back in those bad old days it was (at least thought) possible for one channel to be out of "phase" (actually, polarity) with the other. The switch to the right of the upper valve half (in my very, very poor schematic copy - sorry, can't read the number) does that. Some amplifiers had provision for flipping one speaker's polarity; this one does it in the preamp line stage.
Measurements within 20% are almost always plenty good enough. We normally don't know how the original measurement was taken, or even if it were taken - the number might easily be a design center - and our own measurements inherently change what's being measured.
YOS,
Chris
Measurements within 20% are almost always plenty good enough. We normally don't know how the original measurement was taken, or even if it were taken - the number might easily be a design center - and our own measurements inherently change what's being measured.
YOS,
Chris
I thought I'd attached a copy of the schematic earlier. I guess I didn't. Sorry about that. Maybe this version is more readable.
I thought the combined CF/phase splitter was a bit outside the box myself, particularly when I started looking at different resistors in the two sections (i.e. there is no plate resistor on channel 1, and the cathode load resistors are different). When I look at the readings, though, they're pretty close, so I'm guessing that these values were chosen so that both settings of the phase splitter switch would yield the same output from that triode (which would match the output of the other triode).
The strange thing about all of this is that I retired this receiver ~35 years ago because it kept blowing fuses, and all of this poking and measuring hasn't come up with any obvious failure point. I suppose that I should replace the filter caps, death caps, and add a polarized cord as a matter of course, but I still haven't got the a-ha! moment which would explain the fuses.
I thought the combined CF/phase splitter was a bit outside the box myself, particularly when I started looking at different resistors in the two sections (i.e. there is no plate resistor on channel 1, and the cathode load resistors are different). When I look at the readings, though, they're pretty close, so I'm guessing that these values were chosen so that both settings of the phase splitter switch would yield the same output from that triode (which would match the output of the other triode).
The strange thing about all of this is that I retired this receiver ~35 years ago because it kept blowing fuses, and all of this poking and measuring hasn't come up with any obvious failure point. I suppose that I should replace the filter caps, death caps, and add a polarized cord as a matter of course, but I still haven't got the a-ha! moment which would explain the fuses.
If it were mine, I'd also install 10R 1% resistors in each output valves' cathode leg (pin 3 to ground) to give myself a safe-ish place to measure idling current. And automatically replace the coupling caps feeding the output valves' grids. Anything to try to protect those rare valves. Nobody can really make them anymore.
All good fortune,
Chris
All good fortune,
Chris
I've been calculating the plate current on the output tubes by measuring the voltage drop across the output transformer (having measured the resistance of each winding previously). It's easy enough: clip one lead to the appropriate terminal on the filter cap (it feeds directly to the center tap of each transformer), then move the other sequentially down the row of tubes, reading pin 9 on each. Not only does this seem easier (only need to move one lead), but it measures only the plate current, while measuring at a cathode resistor would also include the screen current (negligible as it may be). Am I missing something here?
I've already scored replacement coupling caps (Cornell Dubilier 930C6S68K-F) for the output tubes. I've heard that the OEM caps rarely go bad, but I have no way of reliably testing ESR, leakage, etc., so it's probably a case of better safe than sorry. $12 for caps is a lot less than replacement tubes.
I've already scored replacement coupling caps (Cornell Dubilier 930C6S68K-F) for the output tubes. I've heard that the OEM caps rarely go bad, but I have no way of reliably testing ESR, leakage, etc., so it's probably a case of better safe than sorry. $12 for caps is a lot less than replacement tubes.
No, no, you're not missing anything. It's somewhat/slightly? safer to have the voltmeter leads near chassis voltage than at B+, but a careful person can get away with it. Many have over the years. You seem like a particularly thoughtful person, but when giving public advice, my strongly held belief is to be as conservative about safety issues as possible, just because.
YOS,
Chris
YOS,
Chris
I learned my B+ lesson years ago: I was sitting on a stool with a guitar in one hand, flicking things underneath the amp chassis to see if I could find out where the crackle was coming from. Next thing I knew, I was flat on my back in the middle of the room.
As they say, "Too soon old, too late smart."
As they say, "Too soon old, too late smart."
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.