Dynaco ST-35 voltage readings

[paart]

Stunch, I'm a little confused by your statement concerning the current measurements. Where, and how are you are getting these readings? The output stages should not be drawing the amount of current you mention. If they are, you are in the danger zone! Way too much current if that's correct, but I suspect it isn't. Your voltages across the 5 ohm resistors is fine, if you are measuring it correctly.

The current of 27ma (per pair) is correctly set by measuring .27 VDC, drop across each 5 ohm cathode resistor. If you are using the single EFB bias circuit, then the potentiometer is adjusted to obtain the reading giving the closest average to .27VDC on each 5 ohm resistor. If you have the separate (individual channel adjustment) EFB circuit, then each channel can be adjusted to .27VDC individually.

To do this, the meter is set on the DC voltage scale. If you do not have an "autoranging" meter, you probably need to use the lowest, or next to lowest VDC range. One lead is placed at each end of the 5ohm resistor. The red lead should be placed on the end of the resistor that's closest to the cathodes of the output tubes. Otherwise the meter will read a "minus" voltage. This is okay though. Just ignore the minus sign on the display.

Current can be measured directly by your meter. But the procedure is different. Current is measured by a voltage drop across a known value resistor. In the case above, the resistor is already in the circuit (5 ohms). If you wish to measure current draw in another part of the circuit, then you can use the current scale on your meter, which uses a resistor that's already built in the meter. The leads are applied differently, however. The black lead is attached to one side of the circuit, usually ground, and the other lead is placed on the area of the circuit to be measured. This sometimes involves temporarily "opening" the circuit, so that the current is drawn through the meter, rather than through another part of the circuit. It's trickier than measuring voltage, and runs the risk of blowing the meter fuse or internal measuring resistor, if the current draw is excessive.

You need to adjust the current of the output tubes, using the procedure, above, then let us know about the voltages of the new 7247 tubes. That was the original concern. The minor reduction in AC line voltage that you mention, should have only a very slight effect on the output tube bias, and probably on the 7247 plate, grid and cathode voltages.

[Stunch]

paart, I set bias the way it was described in the EFB article (which is exactly how you explained it). As mentioned I was able to get bias between channels pretty close. I have a set of test points for each channel using the single EFB circuit as per the article. In my case I mounted the resistors on terminal strips on either side of the chassis on the outside of the amp. Bias adjustment is also on the outside, making it easy to make adjustments. Some day I'll show some pictures.

Readings today were vastly improved. I added the new resistor and new closely matched 12DW7's. Wall voltage was 119.6VAC. Here are the readings:

12DW7

PIN 1 - 214(L) 219(R) 215-230 spec
PIN 2 - 117(L) 110(R) 112 spec
PIN 3 - 130(L) 123(R) 116 spec
PIN 4 - 5.8(L) 5.8(R) 6.3 spec
PIN 5 - same
PIN 6 - 117(L) 110(R) 112 spec
PIN 7 - 0 0 0
PIN 8 - 0.83(L) 0.86(R) 0.85 spec
PIN 9 - see 4&5

Cathode pin 3 appears hot in bot channels. Cathode resistors (R7) are both right on spec. Not sure what's happening here.

On the output tubes, I'm getting better readings today as well. 0 VDC at PINS 1,2,6 and 8 as they should be //// 13.75 VDC (13.5 spec) at PINS 3 //// 5.8VAC (6.3 spec) at PINS 4 and 5 //// 383VDC (370 spec) at PINS 7 //// 387VDC (380 spec) at PINS 9

The quad cap readings are also better: 1st stage - 396 (385 spec) //// B+ stage - 390 (380 spec) //// 20 uF stage - 325 (320 spec)

Thanks for wading through all this stuff. Should I just ignore the inflated cathode readings in the phase inverter stage, or delve into it? They were also high with the other set of input tubes.

[paart]

The output tube voltages look fine. I think it might be a little odd, however, that the heater voltages are on the low side. Do you have any resistors installed in the heater circuit? What brand of power transformer are you using?

The issue with the 7247 tubes isn't the cathode voltage as much as it is with the difference between the cathode voltage at pin 3 and the grid voltage at pin 2. Ideally, pin 2 voltage should be 4 volts lower than pin 3. This establishes the grid bias for the phase inverter. You have a difference of 13 volts on both channels.
Do you know what the wattage ratings are for the R6 and R7 resistors? I'd check the values of each of these resistors in both channels.
You might consider swapping the two 7247 tubes between channels and re-measuring the voltages, and if you want to play around some, try reinstalling your old 7247s, and see what readings you now get with those tubes.
What brand of meter are you using to check these voltages?

[Stunch]

Both R6 are 26.9K (27K spec) and R7 are 33.0K and 32.9K (33K spec). All resistors in the amp measure good. R5 measures 300K for example.

No resistors in the heater circuit as none were called for. I was getting 6.0 volts the other day. 5.8V is within tolerance, but a little low. Power transformer is a PA774 from Dynakit. Everything in the amp is new. Output transformers are from Magnequest. Resistors are Takman and PRP.

The meter I'm using is a Mastercraft. I'll swap tubes around tomorrow and see what's going on.

[Stunch]

Well, the PIN issue followed the tubes, so 110 on PIN2 (L) and 118 on PIN2 (R); and 124 on PIN3 (L) and 132 on PIN3 (R). All other readings were pretty much the same.

I took out the 12DW7's (EH) and installed the JJ's. PIN1 readings were identical at 223VDC, but there were a few changes at PINS 2&3.

PIN2 (L) 110 (same as EH) ///// PIN2 (R) 111 (down 7V from EH)

PIN3 (L) 125 (up 1V) /////// PIN3 (R) 126 (down 6V from EH)

Readings at PIN2 were very close to the 112 spec, but the PIN3 readings are high once again. 13 higher last night; 14 higher an hour ago and now 15 higher with the JJ tubes. But it's consistent channel to channel - always the same voltage difference between channels.

What bothers me is that in the original ST-35 manual it states: " If readings made at any of the points indicated differ from those listed by more than 10%, it is likely that a wiring error or defective component is nearby in the circuit." There is no mention of the PINs 2&3 needing to be 4V apart. You would think that there would be mention of this, but there isn't. Accordingly, the grid could be as low as 101VDC and the cathode as high as 127VDC!

I wonder how many functioning ST-35's are out there that aren't even close to spec. Should I consider adding more resistance to drop the voltage at the cathode? Is it possible that all modern 12DW7's aren't good for this circuit? I can't seem to source any NOS.

[Ty_Bower]

I wonder how many functioning ST-35's are out there that aren't even close to spec.

Without a doubt, tons of them.

Is it possible that all modern 12DW7's aren't good for this circuit? I can't seem to source any NOS.

Possible, but let's consider this: They can build an acceptable replica of the 12AX7. They can build an acceptable replica of the 12AU7. How hard can it be to stuff half of each together inside the same bottle?

Should I consider adding more resistance to drop the voltage at the cathode?

I'd first buy another cheap DVM from Harbor Freight, and recheck the voltages again. It's possible your meter is loading down the circuit and throwing off the readings.

[paart]

Quote:
Is it possible that all modern 12DW7's aren't good for this circuit? I can't seem to source any NOS.

Possible, but let's consider this: They can build an acceptable replica of the 12AX7. They can build an acceptable replica of the 12AU7. How hard can it be to stuff half of each together inside the same bottle?

Probably "harder than you think"!
I have 3-4 examples of "gen-yoo-wine" RCA 7247 tubes, that are "anything but" 7247s. I purchased two of them from the RCA distributor at the end of the era--in the mid 1980s, and they are certainly not 7247's! I have one or two other used ones that I picked up somewhere along the line, and these are the same. The two sections are almost perfectly matched, and my TV-7 "pegs" the pointer at the end of the scale on both sections, when the tester is adjusted for the 7247 tube. I assume that they are 12AT7's, which will work as a substitute, but obviously not as well. Factory mistake in labeling? Personally, I very much doubt it. I believe the substitution was intentional for inventory reasons.

Quote:
Should I consider adding more resistance to drop the voltage at the cathode?

NO!
Stunch-I am waiting to see your results from swapping tubes.
I really think, however, that the issue may be with your meter, since I had that exact problem, and it caused me considerable distress for a long time. Do you have access to (beg, borrow, steal etc.), a better meter?
As an example: Below are the results that were observed from the same 7247 in the same circuit, running at the same line voltage with two different
meters:

Meter #1
Pin: 1 (211V)
Pin: 2 (115V)
Pin: 3 (129V)

Meter #2
Pin: 1 (218V)
Pin: 2 (125V)
Pin: 3 (128V)

The first meter was a "cheap", but assumed to be a
"pretty good" meter.
The second is my Extech meter.
Both of these meters "check out" reasonably well against my Fluke bench meter on routine voltage and resistance checks.
The difference has to be the input impedance of the meters, which isn't of great concern in routine measurements, but becomes critical when measuring high impedance, low current circuits.

TY-check your "PM" on this forum.

[paart]

Stunch- Sorry, I didn't read your complete post from yesterday! The voltage readings that you observed lead me to think that there are no wiring issues in your amplifier. Those variances are the same as I've seen by changing the tubes around.
Dynaco did not specifically specify a difference of 4 VDC between the grid and cathode of the phase inverter, using that specific statement, however if you look at the voltage chart, the difference is 4 volts. A 10% voltage variation, however, would not mean that, say, the plate is 10% high and the cathode is 10% low, as that would be a 20% variation.
As Ty pointed out, once you get too far away from that 4 volt negative bias on the grid of the phase inverter, the desired operating point of the tube starts to change adversely. If you hit that 4 volts, "dead on", then you are lucky. But my amplifiers usually come pretty close; within one volt most of the time, two volts at worst.
Again, I really suspect the meter that you are using is the culprit.

[Stunch]

I'll see if I can source a better multimeter. It's odd though that the readings are consistently high at the cathodes and only there. You wouldn't think a meter would screw up for just that particular reading and be okay at other parts of the circuit. Both JJ and EH tubes read high and in proportion to the grid. I'd really like to get an older 12DW7 and see what's going on. Jim McShane doesn't have any and there are none at Parts Connexion or the TubeStore.

In the meantime, I'll look for another multimeter.

[paart]

I wouldn't worry too much about the cathode voltage, or even the plate voltage at the phase splitter. A lot of this is determined by how "hard" the tube conducts at a certain grid bias. (The "mu" of the tube). Different currents flowing through the cathode resistor will create different voltages at the cathode.
You can see this for yourself by substituting a different (pin for pin, and heater voltage compatible) tube such as a 12AT7, 12AZ7, 12AY7, then checking the cathode voltage. The resulting bias change from the first section of the tube and "mu" of the second section will vary the Pin 2 cathode voltage considerably.

The part that's somewhat amazing about all this, is that typically, the "sound" from the amplifier won't be affected nearly as much as one would suppose!

If you change the value of R7, you will certainly affect the balance of the phase inverter, and increase distortion in the process.

[Stunch]

Well, this may clear a few things up. I borrowed a Fluke 77 Series ll meter and took some more readings. Line voltage was 121.6V. The readings were more or less consistent with my meter EXCEPT for PIN 2 which read 122 with the Fluke and 111 with my meter. Spec is 112. PIN 3 was 126 with the Fluke and my meter as well. Cathode readings were identical as mentioned BUT grid was different by 11 volts! With the Fluke, the difference is now the desired 4 volts between grid and cathode (122V & 126V). Spec is 112V and 116V.

The other significant difference was heater voltage. My meter read 5.9 (which is pretty constant whenever I measure it), but with the Fluke it read 6.38. All other readings were higher than spec, but not abnormally so.

Just for fun - my life is pretty boring! - I hooked up a thermistor which knocked line voltage down to 117. Air conditioners in the area probably helped a bit too. Here are the readings taken with the Fluke meter. They seem to be really good.

Input tubes

PIN 1 - 221 (L) 221 (R) 215-230 spec
PIN 2 - 118 (L) 119 (R) 112 spec
PIN 3 - 122 (L) 123 (R) 116 spec
PIN 4 - 6.2 (L) 6.2 (R) 6.3 spec
PIN 5 - same
PIN 6 - see PIN 2
PIN 7 - 0
PIN 8 - 0.84 (L) 0.83 (R) 0.85 spec
PIN 9 - see 4 & 5

Output tubes

PIN 1 - 0
PIN 2 - 0
PIN 3 - 13.5 (L) 13.5 (R) 13.5 spec
PIN 4 - 6.2 (L) 6.2 (L) 6.3 spec
PIN 5 - same
PIN 6 - 0
PIN 7 - 376 (L) 376 (R) 370 spec
PIN 8 - tied to PIN 9 as per EFB
PIN 9 - 380 (L) 380 (R) 380 spec

Quad Cap

1st stage - 388 385 spec
B+ - 381 380 spec
20 uF - 321 320 spec

The readings for the grid and cathode were high in the phase splitter, but 4V apart. I'm guessing, just leave everything alone now and get on with my life?

[paart]

Great!
I had guessed that all was fine with your amplifier, but it was a matter of getting correct measurements. The change in AC heater voltage isn't too surprising either. AC is somewhat more difficult to measure than DC. Better meters usually produce more accurate AC results, also.

Now you can "sit back " and enjoy the music for awhile!

But if you find yourself getting too bored, perhaps my latest ST-35 project will provide some ideas to keep you busy------


http://www.tronola.com/html/ps_mods_for_st-35.html