DC Voltages look suspicious

Hi!

I just finished building my ST-70 out of an old unbuilt kit and the low-gain VTA board. Bias adjustments went fine as well as the AC balance. I fed the amp with a sine wave generator and, on my scope, everything looked fine.

I gave the amp a music test and the sound seems very good. There is, however a problem that bugs me: the voltages do not come at all to the same values as shown on the schematic:

• B+ to the center-tap of the output transformers' primaries is 450V which is fine.

• The B+ going to the VTA board is 412V which is 20 v above the specified 390V on the schematic but still acceptable.

• Between R13 and R11, I measure 320V against the 310V; again very satisfactory.

• It is when I measure the voltages on V1 plates that I see a huge difference: 227V (pin1) and 224V(pin 6) when the schematic specifies 150V !

• Also the schematic shows 100V DC at the junction of R27 and R17 and I measure 68V !

• The phase splitter tubes show 272V on their plates against 200V on the schematic !

• Bias voltages are normal at -38V.

Is it possible that the tubes supplied with the VTA board are out of spec? They are ECC81 from JJ Electronic? Those measures are identical on both channels. Could someone help me here with a few suggestions?

Yvon

Kudos for finding an "unbuilt" ST-70

Hi Bob,

Thank you for your suggestions.

My Dynaco is indeed a 1967-68 unit. I will try with a 5U4 tube and, if necessary after that, I will raise the value of the voltage dropping resistor on the Quad cap.

I well remember the 5U4. It was the rectifier tube in our old Admiral TV set in the fifties. It used to burn quite often! A quick look on the Net shows a fair choice from quite inexpensive ones to 130$ ! Is there a brand you would recommend?

Finally, you mentioned that 450V on the primary center tap is also a bit high. What would be the ideal value?

Yvon

Bob,

I have a very powerful variac. I decided to use it to check what you're saying about old Dynaco power transformers. I plugged my ST-70 through the variac set at 110V AC.

Indeed, as you predicted, the GZ34 pin 8 came out at 434V, the B+ at the primary center tap at 422V and the B+ to the PCB at 391V. Wonderful!

Alas! The voltage at V1 plates was still 204V. This is way higher that the schematic's 150V.

Then, I noticed that the schematic puts R13 at 15K whereas the part supplied by Roy in my VTA kit was 33K which is double the specified value. As the voltage drop that I measure across R13 is 94V, the current value is around 2.8mA. The schematic puts that value at 5mA.
The same 2.8mA flows across the load resistor R11 which brings the voltage at V1's plate at 203V, (actual measured value 204V).

So my measured voltages seem right. Is there a specific reason why Roy changed the value of R13 to 33K or should I put back the schematic's 15K? I will try to contact him on this matter. A static plate current of 3mA is quite different from one of 5 mA, so I will wait for his opinion on this.

In any case, I feel it is important that PROPER VOLTAGES be given on the schematic. It is an essential debugging tool.

Best regards,

Yvon

two things about this . . . 1- we found out that somehow I had accidently supplied Yvon with the wrong tubes, he got the hi-gain tubes with the lo-gain kit. 2- after testing dozens of these boards that I've built, the voltages seem to be all over the place (which doesn't make sense)
but they have all sounded fine. We know that +/-10% nominal voltage variations are normal and acceptable, most of this variation is due to variations in the tubes and different brands etc.
Maybe we can get more of a consensus here on plate voltages (pin 1 and 6) of the 3 tubes for the low-gain version (non-CCS).
My various versions of schematics thru the years have listed 150-200v on the center tube, and 200-300 volts on the outer tubes.
My current ST70 is in between boards again (lo-gain 9 pin, CCS, and octal versions) so until it's back together let's hear from some other users.

Roy is right here:

Normally, there should be around + or - 10% variations on measured voltages, especially since the passive components on the VTA all have tight tolerances and those voltages are all measured under static DC conditions where the only effect of the signal capacitors is to block any DC path!

Thus the reason for all these variations can only be attributed to diverging DC plate resistances between tubes. Tolerances between units even of the same type and brand seem to be important and it would not surprise me that what is identified as 12AU7, when measured, has little to do with the published data specs.

I'm wondering if the voltages found on the schematic diagram are actual measured voltages or calculated ones based on the data sheets. Roy could help us on this.

It is possible to actually calculate the voltages by using the data sheets and curves for the 12AU7 and see how close one gets to those numbers with real tubes.

I'm tempted to do it!

it's probably been 5 years since I measured actual voltages on the ST70 board. After building thousands of them, it got to the point of why bother? But in the interest of keeping things accurate, let's get some input from customers with the latest version of the board,
feel free to post plate voltages here on this thread (pin 1 and pin 6 on each tube, center, left, and right).
Meanwhile I'll try to get my own amp back in one piece, it's constantly changing from the old version to the new version to the octal version,
to help customers with boards that aren't quite right due to wrong placement of parts or poor soldering.

Hi,

I did my homework...

A original ST-70 that I restored with Roy's VTA board with the 12AU7. The new version.

The amp has the original ST-70 power transformer, and the main here is 117 volts today. This amp sounds great, rock solid Bias.

3 JJ 12AU7
4 JJ 6l6GC
Ruby GZ-34

So here goes:

Center :
Pin 1 :122V
Pin 6 :129V

Left :
Pin 1 :216V
Pin 6 :211V

Right :
Pin 1 :213V
Pin 6 :211V

But I completly forgot to mesure the B+ Ooops!

Anyway, that's what I have on this amp

Cheers

thanks ixe13
is that the latest version with the CCS or is that the older version?

Hi Roy,

CCS, I believe, since it has the 2 LM334. Right?

yes, that's right, so instead of me installing that in my ST70 this weekend, I'll plug in the octal version, which should have identical voltages, but I'll find out for sure!

Hello Yvon,

I have always had similar issue with my low gain (non-CCS) VTA board, in that the voltages didn't match several schematics I was sent, or found here. The B+ was within the 410 to 430 volts going to the board, which one schematic indicates, but the board voltages didn't match. I re-checked all resistor values, and they were within tolerance.

Several inquires regarding the "correct" voltages didn't give me an answer I felt comfortable with. Hopefully bringing the topic up here will lead us to the optimum working voltages for this circuit.

I will dig through my notes, or hook up the amp and re-measure the voltages and post.

Thanks for a great topic!

I took some measurements on an octal version VTA70 the other day, and they should be about the same as the latest "classic" version with CCS
that I've been selling now for about a year.
With 350vdc B+ on the board (it was 400v to the OPTs) I got 120vdc on the plates of V1 (center tube) and 200vdc on the plates of V2 and V3 (the outer tubes). All about 5-10% less than what I show on the schematic that comes with the board kit. That was with 120vac input.
Pretty close to what IXE13 was getting.

I finally received the JJ ECC82s from Roy. (Let me simply mention here that I had been sent the wrong tubes for my VTA board and this mistake was certainly the main reason for the huge discrepancies in measured versus published voltages that I had encountered).

Waiting for the JJ's ECC82s to arrive, I bought locally 3 NOS 12AU7s branded "Canadian Marconi Radiotron" made in the 1950s. I decided then to measure voltages with the JJ tubes first and then the Marconi ones and to compare the numbers with calculated values using the tube data sheets. Here are the results:



- AC voltage input was set at 115VAC by a Variac to take into account that my amp is using an old Dynaco power transformer, designed for 115VAC instead of today's 120VAC.

Point of Measurement	Marconi	JJ	Calculated
V-pin 8 rect.	442V	442V	-
V-choke out	431V	431V	-
V- VTA board	380V	383V	-

Point of measurement	Marconi-Left Channel	Marconi-Right Channel	JJ-Left Channel	JJ-Right Channel	Calculated
V - junction R11-13 and R12-R14	258V	256V	256V	258V	260V
V1(input tube) plate	130V	132V	130V	136V	128V
V1(input tube) cathode	4.2V	4.24V	4.26V	4.16V	4.4V
V2(PS) plate	210V	211V	206V	208V	-
V3(PS) plate	204V	205V	218V	204V	-
V2(PS)-cathode	94V	93V	91.5V	94.5V	-
V3(PS) cathode	94V	93V	91.5V	94.5V	-
R17-R18 tail resistor	90.4V	89.7V	91V	88V	-

As can be seen here, the Marconis and JJs give essentially the same voltages and for the input tube V1 the values come very close to the calculated ones. In fact, the agreement is so close that I decided not to go into the calculations (I don’t use computer programs for these, so they are quite long to make) for the phase splitter tube and take the values as the proper ones.

Comparing these measurements with the values given by Roy in his latest schematic sent to me, the agreement is acceptable and the outstanding differences may stem from the 400V that he puts for the B+ to the board against the 380V that I measured.

Yvon

alright, nice work, now there is very definitive voltage checks on this board, and glad to see they match what I got a year or so ago.
I tested another amp recently, and I'll probably wind up changing that 2.2K 3W dropping resistor on the quad cap to a 1.8K 3w to give slightly more voltage (closer to 400v) the the board, but it works fine even at 350v.