Newb question: which impedance 5k or 8k

[chikov]

I have a simple question that puzzled me and after doing some research I am still not sure. I build a champ AA764 amp. I finally got a hold of a decent Hammond output transformer 1760CP, but it has two primary impedances 5000 and 8000. Which one should I use and why would they want to put two of them if it’s made for a champ as a replacement?

[Stevem]

That's because of different power supply voltages ( you need to learn to look at tube spec charts ) between the Tweed champs and the black and silver face models.
If your building one of the later two models the use the 8k tap to go to pin 3 of the 6V6 output.

[blackeye]

I’d go with 8,000. You’ll get closer to 5W that way. Many report a sweeter sound with the higher impedance. What speaker impedance do you plan?

I don’t cut the 5,000-ohm lead, unless I’m absolutely sure I won’t use it. If I’m saving a transformer lead, I like to snip off the exposed conductor and then seal it off with some heat shrink tubing.

Maybe you will get curious and try out the 5,000-ohm lead to hear how it sounds? It won’t hurt the amp. Your output tubes will run a little hotter and you’ll lose some clean headroom.

[chikov]

Thank you blackeye. Yes that’s what I was not sure about - what difference does it make. I do use 8ohm speaker And my voltage is at about 390 V DC. I’m thinking about installing a toggle switch so it would become an option and I will decide as I go by listening. Thank you very much for your answers - I hope they will have others as they helped me.

[Colossal]

Thank you blackeye. Yes that’s what I was not sure about - what difference does it make. I do use 8ohm speaker And my voltage is at about 390 V DC. I’m thinking about installing a toggle switch so it would become an option and I will decide as I go by listening. Thank you very much for your answers - I hope they will have others as they helped me.

390VDC is high, but not unreasonable if the screen voltage is managed.

For determining single ended primary output impedance, I use this method to find a center-balanced loadline:

Primary Impedance = Plate Voltage^2 / Max Plate Dissipation

ex. 6V6GT power tube
Chikov's plate voltage = 390VDC
6V6GT's max dissipation = 12W

(390VDC)^2 / 12W
152,100 / 12
12,675

The center-balanced loadline for your condition would be a primary of 12.7k. However, the loadline is more of an ellipse under dynamic conditions, working into a reactive (speaker) load, so 8k is possibly reasonable, but at 390V, I would give a substantial drop to the screen voltage (using a fairly large dropping resistance) and individual screen resistors.

To find the optimal plate voltage for your 5k or 8k taps...solve for the plate voltage:

What is the optimal voltage for an 8k primary and a 6V6GT?
8000 = x^2 / 12
8000 * 12 = x^2
x = SQRT(8000 * 12)
x = 309VDC

[blackeye]

Thank you blackeye. Yes that’s what I was not sure about - what difference does it make. I do use 8ohm speaker And my voltage is at about 390 V DC. I’m thinking about installing a toggle switch so it would become an option and I will decide as I go by listening. Thank you very much for your answers - I hope they will have others as they helped me.

Where and how are you measuring 390VDC?

If that’s with no power tube installed, you can expect a drop in volts due to resistance in the rectifier when you install the power tube.

Also, once the power tube is installed don’t forget to subtract the cathode voltage as the datasheet reports measured voltage between plate-cathode, not plate to ground.

[roberto]

It is really supplied at 390V? I'd drop at least 50V to that with 8k Ra and again at least 100V with 5k Ra.
390V means something like 370V across the tube (15 across the cathode resistor and 5 across the primary), something like 30 mA is a reasonable current.
With 8k Ra you have 3.6W before current goes to zero, with 5k Ra you have 2.2W before current goes to zero. In both cases g1=0 brings the plate very very asymmetrical.

[roberto]

What is the optimal voltage for an 8k primary and a 6V6GT?
8000 = x^2 / 12
8000 * 12 = x^2
x = SQRT(8000 * 12)
x = 309VDC

This is a first approximation of a 8k Ra loadline tangent to the 12W max plate dissipation, and then finds the average point between Ia=0 and Va=0.
In reality it is not suggested to work at 100% plate dissipation, plus the plate (even with tubes designed for A2 class like the 6V6GT) stops around 50V (or even before, depending on various aspects).
This, together with some chase in specific harmonic content, moves the working point around that value (usually down-right it, looking at the curves), but it is for sure a good way to reach the ballpark.

Play with screen supply (both statically and dinamically) to find the right responsivness you are looking for.

[chikov]

Thank you for in-depth replies. I measured my voltage between the plate and the cathode (pin 3 and pin eight) this time and got 352Vdc. I used milliamp tester in line with my tube and got 37 mA of the plate current. In order to calculate plate dissipation I have to multiply 352 by .037 and I got about 13W of PD. It’s a little high, but when I flip my toggle switch from 8000 to 5000 ohms the voltage goes up to 356Vdc. I am sure that part of the problem is my transformer that I had made here in Russia and that’s the round air core one (toroidal). I think the voltage is a little too high on this one. I installed 5 W potentiometer on the Cathode to the ground Bias control. But every time amperage lowers the voltage goes up and if I turn it the other direction voltage goes down but amperage gets even worse. I’ll keep it, I guess, at 37 mA and 350 V DC plate voltage, And keep my primary on OT at 8000. This is my small rehearsal amp so I can play it quietly at my apartment , and I think it should be OK even at 13W of PD. I heard that it is possible to lower B plus voltage on a transformer but I would have to install some big , high wattage resistor, right?

[martin manning]

I am sure that part of the problem is my transformer that I had made here in Russia and that’s the round air core one (toroidal). I think the voltage is a little too high on this one.

It does have an air-filled hole in the middle, but the toroidal core is steel.

I installed 5 W potentiometer on the Cathode to the ground Bias control. But every time amperage lowers the voltage goes up and if I turn it the other direction voltage goes down but amperage gets even worse. I’ll keep it, I guess, at 37 mA and 350 V DC plate voltage, And keep my primary on OT at 8000. This is my small rehearsal amp so I can play it quietly at my apartment , and I think it should be OK even at 13W of PD.

You could use a JJ 6V6S, which is a 14W tube. What is your screen voltage? I would adjust the dropping resistors in the power supply to get the screen voltage down to around 310 (that will lower the anode current, so you will have to adjust the bias), and I would definitely stick to the 8k primary. See https://www.thetubestore.com/lib/thetub ... -Rev-B.pdf

I heard that it is possible to lower B plus voltage on a transformer but I would have to install some big , high wattage resistor, right?

You can do that with a MOSFET in the ground return to the PT's CT, but I think you can make it work as is.

[blackeye]

So you went with a ground bias on the cathode instead of the cathode resistor/bypass cap shown in the AA764 schematic? Interesting! But maybe adding extra complexity that is causing trouble?

[blackeye]

I installed 5 W potentiometer on the Cathode to the ground Bias control. But every time amperage lowers the voltage goes up and if I turn it the other direction voltage goes down but amperage gets even worse.

I'd be really interested to know if there was a bias circuit that would solve this problem without the use of transistors. I'm also working on a design that would utilize the bias winding (50vac) to raise/lower the ground reference for some cathode followers and power tubes.

[drew]

I heard that it is possible to lower B plus voltage on a transformer but I would have to install some big , high wattage resistor, right?

You can do that with a MOSFET in the ground return to the PT's CT, but I think you can make it work as is.

MOSFET B+ reducer thread: https://ampgarage.com/forum/viewtopic.php?f=6&t=27035