Estey T12 guitar amp

[Chuck H]

I just did a test. I rigged that 200 ohm series resistor in the heater circuit so I could quick disconnect it while the amp is on. Quick disconnect of the heater AC would be a valid test to see if the heater AC is the problem. The heater would stay hot enough for that one second. So, doing the quick disconnect did change the "tone" of the hum, but the hum remained. How disappointing. SO, I continue to be at a loss here.

I put a like on this one because it's a smart test JSYK I'm not liking that you're still hunting for the problem

[johnhoef]

Chuck, thanks for the thumbs up on my heater test. Indeed, I'm not liking this hum problem either. This is such a simple circuit!! My next thought is to disconnect the entire power supply, and temporarily build an external one using full wave solid state rectification. Then feed the DC to the amp. This external supply would also feed DC to the heaters through a power resistor. Very curious to see the result of this test! This test should eliminate any AC coming from the power supply. Of course, I will report here the result of this test.

[johnhoef]

I ran the above mentioned test this morning with some encouraging results. Absolutely no buzz or hum what so ever!! I wired together a simple power supply using the isolation transformer, solid state bridge rectifier, and a 1500mfd filter. Using my variac on the isolation transformer primary, I used the variac to produce 130 volts DC output. I applied this DC to pin 7 of the rectifier tube socket. I pulled the rectifier tube. I supplied the other two tube heaters from this same DC through a power resistor. Indeed, this is a nice fix, and would be a permanent solution, however, guitar players like the "sag" that a tube rectifier produces. So, need to further investigate this. Good news is that the source of this buzz seems to be from the power supply and/or heater circuit. Think I'll replace the solid state bridge with a diode to produce the half wave rectification.....stay tuned.....

[Helmholtz]

... this is a nice fix, and would be a permanent solution, however, guitar players like the "sag" that a tube rectifier produces.

A single ended class A amplifier essentially doesn't produce sag as its current draw only slightly changes with signal.
In pure class A operation plate current stays constant at all signal levels, so B+ is constant.
There will be some compression by increasing screen current and sagging screen voltage at large output, but that only depends on the screen supply.

Do you need such a large filter cap? A 1500µ/200V cap should be quite big.

[johnhoef]

Helmholtz, I thought the same thing. I'm sure the guitar player would not like the idea of installing a solid state bridge, but you gave me a good idea. I will be running another test. This new test will not be electronics related at all, but rather, a test of the human mind. I will install the bridge temporarily without the guitar player's knowledge, return the amp to him, and wait for his opinion. He plans on using the amp at a gig this coming weekend. Of course, once he is truly convinced that the amp is operating to his expectations, I will reveal what I did. I'll be curious...stay tuned.....
By the way, that large filter cap was temporary...just for the test.

[Helmholtz]

I will install the bridge temporarily without the guitar player's knowledge, return the amp to him, and wait for his opinion.

That's exactly what I was going to suggest.
If you keep the tube rectifier in the heater string it will even glow as before.

[35L6]

Half wave rectifiers were made for AC/DC sets so you can leave the rectifier in . By the way my single ended amp sags under heavy overdrive . The dial light dims . I think it's the screen current . If it didn't sag it would probably burn the screen .

[Chuck H]

Half wave rectifiers were made for AC/DC sets so you can leave the rectifier in . By the way my single ended amp sags under heavy overdrive . The dial light dims . I think it's the screen current . If it didn't sag it would probably burn the screen .

I had a sort of similar/opposite issue once with a single ended build. The screen resistor was burning under clipping conditions and I couldn't understand why. I ended up increasing the screen resistor value and adding a diode clipper to chop off any drive signal a few volts over where the power tube was clipping figuring that might help. That solved the problem but I never did understand the cause. The last couple of posts have me thinking about that now.

Does the screen tend toward becoming the conductor instead of the plate when a pentode is clipping in class A?

[Helmholtz]

Does the screen tend toward becoming the conductor instead of the plate when a pentode is clipping in class A?

Not specific to class A, but screen current increases with:
- screen voltage,
- plate current,
- the difference between screen voltage and plate voltage Vg2-Vp,
- plate load Ra.

With heavy clipping the plate voltage is significantly lower than the screen voltage about half the time, so screen dissipation strongly increases.
Large enough screen resistors help to lower screen voltage when screen current rises.

Always a good idea to watch for "redscreening".

[35L6]

If you look spec sheets that show screen current you'll see when your load line hits the knee (shin ) the screen current shoots way up .

[Helmholtz]

If you look spec sheets that show screen current you'll see when your load line hits the knee (shin ) the screen current shoots way up .

Yes, no wonder as this is the condition of max plate current and lowest plate voltage.
With a sine signal this lasts only for a brief moment but with a flatted top (e.g.caused by grid conduction) it may extend to almost a half-cycle.

Good thing is that with a (sufficiently large) screen resistor the larger the screen current the lower the screen voltage, so screen current and dissipation get limited.

[Chuck H]

Yes, no wonder as this is the condition of max plate current and lowest plate voltage.
With a sine signal this lasts only for a brief moment but with a flatted top (e.g.caused by grid conduction) it may extend to almost a half-cycle.

Good thing is that with a (sufficiently large) screen resistor the larger the screen current the lower the screen voltage, so screen current and dissipation get limited.

On this same subject then...

Could there be any significance to peaks at higher frequencies tending even more toward finding their way through the screens rather than the plate under clipping/peaking conditions? It's been years but I think this was supported by what I saw on the bench. And this might explain why Marshall amps, with their higher Henry rated chokes (with a probable more appropriate impedance load for the screen path option) and using EL34 tubes manages to survive being overdriven in the name of rock n roll all these decades Just a theory form a guy that watches, and has payed attention, but doesn't actually know. In other words, could the impedance of the choke that feeds the screen supply help mitigate over dissipation?

[Helmholtz]

Could there be any significance to peaks at higher frequencies tending even more toward finding their way through the screens rather than the plate under clipping/peaking conditions?

I can't imagine such effect.
What stresses the screen is temperature. It takes a little time to heat up the wire.
So a plate current peak will cause a screen dissipation peak, but screen temperature will depend on the duration of the peak.
Short peaks will get averaged.
Apart from plate current I can't think of an other effect causing plate dissipation peaks.

And this might explain why Marshall amps, with their higher Henry rated chokes (with a probable more appropriate impedance load for the screen path option) and using EL34 tubes manages to survive being overdriven in the name of rock n roll all these decades . ....
...... could the impedance of the choke that feeds the screen supply help mitigate over dissipation?

I think what helps Marshalls and similar designs is considerable B+ sag with clipping (maybe 15% or more).
Reason is that supply current demand almost triples at full power.
When B+ sags, the voltage at the screen node will sag even more (with a little delay).

Regardless if series resistor or choke the screen node filter cap shouldn't allow for fast voltage changes.

[johnhoef]

Hello, everyone,
It's been a while, but I am back at it with some interesting things to tell you all. Firstly, after returning the amp to the guitar player, containing the solid state bridge rectifier, and not telling him about it.....His first reaction - he was absolutely thrilled to be completely hum free! After playing his guitar through this amp, he began to be disappointed. It no longer had the "sweet sound" it used to. I have always wondered if guitar players really did find that tube rectifiers really made a difference. Personally, I can't tell the difference, but he sure can. I finally revealed to him that I installed a solid state bridge in there. So, there is my psychiatric test result! Back to the drawing board. I decided to use a 6X4 tube rectifier. Main reason, it will fit in the same tube socket as the original tube rectifier, with one huge difference - this 6X4 is a dual diode, meaning full wave rectification. Considering the expense of replacing the isolation transformer with a new one that is center tapped, I decided to design a hybrid bridge. This will still create the desirable tube rectification. The result is full wave rectification, and no hum. Haven't yet tested this on the guitar player...thought I would run this by you all first to see what you all think. Here's the circuit diagram of the new power supply:

[Helmholtz]

Looks interesting!

Another thought:
Some vintage amp players seem to like a little ripple modulation. I noticed that using large value filter caps tends to spoil the vintage vibe.

Here's an interesting experiment:
https://www.ampbooks.com/mobile/clas...ass-AB-ripple/

If too technical just jump to the conclusions at the bottom.