Noisy heater CT, quiet artificial CT

[rangdipkin]

Ran into an oddball issue with my latest build, and figured I'd rant about it here and see if anyone else has run into a similar issue!

I've been working away on an AC30 TB inspired build using a hammond 270HX for the PT and a recycled yba1 output transformer. Went overboard on layout and grounding planning with quiet operation being the goal. Fired it up and found that it was buzz city with the volume above halfway! Checked and re-checked wiring and grounding, rolled preamp tubes, rolled power tubes, no change. Tried the heater center tap at the cathode bias resistor, at ground potential, and at a dedicated +100v node, and no change! Still buzzing away like a tornado of bees! George is getting frustrated!
After some curse words and a large glass of scotch, I regrouped, left the heater CT floating and clipped in a pair of 100 ohm resistors as an artificial CT, to the cathode bias resistor. Lo and behold, she's gloriously silent! Must be a defective heater CT!
Here's where its hurting my brain: the suspect defective center tap measures dead-ass balanced in both resistance and voltage, yet is useless in practice! The cheap 100 ohm resistors from my scrap heap make the heater hum inaudible.
Has anyone else run into to a similar issue? Built in center tap is noisy, but artifical just works!

[R.G.]

It is not rare for a heater winding "center tap" to not be really in the center. There's a batch of ways this can be so. Resistor center taps are often more accurate, as you have just found.

If you're after quiet wiring, have you read "Guitar Amp Wiring Notes"?

https://www.dropbox.com/scl/fi/vpijj3al ... k6ln0&dl=0

[rangdipkin]

That must be the case here, I've just never run into such an extreme case. Grounded heater CT is only marginally quieter than leaving it floating. Resistive artificial CT is whisper quiet!

And yes, that is a great reference that i find myself going back to quite often!

Cheers!

[R.G.]

You know, another possible explanation is a high-resistance or open wire lead to the CT on the winding. Your note that grounding it is only marginally better than it being completely open made me think. You could check the resistances from end to end and to the CT on the winding to eliminate this.

[martin manning]

You know, another possible explanation is a high-resistance or open wire lead to the CT on the winding. Your note that grounding it is only marginally better than it being completely open made me think. You could check the resistances from end to end and to the CT on the winding to eliminate this.

In the first post rangdipkin says it appears to be well balanced, but this would be hard to determine without a 4-wire measurement set-up IMO.

[R.G.]

You know, another possible explanation is a high-resistance or open wire lead to the CT on the winding. Your note that grounding it is only marginally better than it being completely open made me think. You could check the resistances from end to end and to the CT on the winding to eliminate this.

In the first post rangdipkin says it appears to be well balanced, but this would be hard to determine without a 4-wire measurement set-up IMO.

Yeah, but I took that "well balanced" to mean that he did a voltage measurement, both ends to CT. A high resistance CT would show up fine on a voltage measurement. A CT with a resistance of, say, 100 ohms or so would still reference a balanced winding fine if there was no other funny business going on. I'll need to think a bit about unbalanced winding resistances around the CT. That is a (dimmer) possibility, but could happen. And yes, you're right; measuring it would need four-wire measurements for the low-ohms nature of those windings.

[rangdipkin]

Well that's something I didn't even consider... I measured resistance from CT to either phase with a good ol fluke mm. Measured .3 ohms on both and figured hey, shes balanced. Just went back and measured phase to phase with the CT floating and also measured .3 ohms. Then shorted my meter probes together, .3 ohms! I guess I'm not performing the measurements to a proper degree of accuracy to confirm whether or not the winding is balanced!

Voltage wise i get a stable 3.24 volts ac on both phases, with slight fluctuations from the wall.

[Kagliostro]

For those resistance measures I've an old but special item by UnaOhm It can measure from a very low value, in the order od 1/10 of ohm to Mohms

I usually use It when measuring Transformers windings

Franco

[R.G.]

Well that's something I didn't even consider... I measured resistance from CT to either phase with a good ol fluke mm. Measured .3 ohms on both and figured hey, shes balanced. Just went back and measured phase to phase with the CT floating and also measured .3 ohms. Then shorted my meter probes together, .3 ohms! I guess I'm not performing the measurements to a proper degree of accuracy to confirm whether or not the winding is balanced!

Voltage wise i get a stable 3.24 volts ac on both phases, with slight fluctuations from the wall.

If you measured what your meter thinks is zero ohms (0.3) with the secondaries open circuited (so as to not measure the resistances of the tube heaters) then the high-resistance CT is out of the running. It's good enough.

Is it possible that the point you're tying the CT to is not really attached to signal ground?

[Kagliostro]

If .3 ohm were measured is because the multineter wasn't able to measures the real resistance and consider It as a short, no open circuit

K

[rangdipkin]

The .3 ohm measurement was taken in circuit but with the tubes and pilot light removed. Center tap was initially tied to the cathode bias resistor, fully bypassed with a 250uF cap. I also tried it at chassis ground and a voltage divider at +100v with no audible change.
If a high resistance center tap is out of the running it must be a balance issue. I'm happy with the performance using the artifical center taps via 100 ohm resistors from heater phases tied to the cathode bias resistor.

Not sure if its pertinent, but I also did try measuring the heater ac voltage referenced to chassis ground, with the CT floating. One side measured 5.8vac, the other measured .6vac. Tying either phase to ground or cathode potential noticeably reduced the noise as well, to a similar degree that the true center tap did.

Artificial center tap remains the winner!

[Kagliostro]

Perfect

Another step, if necessary, would have been to replace the artificial CT with a Humdinger

Franco