Problem with this fizzy sound

[Silvio55]

Ok Chuck, I got it wrong! I thought that the idea was to bring down the value of R19 by using the pot as a variable resitor in parallel. Just have a 500k pot, canīt get a 1M around here soon. Can I try with the 500k wired like that or it has to be a 1M pot?

[Chuck H]

Yes. If possible you would basically substitute the pot for R19 and feed the PI from the wiper.

[Silvio55]

I soldered the pot across R19, one leg and the wiper, but it acts like an on/off switch, I get lower volume at the very first and then full volume if I move the shaft a little. Is it right? I just checked but I canīt do more testing now (got to go to job), it seemed that the fizz was gone but maybe it was because of the lower volume and I couldnīt hear it. Later Iīll do more hearing testing and report.

[sickman82]

Just out of interest Chuck, should it turn out that the problem can be isolated to the PI, where do we proceed from there? Altering voltages and the PI bias?

[Chuck H]

Just out of interest Chuck, should it turn out that the problem can be isolated to the PI, where do we proceed from there? Altering voltages and the PI bias?

Right. But with no promises. A scope would sure help on this one. Then you could literally watch the problem happen. You could see where it's happening and whether it's only the top or bottom of the swing, etc.

Silvio, I sort of figured the pot would act that way. That's why I mentioned the more ideal wiring twice. Just turn it down to where you have some volume with the post master full up. Set the preamp the way you usually do when the problem occurs. Listen closely. The only point is to see if the problem is happening even when the PI isn't being driven too hard.

[Silvio55]

I did some testing today and with the pre PI master set low the fizz is still there, but I couldnīt set the post PI master full up, it was still loud with the pre master set at zero, so I had to turn it down untill I got a lower volume. The fizz is exactly the same at the same overall volume as before.

[Chuck H]

Let's guess then that the fizz might be in the preamp. I might try replacing R13 with a 2.2k and replacing R9 and R26 with 150k. This will raise the plate voltages on the first preamp tube and as a byproduct heat the bias a little on that tube. Gain is reduced slightly with the reduction in plate load value but I'm hoping this will be somewhat offset by the gain increase of the raised plate voltage. But don't be surprised if there is a small reduction in gain.

It occurs to me looking at the schem that R19, which we used as a target point for the pre PI master, is the bias feed resistor for the cathodyne. There is no actual load to 0V in the circuit to access for a proper pre PI master. So a really proper pre PI master would have been more complicated. But I'm hopeful that the test is leading us in the right direction anyhow.

[pdf64]

Yes, I'd be happier if the pre-cathodyne master vol was implemented in a more conventional fashion, before we dismissed it. Pete.

[Silvio55]

I realised that if I conected the other free leg of the pot to ground it acted more like a volume pot instead of an on/off switch. From 0 to 2 (aprox.) the volume is zero, from 2 the volume makes a little jump up and from there to 6 or 7 it goes up and from 7 to 10 it goes down again but not to zero or as low as the pot in 2. So I tested that way with the post master full up and the pre master in 2 and the fizz was still the same.

[Chuck H]

Good. To me that rules out the cathodyne as the source of this particular problem. As I mentioned, I've seen (heard) this in other amps that ran low (relatively) plate volts on 12ax7's. Also, the 220k plate load can sound sort of generally fizzy when the tube is clipping IME. So I'm thinking the problem is in the second gain stage. I would try the above circuit mods to see if it helps. It's only three resistors.

[pdf64]

'I realised that if I conected the other free leg of the pot to ground it acted more like a volume pot'
Did you fit dc blocking caps to the hot track end and from the wiper, ie from the plate an to the grid?

[Silvio55]

No Pete, I didnīt put any caps. The ground thing was accidental and tried that way, but also tried like Chuck told me without grounding and the fizz was always the same

[Chuck H]

This is still the wrong way to wire a pre PI master. But by grounding the unused leg of the pot you did one of two things. You either grounded one side of the cathodyne, effectively rendering it more like a normal class A triode conducting mostly from the plate, or you grounded the input signal and grid of the cathodyne, which removes the bias from the circuit as it shunts the input signal. Either way, if niether of these scenarios made any difference to the crackle I'd bet that the problem isn't with the PI.

Do remove that pot from the amp directly.

[Silvio55]

Iīve been doing more testing with the voltages. The wall voltage in my house varies from 205 to 220 (the 228 I got in the first reading was really unusual as I realised now), beeing 210 the most usual, and measuring from the output of the PT I get nowhere near the 299 volts that the schematic states, itīs between 270/285 volts, and after R2 or R7 I get 255/265. I donīt know if itīs a big difference but if low voltage is the problem, itīs coming right from the PT, also the drop in voltage after R2/R7 seems bigger in proportion than the schematic readings.

EDIT: At 210 volts Iīm getting 5/5.1 volts at the heaters, at 220 I get 5.5volts. Iīm thinking that the PT in my amp is really a 240 volts transformer instead of a 220 volts. On the back of the chassis it sais "220/230 volts 50/60h".

[diagrammatiks]

hang on a minute...

are you sure you're getting 5-5.5 on the heaters?

5volts on the heaters is kinda problematic.