A question on filtering caps

[catnine]

On this AA764 champ I built.

Right now it has a 47ohm between pin 8 of the 5Y3 and the B+ junction to buffer the 5Y3 from the 47uf cap to the plates a 1k dropping resistor to the screen then a 22uf cap 10k dropping resistor to the plate of the second gain of V-1 then a 22uf off to the 100k plate resistor to the V-1 preamp.

Can I add another say 30 uf filter cap after the 47 ohm resistor before the 47uf cap both before the plate or add say a 1k 3 watt before the 30 uf to lower the B+ more then to the 47uf to the plate of the 6V6 ?

So instead of two dropping resistors and three filter caps I would have three dropping resistors and four filter caps and adding 1k ohm before the plate I would lower my B+ on the plate so I could use a lower than the 680 ohm cathode resistor to say 500 ohm bias the 6V6 closer to the mid point of saturation and clipping and at the same time have a safer 30 uf reservoir cap for the 5Y3 and the second filter cap of 47uf would make a stiffer power supply to retain the bottom end I have now?

Is this wrong or over kill ? I ask because I have seen some BF fender amps all push pull with three dropping resistors and four caps and wonder if this cold be done on a single ended champ amp . Or is the reservoir cap on a AA1164 PR the 20uf cap right off the 5u4gb ?

I am not math or tech savy just asking .

[lowell]

It sounds like you don't want to change anything but the power amp stage. If that's correct then you need to drop the B+ at the first cap (reservoir) without changing the series power supply resistance, and also retain voltages down the line through your dropping resistors. If it were me I'd add an amplified zener. Reason being that a large resistor before the reservoir cap might create too much sag, and change the bass response too.

Here's a pic from geofex. I added one to my deluxe to have a class A mode, works great! Oh to only have this affect your power amp, put this in series w/ the CT of the OT. The anode of the zener connects to the CT of the OT and the cathode of the zener connects to where the CT B+ is tapped off the B+ supply line.

[lowell]

oh and if you want to save time let me know, I can get you the part numbers of all the parts you'll need. That would've helped me when I first tried this.

[tedmich]

The Duncan Amps PS design program can save a ton of headaches here IME

[MWJB]

"So instead of two dropping resistors and three filter caps I would have three dropping resistors and four filter caps and adding 1k ohm before the plate I would lower my B+ on the plate so I could use a lower than the 680 ohm cathode resistor to say 500 ohm bias the 6V6 closer to the mid point of saturation and clipping and at the same time have a safer 30 uf reservoir cap for the 5Y3 and the second filter cap of 47uf would make a stiffer power supply to retain the bottom end I have now?"

Firstly how much plate voltage do you have right now? Remember to deduct cathode voltage to get your corrected plate voltage.

All AA764 Champs use a 470ohm cathode resistor, to be biased halfway between cut-off & saturation you'd probably be looking at a plate voltage (measured to ground) of 425vdc+ (400v corrected). Of course, by then the 6V6 will be dissipating 20W, but thousands of them do and have lasted decades doing so.

There's no reason why you shouldn't go larger on the cathode resistor, to calm down plate dissipation but make sure that you are doing it because you like the sound of the amp like that. I have an 800ohm in my SF Champ & 440vdc from plate to ground, for that reason.

[tubby]

Reason being that a large resistor before the reservoir cap might create too much sag, and change the bass response too.

Correct me if I'm wrong, but in a SE class A amp, like the Champ, there can be no sag since plate dissipation is highest at idle. And I'm not sure if bass response would be affected that much, either.

[MWJB]

"Correct me if I'm wrong, but in a SE class A amp, like the Champ, there can be no sag since plate dissipation is highest at idle." Sorry, but you are wrong. Ten minutes spent with a bunch of different rectifiers will prove the point. When you hit a note, plate current can't rise (not by much anyway, depends on bias conditions and most champ style amps are not drawing the max from the PT B+ winding) but plate voltage will drop, it's ohms law. Plus you have the differing internal resistances of the varying rectifiers, they all have an effect on the envelope of sag & decay...effect of cathode bypass cap vs no bypass cap...NFB vs no NFB...fixed bias vs cathode...filtration...

Sag can be more evident however in push-pull amps, biased to class A style plate currents, where the PT B+ winding (maybe rectifier too) is struggling with the demands of the power tubes, and/or if a brute force choke filter is used.

I'm not saying that sag cannot appear to be eliminated in such an amp, but as a blanket statement, it doesn't cut it.

[Wakculloch]

Well spured on by this, as i have a SE, 6v6 amp i decided to check voltage drop on the HT line at idle, full power and at an input level at three times that. Found the 330v dropped to 325v at full power, but then rose again to 330v on overload. Unsurprisingly the current increased by 1mA at full power, but then fell by 2mA on overdrive.
Does this constitute sag?
I can't beleive it would be audible.

[MWJB]

I'm not sure what the test is supposed to demonstrate, sounds to me like you used a constant signal? The envelope of attack, compression & decay is what is heard when you feed a signal full of transient peaks, like when playing guitar.

Did you actually try different rectifiers and listen to the amp?

330v on the plate is very low for a champ style amp, what value cathode resistor do you have? If only 470ohms, the 30 something mA that the tube is drawing is not going to be taxing the power supply, 6V6 or the rectifier much...I wouldn't be surpised if the mA rose a little when a signal was applied.

[Wakculloch]

Tried a guitar signal and there is no drop in HT.
Don't understand.
What has the rectifier to do with it. If your dropping voltage then it's not class A by definition, but if it's single ended it must be class A?
Help.

[MWJB]

"Tried a guitar signal and there is no drop in HT." That's not a total surprise, I don't know what PT you used, but it's probably for at least twice the load it currently has, your 6V6 is also biased (at a guess, if you have a 470ohm cathode resistor) to around 10 to 10.5W & there is the potential for the tube to draw more current under signal & not cause the power supply to drop voltage. Measure mA & B+ simultaneously with a bias probe.

"What has the rectifier to do with it." Everything in the power supply has an effect on sag - The PT, rectifier, filter caps...the weakest link in the chain will sag first. A SS rectifier has no internal resistance and won't sag (the next weakest link is now the focus), a 5Y3 has a high internal resistance and will.

Single ended amps are class A, according to most definitions anyway (not all, please no hair splitting - I'm not looking to find THE definitive class A criteria), as there is no "out of phase tube" to handle any part of the cycle, like there is in a PP amp. Therefore all signal is conducted by the one power tube. Of course, there are varying conditions of bias, ideally according to some trains of thought a class A amp should be biased halfway between cut off & saturation...even in your amp that would require a plate current of around 50mA according to the tube makers data sheets. However, few production amps meet this criteria. SF Champs are usually in the ball park, but you will still hear a difference in envelope if you switch the 5Y3 for a GZ34, or SS rectifier. So there must have been some sag with the 5Y3.

A common definition of class A is that the power tubes should conduct for 360 degrees of the cycle.

[Wakculloch]

Still don't understand why a change in rectifier can alter an amount of sag that basically does not exist. A difference of 2mA between differing drive conditions can only drop 2 volts even if you inserted a 1K resistor into the HT line.
As the AA764's HT is 360V and the valve should draw approximately 40mA under all conditions and the 5Y3 is rated at 125mA, i can't see where sag, even if it could happen, would happen.

[MWJB]

Why does it not exist? I think many folks are reading Aikens definitions of Class A and focussing on a few points out of context...I read it as Aiken saying that a "true class A amp" (biased halfway between cut-off & saturation - yours clearly isn't) won't sag at full, undistorted, rated power (where most folks typically don't play them). A champ is not concerned with being a "true class A" amp, it is a cheap (relatively), cathode biased guitar amp, that the designers wanted to make reasonable sound & power and not eat tubes/fail under warranty. Typically then, they run at higher voltages and lower currents than someone might design when trying to build a "true class A amp".

Who says the AA764's HT is 360v? You have one there, made by Fender, no variac, just straight in the wall and you're getting 360vdc? Fender's voltages on schems & layouts are typically low compared to today's wall AC voltages.

A 6V6 will only draw 40mA @ 360v if that particular 6V6 is biased to draw 40mA at 360v. With the 470ohm cathode resistor, there is no guarantee that your 6V6 will draw exactly 40mA at 360vdc.

Yes the 5Y3 is rated at 125mA, at that 125mA it will drop 50vdc compared to a SS rectifier. At the currents found in champs (real, working, Fender built champs, not described on outdated bits of paper) voltage drop will be considerably less due to the lower current draw...but surely you can see that the internal resistance of the 5Y3 is an issue. Try a couple of lower resistance rectifiers and tell us what you hear? It takes less time than to read this thread, or type a response.

The most pertinent line in Aiken's article is the bit where he tells you not to worry about class. A champ is a single ended, cathode biased guitar amp with NFB.

[Wakculloch]

It does'nt exist cos-
Ive tried to measure it and it's not there.
Theoretically an amp that draws almost constant current under different conditions cannot exibit sag. No matter how much voltage gets dropped across the rectifier or any other impedance in the HT line, unless there is a significant increase in current, that drop will remain the same. The design is single ended. It is class A. There is no sag.
That an amp sounds or performs differently with different rectifiers is irrelevant. That is not sag.

[MWJB]

"It does'nt exist cos-Ive tried to measure it and it's not there." It's there if it can be heard. When you go to a gig do you connect up meters to the performers amps or do you listen with your ears?

"Theoretically an amp that draws almost constant current under different conditions cannot exibit sag." An amp that draws constant current & voltage won't sag, but is that what you have? Most guys I know turn up cathode biased amps until they sag, or find that they are already sagging too early for their taste. But then, most cathode-biased guitar amps are not "True class A" amps.

Give it to me, I'll make it sag...as will any number of players who have tried to reduce sag in their Champs.

"That an amp sounds or performs differently with different rectifiers is irrelevant." How is it? An AA764 is an amp with a 5Y3 as stock, so if the 5Y3 rectifier sags, so does the rest of the amp. IF there is no sag elsewhere, then sure you can control what there is by substituting rectifiers. Anything that affects the tightness of response in the amp, affects the overall sound, you can't listen to the amp without hearing the effect of the rectifier, it's no good saying your amp doesn't sag if someone then says, "Why's it sagging?", you can't say, "oh that's just the rectifier".

You seem to be giving one aspect of this particular amp's design (the accidental part, due to it just having 1 power tube) undue attention/importance. You don't seem to be considering the whole - that it's a little cathode-biased amp, with little trannies that will meet its ceiling faily early & the player will push it beyond clean rated output to make up for this.

Prove me wrong by all means, but you'll need to give me fuller & more detailed specifications of your test.