6G6a PI voltages

[pontiacpete]

I'm building one of these, voltages are pretty close around the board except for the PI. THe schematic says that the plates have 330v and 320v. I can't seem to get more than 260v /265v . Theres 450v at the top of the 82K and 100K plate resistors, but no voltages in the 300's. Any thoughts? Is this a schematic error? Or?
thanks
pete

[pontiacpete]

no takers here on this one, but for whatever purpose, this build has had a problem getting a good tone, lots of crossover distortion in the low end and nasty growl for A or C down to low E. I thought it might be the PI voltages. The output transformer is not stock for a 6G6 so I thought it might have to be replaced. Instead I ended up changing the NFB from the 56K by paralleling .22uf cap and 100k pot and tried dialing in a better sound. This helped a lot. I'm using a 15" weber 8 ohm 80W california bass speaker, which now I suspect might be too light for this amp. Today I'll bring it over to a friend with an assortment of cabinets and heavier speakers and listen to it.
This amp sounds great for a guitar but for bass so far it's nothing but a problem, in fact guitar amps, I conclude, are way easier than bass amps to build .....

pete

[tubeswell]

450V on the Ht for the PI you say. What are the voltages like in the tail?

[MWJB]

Given the 450v supply to the PI tube I would expect around 300v (+/- 10-15v) on each plate, plus tail & cathode voltage. You definitely have a 12AX7 in the PI, not a 12AT7? If your PI voltages are low, fix this first, otherwise you are just muddying the water.

6G6A does seem much more popular for guitar than for bass. If you really want to use this amp for bass, I'd suggest that you could really do with more b+ voltage than 470vdc (if this is what you have), more filtering at the main B+ node (at least 2x220uf 350v caps wired in series, each cap bypassed with 220K 2W resistor) and 6550 power tubes (upgrade screen grid resistors to 1K 5W, make sure the PT can take the extra heater current, a hi voltage 370-0-370VAC twin style PT would be good). Maybe sub the PI tube for a 12AT7 (once you have the expected voltages with the 12AX7)

The cathode follower bass channel can sound a bit farty with the controls set low.

Crossover distortion? Low-ish PI plate voltages shouldn't be causing this, what plate current do you have on the 6L6s?

[pontiacpete]

" You definitely have a 12AX7 in the PI, not a 12AT7?"

Yes, it is a 12AX7, in fact I tried a 12AT7 and got even lower voltages. I cant' see why this is so. 100K and 82k split the 450v from where they meet. Almost makes sense that about 260v would appear on the plates. No?
I combed through this part of the circuit , even had help from a friend to check double check the work. mmm...

"6G6A does seem much more popular for guitar than for bass. If you really want to use this amp for bass, I'd suggest that you could really do with more b+ voltage than 470vdc (if this is what you have), more filtering at the main B+ node (at least 2x220uf 350v caps wired in series, each cap bypassed with 220K 2W resistor) and 6550 power tubes (upgrade screen grid resistors to 1K 5W, make sure the PT can take the extra heater current, a hi voltage 370-0-370VAC twin style PT would be good). Maybe sub the PI tube for a 12AT7 (once you have the expected voltages with the 12AX7)"

I really like this idea of putting in 6550's . Before i had put in the bucking transformer I was getting somewhere around 515vdc B+. I thought that might be high for a pair of 6L6s.(although I have seen super's run as high as that). Now I'm getting 470v. Suppose I reverse the wires on the little bucking xfmer and bring up B + to about 545v or 550v? Do you think a pair of 6550's would operate well with there? Do you think the OT would need to be changed? Would there be a mismatch there? The PT and OT came out of an almost gutted Earth Sound Research amp that I suppose was running 4 EL34 with a SS preamp(in case you haven't already seen my post on this project)


"The cathode follower bass channel can sound a bit farty with the controls set low."

"farty" better describes the low end of this amp now. I tried it at a friends today with some heavier more powerful speaker cabinets and the same farty sound.

"Crossover distortion? Low-ish PI plate voltages shouldn't be causing this, what plate current do you have on the 6L6s? "

I think my wave has smooth out a bit with the adjustment to the NFB. Need to put it backon the bench and give it a look. I was getting oscillation also, so I put a 100pf cap across the volume control, that helped a bit. Then put another on the 1st 12AX7 pin 3 to pin 1, and the oscillation disappeared.

Thanks greatly for the help
regards
pete kc2wid

[MWJB]

"100K and 82k split the 450v from where they meet. Almost makes sense that about 260v would appear on the plates. No?" Not really. Do you have 35v-ish at the cathode, 1.5-2v...actually make that more like 2.5-3v between cathode & tail resistor?

Yes, the 6550 will easily take 550vdc, but be aware that all your power supply voltages will rise accordingly, rewire the screen supply as a totem pole (like the recommendation for the main filters) use 2x100uf 350v caps. I'd see how the amp sounds at 515v (close to what aa original 6G6A would run).

If the OT was originally meant fior 4xEL34 then it should take the current requirements of the 6550 (50-55mA per tube). I assume that the OT was originaly wired for 4ohms (1.8K:4ohms?) and that you have switched to 8ohms with 2 power tubes to redress impedance issues (3.6K:8ohms)? With an 8ohm output Fender would have used a bigger NFB resistor...just like you have done.

[pontiacpete]

"Do you have 35v-ish at the cathode, 1.5-2v...actually make that more like 2.5-3v between cathode & tail resistor?"

the cathode pin 3/8 has 50v.
the grids have 34v.
between the tail and cathode, do you mean the top of the 220 ohm and the cathode? that has 1v. From each grid to cathode there's .9v.

"I'd see how the amp sounds at 515v (close to what aa original 6G6A would run)"

I'll give this a try in the day time, too late for the neighbors.

thanks for the help
pete

[MWJB]

"between the tail and cathode, do you mean the top of the 220 ohm and the cathode? that has 1v." There is no 220ohm resistor in the 6G6A PI circuit, cathode resistor is 820ohms, tail resistor is the 6.8K that connects to it. If you really have a 220ohm resistor there, then 1v/220ohms= 4.5mA. 2.25mA (for one triode)*100K = 250v drop. 450v at the B+ node, less 250v drop = 200v, plus 50v at the tail will see a plate voltage of 250c +/- a bit.

Ergo, you have the wrong value cathode resistor at pins 3&8 of the PI tube.

[pontiacpete]

Thank you MWJB, I looked at that many times, zoomed in on both the layout and the schematic, and thought it was a 220 ohms. Should have done the math, duh.
I'll put that in and listen to it. And also try it with the B+ raised to 515v

Still want to try out the 6550s , though, I like the idea of getting more power out of this amp.
thanks again.

petekc2wid

[pontiacpete]

It's sounding a lot better, but still has a big a** sag on the low end. I may try tweaking the NFB some more. But the voltages are looking more normal.

Plates - 519v
Screens- 517v
-54v on the grids

PI plates 337v
cathode 38v
grids 26v

I'm going to build up the power supply like you suggested, with the 2 totem filters, 220/350v for the main supply and 2 100/350s for the screen supply and give the 6550s a go, of course when everything comes in from the supply house. In the meantime I'll do some guitar playing through it. : )
thanks , will be back with the results.

pete

[pontiacpete]

6550s

With B+ at 515v, I put the 6550s and my new caps in 2 totems, the 1st 220/350v for the plates and the 2nd a pair of 100uf /350v for the screens, each in series. I put two 220k 2 watt resistors but wasn't sure exactly what you meant by 'each cap bypassed' so I followed as in a bassman aa165,(schematic attached if necessary), and paralleled each of the 220ufs. Turned it on and the 220ks fried,,,, so I must have done something wrong.
So I clipped them out and turned it on and it sounds much much better than the 6L6's, I like it a lot.
I didn't play it for long , but will I need to use bypass resistors on these caps in the long run? and how do you mean?
Also, is it necessary to bring B+ up any higher?

thanks
pete

[MWJB]

"I put two 220k 2 watt resistors but wasn't sure exactly what you meant by 'each cap bypassed' so I followed as in a bassman aa165,(schematic attached if necessary), and paralleled each of the 220ufs. Turned it on and the 220ks fried,,,, so I must have done something wrong." Yes, you must have. This works fine in practice, as shown on the AB165 schem, recheck your wiring. The 200uf at the screen supply node should also have the 220Ks. These resistors ensure that the B+ voltage is split evenly between the caps, rather than one cap taking most of the voltage, the other taking very little.

"Also, is it necessary to bring B+ up any higher?" Not if you like the tone as it is, maybe if you needed more headroom.

[pontiacpete]

figured out the mistake i made, ,, now the power supply is set and working as it should. There was 570v on the plates and -55v on the grids, but the sound wasn't quite there, too much crunch for bass. I changed out the bias resistor that had originally been changed from 56k to 22K back to 56K and this brought the grids more negative to -65v but the plates in turn dropped to 530v. It does sound pretty darn good way but i can't help but wonder about what you had suggested that bringing the plate voltage up would give me more head room.
How can this be done? Or should I not bother and just call it done?

[MWJB]

"It does sound pretty darn good"...this suggests that the voltage you have is fine. A new PT is the best way to up ther B+ voltage.

[pontiacpete]

OK, thanks again for the help!

regards
pete