Tube pa conversion - volume level

[MikeR670]

Hi all, it's an old Garnet pro vocal 400 pa head - 5 channels. Volumes are pretty low with guitar, but I noticed that once I cascaded two channels together (with a lot of help from David Root, thanks!) - the nice overdriven result is also hugely louder. But to maintain a clean channel I'll somehow need to increase the output volume (it can't hold its own in my band setting. The new drive channel is capable of drowning everyone else out, including the drummer lol) Any ideas on a mod? Could it be as simple as changing the volume input resistor (27 k right now), and the first input cap (.05 right now)?

[MikeR670]

Here's the schematic

[JazzGuitarGimp]

It sounds like you need more signal level from your guitar. Do you have a clean boost pedal, and if so, have you tried it between your guitar and the amp? Also, in another post, you mentioned that when you're plugged into the cascaded channel, the amp "cleans right up" when you roll the guitar's volume down - can you get a clean sound that is loud enough for the band setting by using the cascaded channel?

[MikeR670]

Clearer pics

[JazzGuitarGimp]

Also, the suppressor grids are connected to -Vbias. I've never seen it done this way (the suppressors usually connect to ground). I don't know if this would have an effect on the gain of the output stage though - perhaps someone more-in-the-know will address this for us.

[MikeR670]

No, it only cleans up when you drop down to five on the guitar. So too low for the band. I do have an EQ pedal which has a boost effect slider. I tried that on the clean and it brought volume up but also created distortion - kind of growling and harsh sounding, so not useable. I've been looking at other Garnet guitar amp circuits and they seem to commonly use a pair of 68k resistors at the volume/ inputs and have no input cap. But they still have the .02 output cap from the first tube stage like mine does. Am I being too simplistic here? I've just ordered the Dave Hunter book - guess I better get some real understanding, this may become a lifelong obsession lol. Thanks for replying!

[David Root]

I hadn't noticed that about the suppressor grids. They are referenced to ground but indirectly, thru the bias circuit 82k tail resistor. So at the top of the 82k, the suppressor grids are indeed seeing the -ve bias voltage.

I've never seen that before either. If the suppressor grids are more negative than ground that should cause them to pick up fewer electrons from the plate area, no?

Maybe not....see attached page from my Garnet book. It says doing this makes the suppressor grid "better attract secondary emission", and "seems to make the tube more stable with less parasitics. Tube life is extended".
Note this only applies to EL34/6CA7, which is what this PA amp uses.

[JazzGuitarGimp]

I hadn't noticed that about the suppressor grids. They are referenced to ground but indirectly, thru the bias circuit 82k tail resistor. So at the top of the 82k, the suppressor grids are indeed seeing the -ve bias voltage.

I've never seen that before either. If the suppressor grids are more negative than ground that should cause them to pick up fewer electrons from the plate area, no?

So, perhaps miving them to ground will increase the gain of the output stage?

[David Root]

Please note my addition in my post above-your response beat me to it!

That page explains Garnet's reasoning, but doesn't answer the ? about gain/power, and also muddies things up by showing a connection of the screen to the suppressor thru pin 1, so he can use 6L6GC tubes too with proper bias, using the +ve screen voltage to reduce the bias voltage.

I'm inclined to agree with you on the gain/power question, that going directly to ground should increase power. Garnet says he wrote this book in 2003/4 when modern EL34/6CA7 production was worse than it is now so he was trying to make the amp more reliable by doing this. If so, he wouldn't be too concerned about power loss I think.

Later in the book he indicates that Pete Traynor applied the bias voltage to the suppressor as well. Whether Pete Traynor got this from Garnet Gillies or vice versa we don't know.

[JazzGuitarGimp]

He doesn't directly address gain / power, but he does say that tube life is extended, which might suggest power output and gainis reduced.

[MikeR670]

Thanks guys, excellent info! Getting pretty technical for me, unless it's a simple change over? Sounds solid though.

Something I just noticed which may be important: the four power tubes are actually 6550 tubes. With the date 07 on stickers on their tops.

[MikeR670]

Oh I just googled, doesn't look relevant - just a tonal variation re 6550 vs 6ca7 vs el34 I guess. I keep worrying about bias issues, tube life etc especially since the mods I've done. But.....drove it hard last night till 2am and no issues apart from some ringing in my ears today It hooks through a Garnet 2x12 cab I have, with those silver cone speakers. They never fart out and the bottom end seems to reach down to the depths of hell.

Anyway, I really appreciate the input and if the suppressor mod is technical I'll learn, hopefully through the Dave Hunter book!

[JazzGuitarGimp]

Unlike the EL34, the 6550 doesn't have the suppressor grid brought out to pin 1 on the tube, so this discussion is a moot point. However, swapping the 6550 for EL34 may require more signal swing from the phase inverter. Do you have any test equipment?: oscilloscope, signal generator, digital multimeter, dummy load?

[MikeR670]

I only have a good multimeter, it has resistance, capacitance, temp etc. I think I can build a dummy load unit no problem. Could I get by with those?

[Malcolm Irving]

Connecting the suppressor grid to the negative bias supply (for tubes such as EL34 which bring the suppressor grid out on to a pin) is mentioned in the recent book:
‘Audiophile Vacuum Tube Amplifiers, Vol. 1’, Igor S. Popovitch, p279
It is supposed to increase efficiency, linearity and reliability - but no mention of gain.