New Guy Needs Help With Gibson-esq Build

[Matt D]

Hi Guys,
I've been working on a single channel amp that is kind of a combination of a GA-40 and a GA-20. The goal has been to make it out of mostly donated and salvaged parts.

The amp uses a 5879 for the input, a 12ax7 for the phase inverter, 2 6V6s, and a 6X5 rectifier. I also made a solid state rectifier in a tube socket with the same pinout as the 6X5, but it seems to make the problems worse.

The amp uses just a volume and tone control, like one channel of a GA-40 without the tremelo.

With the tone control at 0 resistance, the amp almost sounds pretty good, if a little muffled. As soon as I turn the pot though, a loud, high-pitched oscillation and a significant amount of noise are introduced. Also, the attack of the notes has farty, papery sound.

My voltages are:

5879: 3v on the cathode and suppressor grid, 68v on the screen grid, 124v on the plate.

12AX7: Around 100v on the plates, and 1v on the cathodes

6V6s: 239v on the plates, 202v on the screen grids and 11v on the cathodes.

The voltages are measured with the volume and tone turned down.

Additionally, when the DMM probe touches the screen grid of the 6V6 directly next to the 12AX7 a loud squeal comes out of the amp.

I've tried changing the tubes, changing out the tone pot, and changing out the capacitor attached to the tone pot.

Here is a picture of my schematic and some pictures of the amp. It is my first build. The schematic gets kind of crowded towards the right side of the page.

[Phil_S]

Try reversing the leads of the output transformer or swap the inputs from the phase inverter to the power tubes. Do whichever is easier, as it has the same effect. The squeal is oscillation, typically the result of a phasing problem.

I can't help but notice that the wiring is all over the place. You may simply have parasitic oscillation from poor lead dress. Wires should be placed with purpose and cross at right angles. Heater wires should be separated from signal wires. There are "rules."

IMO, this design sounds better with the input pentode running at higher voltage, though yours is not all that low. Here is my take on it: http://home.comcast.net/~psymonds/GA20.htm

I'd look to boost the plate voltage on the 12AX7.

Let us know what happens. If reversing phase doesn't do it, we can dig in and see if we can figure out what's going on, maybe look at lead dress issues.

[Tonegeek]

Looking at the schematic, shouldn't the ground be on the other side of the 200ohm/25uf network off the 6v6 cathode? U may have corrected this in your build or you would smoke the output tubes pretty quick.

[Firestorm]

I agree with Phil that lead dress is your biggest challenge; you've got the signal criss-crossing all over the place. Mark the leads that are in phase with each other (or get some different color wire!) and separate them; keep the grid leads as much away from everything else as you can (or shield them).

I see the connection that Tonegeek mentioned, but I think you have it wired right, but drawn wrong. One thing though; you have your preamp and phase inverter running off the same power supply node. You do not want to do that. You'll need another filter cap and a dropping resistor to decouple them.

[Matt D]

Thanks for the replies and advice. Tonight when I get back from work I'll check the ot, then take a look at the lead dress.

[martin manning]

Since there is no global negative feedback in this circuit there is no reason to think that swapping the OT leads will help. I agree its likely a lead dress problem.

[Matt D]

Ok, So I spent today working on the lead dress. I moved and secured the heater wires, output jack, and ot wires. I got some taller standoffs and moved the board higher to get it further away from the heaters and and ot leads. I made sure to keep different signal, b+, and heater wires away from each other as much as possible and......

It worked! This amp still probably will never win a beauty contest but it is relatively quiet and sounds good.

I had just a half hour or so to play it before before my domestic situation necessitated that I keep the noise down. Here is what I noticed:

The amp begins to distort at a fairly low volume. It is a pretty good sounding distortion, but I would like to get more clean headroom. Any suggestions?

The tone control isn't as effective as I would like it to be. How can I achieve a little more tonal variety?

I actually have no idea what impedance the OT wants to see. It is a donated part I just thew in there. It sounded good through an old Jensen 8 ohm 1x12 but perhaps my impedence is mismatched. How do I measure this?

I have built a solid state rectifier with the same pin-out as the 6x5 on a male tube base. I'll give that a try tomorrow. Also, I want to figure out the best way to decouple the preamp and phase inverter given the already cramped layout.

Thanks again for your help!

[martin manning]

Do a search on measuring OT impedance ratio. Basically you put an AC signal in one side and measure the voltage there and on the other side, calculate that volatge ratio and square it to get the impedance ratio.

Re the tone control, I would think some kind of tone stack between the first stage and the PI would be better than the post-PI treble bleed. Maybe do something like a 5E3 Deluxe? You could use the same two pots that you have, just replace the 5E3's chanel 2 volume pot with a 1M to ground. Also, I have to believe that the 470k series resistor feeding the PI is rolling off a lot of treble.

BTW, your power stage cathode resistor is probably grounded at the oposite end from the one shown in your schematic ;^)

[Structo]

Hopefully the rectifier isn't wired as shown in the schematic either.

[Matt D]

The power resistor is in fact grounded on the other end. I think I must have been pretty tired when drawing the schematic. I'll take a look at how I have depicted the rectifier.

[overtone]

...The amp begins to distort at a fairly low volume. It is a pretty good sounding distortion, but I would like to get more clean headroom. Any suggestions?...

+1 to Phil_S suggestions to raising the voltages in the pre-amp.
Your voltages are fairly similar to the ones I measured on a '54 GA-40 that I picked up last year and that had zero headroom too.
The voltages in that amp were actually too low because the PT was having trouble.
Getting all the voltages up in that amp gave it some headroom, but don't expect much of that from this circuit!
Ditto on any expectations of the tone control too.
This GA-40 still only has 285V on the 6V6 plates.
It has the original OT and it measures 10,5k into an 8ohm load. That 10k ball park has been confirmed by others.

Good luck, tony

[Phil_S]

What Overtone says on headroom....

To get the maximum plate voltage from the PT, leave the solid state rectifier in place. It's time to forget about the tube rectifier. IMO, there is no real advantage to using one in this situation.

You've got the PI and the input pentode on the same power node. As someone else pointed out, these should be separated. Add another dropping resistor and a cap.

The goal, IMO is to maximize plate voltage to the input pentode. To do this, I would make the dropping string one 5K and two 1K resistors, or 5K-1K-470.

I think the amp presently pulling about 10-11mA across the second 10K resistor. It is difficult to predict exactly, but I think going to a 1K resistor will raise plate voltage on the PI to the upper 100's and on the pentode to the middle 100's.

I might dress the 5879 a bit differently, as well. Some experimentation may be needed. Suggestions, but without real knowledge of possible result: change the 750K to 1.5M or 2M and reduce the cathode resistor considerably from 3.3M to as low as you can go. The cathode is easy to experiment with and you can start by tacking in a jumper directly to ground, bypassing the R/C.

I am glad to hear better lead dress has worked in your favor. This is the reason that neatness counts in amp building -- and it is not obvious to the beginner. Good job fixing things up!

[Zippy]

What Overtone says on headroom....

To get the maximum plate voltage from the PT, leave the solid state rectifier in place. It's time to forget about the tube rectifier. IMO, there is no real advantage to using one in this situation.

You've got the PI and the input pentode on the same power node. As someone else pointed out, these should be separated. Add another dropping resistor and a cap.

The goal, IMO is to maximize plate voltage to the input pentode.

Regarding maximizing plate voltage on V1: Might this be a good place to use a parallel power supply to the input and PI rather than a series dropping string?

[Phil_S]

Regarding maximizing plate voltage on V1: Might this be a good place to use a parallel power supply to the input and PI rather than a series dropping string?

I don't know, but did consider it without any calculation. My gut (not the most reliable) was telling me that a voltage divider (what you basically have with a parallel feed) would not be advantageous or only minimally so. The whole current draw downstream from the power tube screens is only about 10-11 mA. Current draw for the input and PI will probably drop as voltage rises due to the lower R values in the supply ladder. I keep thinking it's not justified here, but I suppose you never actually know without trying it.

[Matt D]

Thanks again guys for the help and advice.

There may have been some confusion about this, but the previous voltage readings were with the 6X5 rectifier. I put in the solid state one tonight and got the following:

B+ : 285v

6V6's: Around 280v on the plates, 235v on the screens, and 13v on the cathode

12AX7: 113v on the plates, 1v on the cathodes

Pentode: 145v on the plate, 81v on the screen, and 3v on the cathode and suppressor.

I definitely have a lot more clean headroom with the SS rectifier.

In the interest of saving space I'll probably look for a three section can cap. 20/20/20? I'll use that to split the PI and preamp try a different power resistor dropping string.

The amp has a nice warm sound that overdrives easily. If anything, it is a little muffled sounding. If I wanted to add some treble (I hesitate to use the unspecific "sparkle" or "presence") to the overall sound of the amp, should I start with the preamp dressing or the 470k resistor feeding the PI?

This amp is turning out to be what I was hoping it would be based on what I read about the GA-40: A good one-trick pony amp for David Lindley-style electric lap steel.

I'll keep posting here as I tune the amp up.