Gibson GA-95 RVT

[Travst]

I'll turn this in for a grade... idle current dissipation.

CT to Pin3 of V8 = 37.9 ohms, voltage = 2.9v = 13 watts
CT to Pin3 of V7 = 39.4 ohms, voltage = 3.3v = 11.9 watts

In anticipation of the next question, I have 4 unmatched power tubes. When I first got the amp, I grabbed some cheap Rubys... only ones available. From what I've read, the readings are on the hot side, is that correct?

[Phil_S]

I am notoriously bad at the toob math. Someone should say they agree with me.

We are looking for how many mA.
Ohm's Law V= I*R; I=V/R
2.9/37.9/2=38.3mA per tube average (divide by 2 because the reading is for a pair)
39.4/3.3/2=41.9mA per tube average
Actually, probably a fairly well matched quad, but we don't have individual tube readings to be certain.

This is about ~ 17-18W per tube (rounded: 40mA * 460V *4 = ~70W+). 70W is still ear splitting volume. What could be wrong?

I might be tempted to bring bias voltage to around -37 and see if it makes a difference, but I don't expect that's the answer to the problem. To do this, I would roughly double R54. See if a 15K resistor there makes a difference or maybe a 22K. Shoot for -37V.

Edit: Get the data sheet here.
http://tubedata.itchurch.org/sheets/093/6/6L6GC.pdf
Your amp is idling at about 65%. The rule of thumb is 70%. It's just a tad on the cold side and based on the schematic, I think Gibson intended that.

[martin manning]

Your math is fine Phil, but the idle dissipation doesn't have anything to do with power output- it's just sitting there wasting 70W without making a sound. What's the problem at this point? Low volume? Maybe it's time to put a signal into the input and measure the AC voltage at the output of each stage?

[Firestorm]

This is about ~ 17-18W per tube (rounded: 40mA * 460V *4 = ~70W+). 70W is still ear splitting volume. What could be wrong?

That's watts of heat, not output power. But yes, the amp should be very loud.

Your amp is idling at about 65%. The rule of thumb is 70%. It's just a tad on the cold side and based on the schematic, I think Gibson intended that.

70% idle is pretty much the maximum you can bias in Class AB without cooking the tubes with high signal. 70% has become an article of faith among some (Aiken is with you), but people like Kevin O'Connor suggest something much closer to 50%. For fixed bias amps, Fender, Gibson, Ampeg all typically biased cold.

[Firestorm]

Maybe it's time to put a signal into the input and measure the AC voltage at the output of each stage?

I don't think we have signal generator to play with. So I was thinking portable CD player and mini speaker on a stick.

You can, however, get a fairly consistent level 60Hz signal by putting a finger on the tip of your guitar cord. You should have some one else do that so you don't have two hands touching a live amp.

[Travst]

Ok... this is embarrassing. What exactly is required here?

[Firestorm]

With a signal generator, you can input a signal of known voltage and then measure the AC voltage at each preamp stage. It should get progressively larger. When you find the stage where it stops getting larger (or doesn't get as large as it should based on the gain of the stage, you've likely found the problem.

Without a signal generator, just plug in a cable and have someone touch the tip. This will produce a fairly stable 60Hz "signal." But it won't be very large. Measure its amplitude (AC millivolts) and compare that to the signal coming off the plate (AC volts -- usually best to measure after a coupling cap). The AC voltrage should be appreciably higher. Continue through the amp looking for the place where the gain falls off.

A much more jerry-rig way of doing it is to tape a .047 cap to the end of chopstick, connect one end of it one terminal of a miniature speaker (like a headphone speaker) and connect the other speaker terminal to ground. Get a mini-plug to phono plug adapter and plug the headphone out of a CD player, iPod, or something like that into the amp input. Keep the volume pretty low. You can touch the open end of the .047 cap to anything inside the amp (because the cap blocks DC), but you're interested in the plates of the preamp tubes mostly. You can actually hear the signal from the input through the little speaker. Not very scientific, but you should be able to tell whether the signal is getting louder at each stage.

In all these cases, it's probably best to pull the output tubes to do this and just deal with the preamp signals.

[martin manning]

There are several tone generators available for download which can produce a sine wave at the headphone jack of your computer. This can be fed into the input of the amp, and then your DMM can be used to measure the signal strength at points along the signal path. You can also listen to the signal at various points using a probe made from a capacitor and a piece of shielded cable, which you can feed into another amp. I'm still not clear as to what the status of the Gibson amp is. Is it not producing any sound, a very low level sound, or what?

[Structo]

Martin, I think they have it sounding OK but the output isn't near what it should be.
I think we figured it is a 90 watt amplifier which should really shake things up but it isn't near as loud as it should be.

[Firestorm]

Martin, I think they have it sounding OK but the output isn't near what it should be.
I think we figured it is a 90 watt amplifier which should really shake things up but it isn't near as loud as it should be.

Low output on both channels, too. Nothing apparent at the common stages (except possibly weak 12AU7s and I think Travst doesn't have spares to sub).

[martin manning]

So it's not established that all the tubes are good then... Is there an up-to-date voltage table with all tube pins and supply nodes?

[Travst]

Martin, I think they have it sounding OK but the output isn't near what it should be.
I think we figured it is a 90 watt amplifier which should really shake things up but it isn't near as loud as it should be.

Low output on both channels, too. Nothing apparent at the common stages (except possibly weak 12AU7s and I think Travst doesn't have spares to sub).

My spares did come in, so I have a fresh 12AU7 in V5. I'll play around with the methods described here and see what I can find out.

[Firestorm]

Especially need a fresh 12AU7 in V4, too.

[Travst]

Especially need a fresh 12AU7 in V4, too.

Interesting... I just noticed that I've been following the Gibson schematic for tubes, and it shows a 12ax7 in V4, while the Epi schematic shows a 12au7. I'll sub in the 12au7.

[Firestorm]

The voltages shown on the Gibson schematic for V4 are clearly wrong for a 12AX7. It looks like they just picked them up from a previous/different schematic even though they changed the tube type.