1969 Gibson GA5-T Amp PS Caps

[texstrat]

Yes, it seems wrong, but I thought we need to concentrate on the ripple at node 1 before looking downstream. Now that you raise it, maybe the OP isn't measuring it correctly? What else might explain this? This is hard from far away.

Texstrat, you are taking these measurements with the black probe grounded to the chassis (and not the bolt for the green line feed ground), and the red on the + terminal of the cap?

I have the neg probe grounding to the chassis and I might have tested the leads where they terminate at the lugs with the pos probe. Anyway here are the measurements taken at the lead as it comes out of the cap. These measurements are taken after replacing the 6X4 and 6BQ5 tubes. All tubes are new. For reference those tubes are V1 Mesa Boogie, V2 Tesla JJ, V3 GE, V4 Groove Tube.

Red
0.3 VAC
292 VDC

Green
0.2 VAC
164 VDC

Blue
0.2 VAC
270 VDC

Voltage Readings. Values in parenthesis are last set before changing out the V2 and V3 tubes.

V1
Pin 1 - 116VDC (122VDC)
Pin 2 - 0.2mV (0)
Pin 3 - 0.92VDC (0.87VDC)
Pin 6 - 114VDC (110VDC)
Pin 7 - 3mV (0.3mV)
Pin 8 - 1.03VDC (1.0VDC)

V2
Pin 2 - 23mV (10mV)
Pin 3 - 8.16VDC (8.20VDC)
Pin 7 - 277VDC (264VDC)
Pin 9 - 274VDC (260VDC)

V3
Pin 1 - 262VAC (258VAC)
Pin 6 - 263VAC (258VAC)
Pin 7 - 296VDC or 140VAC (258VDC or 113VAC)

V4
Pin 1 - 134VDC (136VDC)
Pin 2 - 13mV (11mV)
Pin 3 - 0.93VDC (0.87VDC)

The hum/noise is gone.

So what do we have going on here?

[Phil_S]

Green
0.2 VAC
164 VDC

This one, voltage is too low. The schematic says 200V. Your ladder drops from 292 to 270 to 164. This is a drop of 106V across a 100K resistor, or 1mA. That doesn't make sense to me. You are supplying one two section triode (12AX7). Downstream calculations are roughly equal, dropping 50v across 100K is 0.5mA per triode. These triode sections should each be drawing 2-3mA or 5-6mA total for the whole tube. Since the second calculation is consistent with the first calculation, I think the trouble is not with the tubes or tube sockets, but with the B+ supply ladder. Something is bleeding off the voltage and limiting current.

I haven't liked that 100K resistor in the B+ ladder from the begining. That's the one between the nodes reading 270v (blue) and 164v (green). We are supposed to see a 50v drop there not 106v. Ignoring for a moment the low current problem, backing into this, 270V - 50V = 220V. If we assume 5mA for that last leg, then R=V/I or 50/.005 = 10K. I would change that resistor to 10K. There will be some heat generated there and I'd use a 1W or higher rated if I had one even though the math says a regular 1/2W should be able to do the job.

I think we are dealing with a misprint in the schematic and it is possible whoever built the amp at the factory followed the schematic. I would change this and see what happens.

V3
Pin 7 - 296VDC or 140VAC (258VDC or 113VAC)

Does the 140VAC stay the same or decline as you check at each of the filter caps (red-blue-green wires, in that order)? This just ain't right.

BTW, I am not troubled by 296VDC. I think it is easy enough to account for +8% for line voltage difference from yesteryear to today and the rest to just ordinary random +/-5% variance that we expect.

V4
Pin 1 - 134VDC (136VDC)

This is another half of a 12AX7 not drawing much current. I get almost 1mA. (Tremolo should be switched off to measure this one.) I suspect this is may be related to that 100K dropping resistor between the green and blue wires (B+2 and B+3).

OTOH, the 4.7K cathode resistor is a relatively large value. It's going to shift the bias point of the tube and may actually be doing its job as intended there.

The hum/noise is gone.

So what do we have going on here?

Hmm....it is hard to argue with success. Things seem wrong, yet maybe they are OK? You run the risk of fixing it 'til it's broken.

I think the hum is gone because you changed the rectifier. It is simple to test. Put the old one back in. If that's not it, then it's the power tube.

Noise is gone, but how's the tone? This amp should have killer tone.

Martin, if you see this, I'd sure be happy to hear your thinking on this one.

[martin manning]

OK, the red, blue, green voltages look reasonable (for where we are), but the ripple is extremely low on all nodes- only ~1%. And, the red wire should be the same as pin 7 of V3, but there is a huge difference in the ripple. Something is wrong with one or both of those measurements. Let's just ignore that for the time being, because we know all of the associated components are new.

Re the green voltage, I agree with your math Phil, between the two triodes supplied by that node there is only 1.06 mA current. But look at the input stage as shown on the schematic: 200V B+ with a 100k plate load and 2.2k Rk... that will only be about 0.6 mA at the quiescent point. Then there is the 4k7 shared Rk for V1b and V4. I'll bet if the 4k7 shared Rk value were restored, V1b's current draw would go down and that third node's voltage would come back up. Can you try that texstrat? Just remove the 820R resistor (or just lift one end if you like) that is across the 4k7 now.

It may quit working again, but maybe not. I think the two pre-amp stages may have been biased at very low idle current for the tonal quality, and likewise with the 100k dropping resistor. If it won't work with a 4k7 Rk on V4/V1b, then maybe find the largest value that does work, and see what that sounds like.

[martin manning]

dupe

[texstrat]

Guys, had to work last night.

I did not de-solder the positive lead on the ps caps when I tested them. The caps were in circuit and power was on. Also, the voltages that I listed are where it settled, meaning I got voltages in the 125+ range, and then it steadily decreased until the voltage drop stopped. So the strange values are more than likely my mistake reading them. The DMM was grounded to the chassis and pos probe testing the lead coming out of the cap.

This is new territory for me, so if I need to remeasure the voltages, please advise the proper way and I will get that data.

Let me know.

[texstrat]

Ok, I desoldered the 820R resistor and I still have sound. The tremolo sounds real funky, but the voltages look a bit better. Whereas before, the amp was stone quiet, no hum/hiss, now I have a small amount of hum/hiss.

Here are the new voltages, values in parenthesis were yesterdays numbers.

V1
Pin 1 - 186VDC (116VDC)
Pin 2 - 0.0 (0.2mV)
Pin 3 - 2.2VDC (0.92VDC)
Pin 6 - 136VDC (114VDC)
Pin 7 - -11mV (3mV)
Pin 8 - 1.22VDC (1.03VDC)

V2
Pin 2 - 21mV (23mV)
Pin 3 - 8.16VDC (8.16VDC)
Pin 7 - 274VDC (277VDC)
Pin 9 - 274VDC (274VDC)

V3
Pin 1 - 261VAC (262VAC)
Pin 6 - 262VAC (263VAC)
Pin 7 - 295VDC or 140VAC (296VDC or 140VAC)

V4
Pin 1 - 217VDC (134VDC)
Pin 2 - 0.0 (13mV)
Pin 3 - 2.23VDC (0.93VDC)


Rechecked ps cap values and got the following:

Red
42 VAC
294VDC

Green
42VAC
198VDC

Blue
42VAC
274VDC

I also reflowed the old solder on the lugs and terminal strips and added new.

So are we moving in the right direction?

[Phil_S]

So are we moving in the right direction?

Yes. I like these voltages better. Can you elaborate on how the tremolo sounds? What happens as you rotate the pot? How's the tone of the amp, in general?

[texstrat]

Yes. I like these voltages better. Can you elaborate on how the tremolo sounds? What happens as you rotate the pot? How's the tone of the amp, in general?

Phil it has more of a whump, whump loose bassy sound. Before it was tighter and had more treble to it.

[martin manning]

So now you have a baseline using the original circuit values, and it leaves a little to be desired. Probably the 4k7 cathode resistor needs to be smaller, but that will drop the supply voltage below 200V as you saw when the 820R resistor was across it. I guess it worked, but the plate voltage on the pre-amp tubes was low and you won't have much headroom. Because of that I think it's a good idea to try decreasing the value of the 100k dropping resistor, but there is an infinite number of possible combinations of that dropping resistor and the V1b/V4 cathode resistor.

Here's one way you could explore the possibilities while staying close to the original circuit's "intent":

Change the value of the 100k dropping resistor by tack-soldering another resistor across it, starting with say a 330k, which will give you 100*330/(100+330)=77K. Then clip-lead a pot (at least 25k) as a variable resistor across the 4k7 cathode resistor and find a setting that results in a voltage at the green power supply node of 200V (trem off). The cathode voltage is low, but make sure that your clip leads are insulated and that they won't touch any high voltage leads. See how that sounds, then try a smaller resistor across the 100k, like a 220k, and repeat.

I'd also like to see pin 3 of V1 and V4 at about 1.2V, so look for that too.

[texstrat]

So now you have a baseline using the original circuit values, and it leaves a little to be desired. Probably the 4k7 cathode resistor needs to be smaller, but that will drop the supply voltage below 200V as you saw when the 820R resistor was across it. I guess it worked, but the plate voltage on the pre-amp tubes was low and you won't have much headroom. Because of that I think it's a good idea to try decreasing the value of the 100k dropping resistor, but there is an infinite number of possible combinations of that dropping resistor and the V1b/V4 cathode resistor.

Here's one way you could explore the possibilities while staying close to the original circuit's "intent":

Change the value of the 100k dropping resistor by tack-soldering another resistor across it, starting with say a 330k, which will give you 100*330/(100+330)=77K. Then clip-lead a pot (at least 25k) as a variable resistor across the 4k7 cathode resistor and find a setting that results in a voltage at the green power supply node of 200V (trem off). The cathode voltage is low, but make sure that your clip leads are insulated and that they won't touch any high voltage leads. See how that sounds, then try a smaller resistor across the 100k, like a 220k, and repeat.

So I can remove the 820R from the 4.7K resistor between V3 and V4? Does the wattage matter on the 25K pot? Do I solder the leads to lug 1 and 3?

Is the 100K dropping resistor, coming off V1 pin 8 to the terminal strip where the green ps cap lead is soldered (refer to my layout)?

Doing both of these things, which is going to effect the green ps lead voltage? And what change has caused the tremolo to sound odd?

Two resistors that are not on the Gibson schematic that have been added at some point are two 100 ohm on V2 pins 4 and 5 that are soldered to the terminal strip center lug where the blue and green ps cap leads are soldered. I don't know if these resistors are causing any issues.

I'd also like to see pin 3 of V1 and V4 at about 1.2V, so look for that too.

What I am reading on the schematic voltage chart, V1,V2 and V4 pin 3 should be 2.3V and V1 pin 8 should be 1.6V. What I have now is close to these values. Because I don't understand, why would you want to reduce V1 pin 3?

Thanks.

[Phil_S]

So I can remove the 820R from the 4.7K resistor between V3 and V4? Does the wattage matter on the 25K pot? Do I solder the leads to lug 1 and 3?

Remove the 820R. Wire the pot as a variable resistor. Jumper the wiper with one outer leg. Then use the other outer leg and the wiper (middle lug). This on the cathode, so your typical 1/2W pot will work fine.

Is the 100K dropping resistor, coming off V1 pin 8 to the terminal strip where the green ps cap lead is soldered (refer to my layout)?

That layout is hard for me to use. The 100K resistor is between the green and blue cap wires. It should be on the terminal strip.

Doing both of these things, which is going to effect the green ps lead voltage? And what change has caused the tremolo to sound odd?

We are trying to re-arrange the distribution of voltage here to trigger the tremolo. I think, and suspect Martin agrees, that the voltage swing isn't enough to get the oscillation needed. Everything affects everything else -- it's interactive.

Two resistors that are not on the Gibson schematic that have been added at some point are two 100 ohm on V2 pins 4 and 5 that are soldered to the terminal strip center lug where the blue and green ps cap leads are soldered. I don't know if these resistors are causing any issues.

Often not shown on the schematic, but typical to have. This is the artificial center tap for the heater supply and it's purpose is to eliminate hum. Leave it alone. You will regret removing it.

I'd also like to see pin 3 of V1 and V4 at about 1.2V, so look for that too.

What I am reading on the schematic voltage chart, V1,V2 and V4 pin 3 should be 2.3V and V1 pin 8 should be 1.6V. What I have now is close to these values. Because I don't understand, why would you want to reduce V1 pin 3?

V1 pin 3 voltage affects the oscillation of the tremolo. Martin understands this better than me, but I fairly certain he thinks this will better trigger oscillation, so the tremolo will sound decent. It shares the 10uf cap and the 4.7K resistor with V4 pin 3. At this point, I guess trial and error is the easiest path to making it work. The pot allows you to dial in the tremolo. Then you read the resistance on the pot and you'll know what resistor is the best to place there. Otherwise it's just too much work to figure it out. Remember, when you are done, the pot is parallel to the 4.7K, so do the proper calc to determine the actual value. Martin gave you the math above for the 100K dropping resistor.

[martin manning]

So I can remove the 820R from the 4.7K resistor between V3 and V4? Does the wattage matter on the 25K pot? Do I solder the leads to lug 1 and 3?

Yes you can remove that for this experiment. An ordinary linear pot will be fine since there will only be a few volts across it. I would use lug 2 (wiper) and lug 3 (CW) so the resistance goes down as you turn the pot CW, which will increase the current flow through V1b and V4, and lower the Green lead voltage.

Is the 100K dropping resistor, coming off V1 pin 8 to the terminal strip where the green ps cap lead is soldered (refer to my layout)?

The 100k dropper is the one with the green and blue power supply leads connected to either end, mounted on one of the terminal strips.

Doing both of these things, which is going to effect the green ps lead voltage? And what change has caused the tremolo to sound odd?

Reducing the value of the 100k dropping resistor will raise the green lead voltage (less voltage is dropped for a given current flow), and reducing the value of the common cathode resistor will lower it (by increasing the current flowing through the dropping resistor). The operating points for the trem oscillator and the second pre-amp stage are tied together by the common 4k7 cathode resistor. I think both V1b and V4 are biased too close to cut-off now. The voltages on the schematic are a good place to start, but I think what you need to do now is tweak things to get the best tone.

What I am reading on the schematic voltage chart, V1,V2 and V4 pin 3 should be 2.3V and V1 pin 8 should be 1.6V. What I have now is close to these values. Because I don't understand, why would you want to reduce V1 pin 3?

Centering the bias point will require a voltage in that neighborhood. Is the tone worse now than it was with the 820R resistor across the 4k7 common cathode resistor? This little exercise will let you hear the difference as the bias points of V1b and V4 are moved back toward the center of the characteristic curves while maintaining the 200V supply. You should get less distortion and fuller tone as you move in that direction, and I expect the trem sound will clean up too.

[texstrat]

Remove the 820R. Wire the pot as a variable resistor. Jumper the wiper with one outer leg. Then use the other outer leg and the wiper (middle lug). This on the cathode, so your typical 1/2W pot will work fine.

So if I understand this correctly, the wiper, middle lug, two leads will come off it, one lead with lug 1 lead and the other lead with the lead from lug 3

Like this:

[martin manning]

you only need two wires from the pot: one from the wiper and one from lug 3; leave lug 1 unconnected. A larger value pot will keep you closer to the 4k7 as a maximum effective Rk, so maybe a 50k or 100k pot would be better. Start each trial with the pot full CCW and you will see minimum current flow and maximum green node voltage. Turning the pot CW will increase the current and lower the voltage at the green node.

A little more explanation: I'm trying to find combinations of dropping resistor value and V1b/V4 cathode resistor value that will hold the green node voltage at 200V while moving the operating points of those two triodes towards center. V1a (the input stage) uses the same supply node, and it is doing fine right now. If we keep that supply at 200V, then the voltages on V1 pins 6, 7, and 8 will stay as they are. Whomever put the 820R across the 4k7 had the right idea but, as you saw, without lowering the 100k dropping resistor, that caused the supply voltage to drop 50V- too much, IMO.

Tubes from 1969 may have worked fine with the resistor values shown on the schematic, but these don't seem to like them. This will probably be the case with most other new production tubes too. Like Phil says there are several things that are interdependent here, and this is the quickest but still systematic way I can think of to explore the tonal landscape.

[texstrat]

Martin/Phil:

Here are the results using a 50K pot and starting with a 330K resistor tacked onto the 100K resistor between green and blue ps cap leads.

Here are the new voltages, values in parenthesis were yesterdays numbers.

V1
Pin 1 - 186VDC (186VDC)
Pin 2 - 0.0 (0.0mV)
Pin 3 - 2.09VDC (2.20VDC)
Pin 6 - 140VDC (136VDC)
Pin 7 - 0.0mV (-11.0mV)
Pin 8 - 1.25VDC (1.22VDC)

V2
Pin 2 - 12mV (21mV)
Pin 3 - 7.9VDC (8.16VDC)
Pin 7 - 268VDC (274VDC)
Pin 9 - 296VDC (274VDC)

V3
Pin 1 - 257VAC (261VAC)
Pin 6 - 257VAC (262VAC)
Pin 7 - 290VDC or 248VAC (295VDC or 140VAC)

V4
Pin 1 - 209VDC (217VDC)
Pin 2 - 68mV (21mV)
Pin 3 - 2.11VDC (2.23VDC)

Rechecked ps cap values and got the following:

Red
267 VAC
291VDC

Green
276VAC
206VDC

Blue
248VAC
270VDC

The hum is back and the tremolo seems to be working like it was earlier. I did do a search for Gibson tremolo sound clips and mine amp does sound similar to what I am hearing with other vintage Gibson amps. As soon as I turn on the tremolo, I can hear the oscillation through the speaker before I strum the guitar strings, is this supposed to be?

It is interesting to hear the difference using the pot as a variable resistor. With it off, the tone is muddier and as I add resistance, the tone has more clarity. The tone I like is with the pot set at 22K, almost half way. So if I understand correctly, the total resistance between V3 and V4 is now 3.9K? If I change the 330K to 220K what can I expect to hear from the amp?

Here are two pictures showing how I connected things. The 330K resistor is left long until the correct value is determined, plus it is tight in there. Right now I have the 330K tacked on the 100K leads. Is this the same as soldering the resistor to the two outside lugs on the terminal strip?