Hot Rod Deluxe volume loss

[rwhenshaw]

So I bought 2 types of resistor at Radio Shack because I wasn't sure which one I needed to do the test/replace old one. One is 100k ohm at 1/4 watt and 5% tolerance. The other is 100k ohm at 1/2 watt. Which one of these (if either) would be the right one to use and why? Also, in parallel means I would clip each lead of one resistor onto its matching lead on the other resistor?

[JoeM]

You want (at least) the 1/2 watt in this application. It's not so much the watt rating, but the higher wattage most likely has a higher voltage rating. I usually use 1W for tube plates, but most Fenders these days use the 1/2 watts, as does your HR amp. Since you already have the new resistor, I'd just go ahead and replace R4.

[Leo_Gnardo]

So I bought 2 types of resistor at Radio Shack because I wasn't sure which one I needed to do the test/replace old one. One is 100k ohm at 1/4 watt and 5% tolerance. The other is 100k ohm at 1/2 watt. Which one of these (if either) would be the right one to use and why? Also, in parallel means I would clip each lead of one resistor onto its matching lead on the other resistor?

Good move! I'd clip in a half-watter. Here's my "theory" - if R4 is faulty, and you clip in a 100k in parallel to it, your plate voltage will rise to @ 200V and your signal should return back to normal. With the pre tube in its socket. If not, the test resistor is less likely to go up in smoke, always a disconcerting event.

g-one mentioned "lifting" one end of the suspect resistor, BUT I'm trying to prevent you needing to remove the whole %$#@ board to replace that resistor, a time consuming and aggravating job.

- - - - - -

If there's no improvement with the clipped-in resistor, "the game is on" and we have to brainstorm why your problem is happening and what's the solution.

[rwhenshaw]

I'm not sure if I did this correctly. I took the second resistor and bent the leads around the leads of the resistor on the board - in parallel, to the best of my knowledge. My readings were: plate side= 10v and supply side = 276. Note: the 276 started around 330 and slowly made its way down to 276. Does this make any sense?

[g1]

Yes it makes sense. It means the resistor is ok and not causing the problem (provided the parallel resistor was making good contact with the original resistors leads).
Can you do a resistance check (power off) from pin3 to ground?

[rwhenshaw]

1.48 k ohms

[Leo_Gnardo]

I'm not sure if I did this correctly. I took the second resistor and bent the leads around the leads of the resistor on the board - in parallel, to the best of my knowledge. My readings were: plate side= 10v and supply side = 276. Note: the 276 started around 330 and slowly made its way down to 276. Does this make any sense?

Means something's drawing a lot more current than it should on the tube end of R4. How about a resistance reading from pin 1 to ground? This is getting mysteriouser by the minute.

1.48K pin 3 to ground sounds OK.

[Leo_Gnardo]

By way of explanation rw, we were searching for more common failure modes. What I'm suspecting now is much less common: a shorted or very leaky coupling cap from that 1st stage pre plate to the next stage.

Don't worry. If that's what it turns out to be, that cap is still an inexpensive part. Removing the old one and replacing it will probably mean pulling out that big circus board, a headache I was trying to spare you.

Some med students get drilled with this notion: "When you hear hoofbeats, think horses." Whaaat? OK let's expand that, "When you hear hoofbeats, think horses, not zebras." Most symptoms resolve to commonly found problems. Not much point in making the case for exotic, rare diseases or conditions. As in medicine, so it goes in electronics repair. Failure of the cap I'm suspecting is definitely rare. We've pretty much ruled out everything else.

[rwhenshaw]

There is no continuity from pin1 to ground.

Any ideas of how to proceed? Thanks again to all for the help.

[Leo_Gnardo]

I took the second resistor and bent the leads around the leads of the resistor on the board - in parallel, to the best of my knowledge. My readings were: plate side= 10v and supply side = 276. Note: the 276 started around 330 and slowly made its way down to 276. Does this make any sense?

no continuity from pin 1 to ground

2 possibiities for the drifting-down voltage you observed. First, as the tube filament warms up, the tube becomes more conductive and that process takes @ 20 seconds. Second, having measured no continuity from pin 1 to ground, possible the coupling capacitor to the next stage measures "open circuit" as it should, at low voltage as applied by the ohm meter. Then starts to get leaky as high voltage is applied with the amp on.

The cathode voltage @ 4V has me scratching a bigger bald spot atop my head.

About the only thing we haven't asked for yet is a voltage check on the control grid pin 2, when the amp is on. Can't imagine what V applied here would be enough to force the plate V all the way down to 6V with the pin in place. Pin 1 at 85V with tube out of socket still tells us current is being sent somewhere it shouldn't be.

Reaching out for ideas. Anyone got any?

Where are you in this world rw? Maybe one of our TAG correspondents could have a look. I'm mid Hudson Valley, if anywhere nearby send me a PM. This is an intriguing case - bound to learn something.

[rwhenshaw]

Voltage at pin 2 = 11v. I live in central New Hampshire.

[g1]

Check that pin2 voltage with your black probe connected to pin3 rather than ground. Sometimes grid to ground voltages are tough to get a proper reading, grid to cathode will always show the correct voltage between them.

[Leo_Gnardo]

Check that pin2 voltage with your black probe connected to pin3 rather than ground. Sometimes grid to ground voltages are tough to get a proper reading, grid to cathode will always show the correct voltage between them.

Good advice g-one, and if the grid is biased +7V wrt cathode, that would explain a lot of unexpected behavior. Now where in sam hill is that coming from?

Any chance of contamination on the circuit board? Somethin' got spilled in? Unplanned conduction paths could do what we're seeing here.

Could try cleaning the tube socket area with a cotton swab soaked in solvent. I use ethyl alcohol - high concentration - AKA Spirytus Rektyfikowany (96% from Poland) or Everclear 95% USA - both could be considerd dehydrated vodka. Denatured alky would do too, or a spritz of non-residue contact cleaner spray on the q-tip. In a pinch, a good highly pure vodka like Tito's would do. (Beware - a lot of vodka contains a little sugar, up to 1/4% just enough to goof things up.) Let area dry thoroughly before applying power. Please be careful. Those hi-proof items are flammable. VERY flammable. Have had to use things like this for a field fix.

[g1]

If indeed the grid is getting positive with respect to cathode, as you say it could be the socket area, but I have a hunch it could be related to the ribbon cable, maybe a wire or solder whisker or flux or something.
Unfortunately, the whole board will likely have to be removed.

[Leo_Gnardo]

socket area, but I have a hunch it could be related to the ribbon cable, maybe a wire or solder whisker or flux or something.

Or the socket itself or under it.

Oh well, rw, you're in some scenic area I take it. Take a break & have a look at the great outdoors. Central NH must be beautiful this time of year. Between monsoon rains (like here.)