carvin x100b reoccuring issue... heat related?

[methodofcontrol]

i've got my friend's X100B here for the second time in 6 months. it seems R121 (350ohm/10w) has trouble staying soldered on the board.
i first looked at this amp in September when i started playing music with my friend, i let him borrow my amp because his amp, "...broke on tour like
5 years ago and was repaired on the road in some nameless town. repair guy said something just needed to be re-soldered. then it worked great for about 2 weeks on tour and broke again. i finished the tour on borrowed gear and it's been sitting ever since."
so, there's some back story.
now, back in September i opened the amp and sure enough, there's a resistor hanging by one leg. R121! i re-soldered it and found no other issues. i assumed the previous repair must've been a quick tack-on type solder job and that coupled with some bumpy van rides might explain the resistor coming up. i did a proper soldering and buttoned everything up. the amp has been working with no issues up until a few days ago where, on stage @ a venue, it started making an intermittent violent crackle sound but otherwise passed signal. by the end of that song (thankfully, the last song of the set) there was no signal.
the amp is open once again on my bench and R121 is hanging by one leg. the side that connects to the standby switch is the side that comes out.
the amp works without issue aside from this resistor coming undone after 30 hours or so of use. is this a design flaw that can be helped w/ the addition of a fan or is there something else going on?

this is a mid 90's amp w/ EL34's. schematic found here: http://www.carvinmuseum.com/pdf/amps..._schematic.pdf

[52 Bill]

It happens. The resistor heats from the current being drawn through it, until it softens the solder. The heating and cooling cycle causes the solder to crack and the connection fails.

You could replace the resistor with a higher wattage one that will dissipate the heat better, you could heatsink the resistor, you could mount the resistor off the board, you could resolder the connection with a higher temp solder like lead free or silver, etc.

[Enzo]

And you could measure the voltage drop across it to see if perhaps there is excess current flowing from a bad tube perhaps.

[methodofcontrol]

i get 450 @ the standby switch and 442 @ C77.
8v drop across 350 ohms is only about 20mA or so...is that right? i wouldn't think that an issue.
then again, my ohms law theory is pretty weak— i could be way off.

[52 Bill]

i get 450 @ the standby switch and 442 @ C77.
8v drop across 350 ohms is only about 20mA or so...is that right? i wouldn't think that an issue.
then again, my ohms law theory is pretty weak— i could be way off.

That would be 22 mA at idle. The rest of the tube stages draw through that resistor, output tube screen grids, driver stage, etc., so when the amp is really pushed there will be a lot more current drawn through it.

As always Enzo's point is a good one, what condition are the output tubes in?

[methodofcontrol]

should i hook everything up and measure it while someone plays through the amp to get a better idea?

chances of a 12ax7 causing this issue?

talked to my buddy, he says the tubes are 10+ years old with sporadic use. EL34's are winged C svet's that test well on a tube tester, FWIW.
maybe a new quad is in order?

tangent question: i noticed the power switch on back giving options for 50% or 25% power output. doesn't appear to have anything to do w/ pulling output tubes as i suspected it would...
it looks like it's acting as a sort of PPIV MV using R129 and R130 as limiting resistors, am i reading this schematic right?

[g1]

i noticed the power switch on back giving options for 50% or 25% power output. doesn't appear to have anything to do w/ pulling output tubes as i suspected it would...
it looks like it's acting as a sort of PPIV MV using R129 and R130 as limiting resistors, am i reading this schematic right?

No, I don't see any connection to the signal lines, It looks like the switch is adjusting the screen voltage via R129 & R130.

[methodofcontrol]

is that like voltage starving the output tubes to attenuate the volume? makes for a spongier sound, right?

[Gingertube]

Let me make a wild guess - The EL34 output tubes don't last too long either. At high powers EL34 draw a lot of screen current which is sourced thru' R121.
I would be changing R113, R114, R115, R116 to 1K 5W to make the EL34s last longer. That should aslo drop the dissipation in R121 a bit by reducing total screen current draw.

Also when you resolder R121 in the board make sure the resistor body sits 2 to 3mm clear of the PCB to allow airflow all the way around it.

Another "old farts" trick is to run each resistor lead away from the resistor body, loop it upward and over back toward the resistor body again and then down into the PCB, that is put a "loop" in the lead off to the PCB, That loop acts as additional heatsinking.

Cheers,
Ian

[Enzo]

Not only that, if you have a half an inch or whatever of extra wire lead on the new resistor, instead of trimming it off, clean the coating off the copper trace on the board, and lay the extra wire length down along the bare copper and run a solder bead along it. That also acts as more heat sinking.

[methodofcontrol]

Let me make a wild guess - The EL34 output tubes don't last too long either. At high powers EL34 draw a lot of screen current which is sourced thru' R121.
I would be changing R113, R114, R115, R116 to 1K 5W to make the EL34s last longer. That should aslo drop the dissipation in R121 a bit by reducing total screen current draw.

Also when you resolder R121 in the board make sure the resistor body sits 2 to 3mm clear of the PCB to allow airflow all the way around it.

Another "old farts" trick is to run each resistor lead away from the resistor body, loop it upward and over back toward the resistor body again and then down into the PCB, that is put a "loop" in the lead off to the PCB, That loop acts as additional heatsinking.

Cheers,
Ian

well...no, see the EL34's in the amp have been in there for 10+ years!
got some more info from my friend today: the amp was retubed MORE than 10 years ago and the tour on which the amp first had problems was in 2004. it was fixed on the road. and was fixed once more at home. then the amp sat dormant for 7-8 years until i got my hands on it in september.
i like the resistor lead "loop" trick as a heat sink... i'll have to try that if there's enough lead left!

new EL34's are on order.
anyone else in favor of me swapping the screen grid resistors for 1k/5w's?

[Zouto]

well...no, see the EL34's in the amp have been in there for 10+ years!
got some more info from my friend today: the amp was retubed MORE than 10 years ago and the tour on which the amp first had problems was in 2004. it was fixed on the road. and was fixed once more at home. then the amp sat dormant for 7-8 years until i got my hands on it in september.
i like the resistor lead "loop" trick as a heat sink... i'll have to try that if there's enough lead left!

new EL34's are on order.
anyone else in favor of me swapping the screen grid resistors for 1k/5w's?

That's a good idea, best thing to to...

[Gtr_tech]

That resistor is in the screen circuit in place of where a choke would be. If its getting hot enough to desolder itself, you have a failed output tube.

[methodofcontrol]

well finally got the tubes and resistors. everything is in and the amp is functioning. i forgot to order a new 350/10w... so the original one is still in there. i considered taking the 4x pulled 350/5w screen grid resistors and going series-parallel for 350/20w but i figured that to be as much a liability as sticking with the 10w original. i plan to order a new 10w with my next repair order and i'll swap it out then. thanks again for the helpful advice!

[Tooboob]

Yes 1K screen resistors will help, though they smooth out the power tube distortion a bit (I think it's a good thing). That 350 Ohm resistor tends to cook up good when power tubes short, the smoke escapes, and the board cooks. One of the only ways to kill the beasts. You can get to the top of the board easily. Make sure there's no trace from the resistor on the bottom of the board with a damaged pad. Make sure to add a bias pot while you're in there. Gain low, cranked, it should make the sound of rawk. You could run some wire from the resistor to the other end of the traces using lead-free solder, staying away from anything lower voltage, and you wouldn't have to worry about funky pads with low thermal mass, but the failure you're reporting does indicate something wrong, probably power tubes. Beef it up, and you might just fry something else in the chain.

I've got lots of X100Beasts. I'm particularly fond of the grey fuzzy XT112s and XV112s, though I frequently convert them to 6L6 or 6V6. The heads on a 4x12 get particularly monstrous. Love the clean and GEQ.