12ax7 cathode resistor burning up?

[aab0mb]

I have here a Vox v941 which is a tube driven effects loop similar to a Dumblator.

https://el34world.com/charts/Schematics ... ematic.pdf


The 1k cathode resistor on v1b is fried and has scorched the circuit board. Bad tube killed by the high stress of v1a cathode follower? Bad .47 coupling cap before the SEND? Bad IC brand filter caps?

I haven’t seen many fried cathode resistors on preamp tubes. What might have caused this and what should I look towards when repairing this unit so it doesn’t happen again?

[dorrisant]

Check that the value of the 100k plate resistor for that 2nd stage is correct. If it were too low you'd be pulling too much amperage through the he cathode resistor. I don't think that the bypass cap would be causing it. Most IC brand caps are perfectly fine, unless they come stock in a Fender amp. Fender's bean counters have squeezed the good out of the design, but the IC caps you get through vendors are much better quality. Fender gets what it pays for.

Also, check for trouble in the tube socket itself.

[aab0mb]

Check that the value of the 100k plate resistor for that 2nd stage is correct. If it were too low you'd be pulling too much amperage through the he cathode resistor. I don't think that the bypass cap would be causing it. Most IC brand caps are perfectly fine, unless they come stock in a Fender amp. Fender's bean counters have squeezed the good out of the design, but the IC caps you get through vendors are much better quality. Fender gets what it pays for.

Also, check for trouble in the tube socket itself.

Recapped it since it’s like 30 years old, new cathode caps and resistors, new plate resistor, random sovtek 12ax7 and cleaned up the scorching on the board. Fired right up and voltages look good. Can’t wait to run it with my high plate skyliner.

[nuke]

To do that kind of damage, you'd need to see 30-40 or more volts across that resistor. It wouldn't come through a functioning plate of a 12AX7/ECC83, as the plate resistance is about 60k ohms.

I'd tend to suspect a heater-cathode short in the tube, possibly triggered by excessive voltage elsewhere.

It looks like the 470-ohm is toast as well as the 1k resistor.

The voltage doubler power supply in this schematic tends to be stressful on the 80uf caps. They get a lot of ripple. Use a high-ripple current rated capacitor, rated for long life and higher temperatures to get more life out of them. When these caps fail in the power supply, all bets are off on how the rest of the circuit works.

The schematic depicts the filament supply with no ground reference. So it is just floating, and is subject to whatever H-K leakage happens. That's usually not good. With V1-a in cathode follower configuration, I think elevating the heater supply a bit might be in order. The old 12AX7 datasheets usually list a 180v H-K maximum voltage, but a lot of modern day tubes don't live up to that.

If you setup a simple voltage divider from the "A" output of 260v, with a target of about 30-50v, and set that as a reference between a couple 100-200 ohm resistors across the filament supply, it would probably enhance tube life and reduce hum.

[aab0mb]

To do that kind of damage, you'd need to see 30-40 or more volts across that resistor. It wouldn't come through a functioning plate of a 12AX7/ECC83, as the plate resistance is about 60k ohms.

I'd tend to suspect a heater-cathode short in the tube, possibly triggered by excessive voltage elsewhere.

It looks like the 470-ohm is toast as well as the 1k resistor.

The voltage doubler power supply in this schematic tends to be stressful on the 80uf caps. They get a lot of ripple. Use a high-ripple current rated capacitor, rated for long life and higher temperatures to get more life out of them. When these caps fail in the power supply, all bets are off on how the rest of the circuit works.

The schematic depicts the filament supply with no ground reference. So it is just floating, and is subject to whatever H-K leakage happens. That's usually not good. With V1-a in cathode follower configuration, I think elevating the heater supply a bit might be in order. The old 12AX7 datasheets usually list a 180v H-K maximum voltage, but a lot of modern day tubes don't live up to that.

If you setup a simple voltage divider from the "A" output of 260v, with a target of about 30-50v, and set that as a reference between a couple 100-200 ohm resistors across the filament supply, it would probably enhance tube life and reduce hum.

Wonderful info! Thanks a ton.

I was wondering about the heater supply so that’s very illuminating. I’ll have to install a terminal strip and try that.

Hopefully the f&t 80uF’s a dropped in are adequate for the job. It had caps from the mid 90s in it so a refresh was in order.

I’ve seen people around here mention modding these units. Any insight into tweaks that could further enhance this unit’s performance? Is the dumblator design vastly superior?

[Stevem]

If the amp is second hand then someone might have tried out a 5755 which has a different pin out.

[aab0mb]

If the amp is second hand then someone might have tried out a 5755 which has a different pin out.

It’s just a 12ax7 effects loop in a 1u rack box like a dumblator. The installed tube was a basic 12ax7. Good to make sure though! I’ve definitely bought broken gear with incorrect tubes installed before. It’s always an adventure.

[LOUDthud]

Those carbon comp resistors don't look original, all the other resistors are carbon film. What's going on with the 1K ? It looks like someone tried to attach a wire there.

[aab0mb]

Those carbon comp resistors don't look original, all the other resistors are carbon film. What's going on with the 1K ? It looks like someone tried to attach a wire there.

Yeah they probably blew this thing up once before and repaired it or had a bad tech try it. They didn’t recap it, maybe used the same bad tube, generally bad troubleshooting and service.

I cleaned up all the scorching, recapped it, new tube, new cathode and plate components. She’s happy now. I more so wanted to learn why this unit would fail in this fashion as I haven’t seen many toasted cathode resistors.

[nuke]

As far as "improving" it goes, I'd probably try the simple mod of creating an artificial center tap of the filament supply and attaching that to a well-bypassed elevated DC of around 50v as I mentioned.

The power supply is a voltage doubler, and like I said, they are pretty rough on capacitors. Go to digikey or mouser and pick out some extra-long life, high-temperature rated capacitors. I've been using Kemet capacitors designed for LED light fixtures. Very low ESR, and they're rated for 20,000 hours at 105c. Total overkill in a Blues Junior or what not, but they fit perfectly and they'll outlive me most likely.

Just use the parametric selector at Digikey or Mouser and pickout some replacements. The microfarad value isn't as important as the voltage rating.

Beyond that, it is kind of what it is. The majority of the weight and cost of the unit is the rack-mount enclosure it is housed in.

I'd think other improvements would likely be found in layout and lead dress, if needed, to improve grounding and noise. But I don't know what's in there or how it acts at this point.

To be honest, other than "tube purity", an effects loop is a good place for a high-quality solid state implementation. Pretty much anything you put in the loop is gonna be full of crappy, bargain basement, solid state parts anyway, so if you spent $7.00 on a luxurious OPA2134 chip, it would far exceed anything you might put in the signal chain.

[aab0mb]

As far as "improving" it goes, I'd probably try the simple mod of creating an artificial center tap of the filament supply and attaching that to a well-bypassed elevated DC of around 50v as I mentioned.

The power supply is a voltage doubler, and like I said, they are pretty rough on capacitors. Go to digikey or mouser and pick out some extra-long life, high-temperature rated capacitors. I've been using Kemet capacitors designed for LED light fixtures. Very low ESR, and they're rated for 20,000 hours at 105c. Total overkill in a Blues Junior or what not, but they fit perfectly and they'll outlive me most likely.

Just use the parametric selector at Digikey or Mouser and pickout some replacements. The microfarad value isn't as important as the voltage rating.

Beyond that, it is kind of what it is. The majority of the weight and cost of the unit is the rack-mount enclosure it is housed in.

I'd think other improvements would likely be found in layout and lead dress, if needed, to improve grounding and noise. But I don't know what's in there or how it acts at this point.

To be honest, other than "tube purity", an effects loop is a good place for a high-quality solid state implementation. Pretty much anything you put in the loop is gonna be full of crappy, bargain basement, solid state parts anyway, so if you spent $7.00 on a luxurious OPA2134 chip, it would far exceed anything you might put in the signal chain.

Excellent advice and your points are well received. Wonderful tip about the capacitors and I’ll be trying those in the future.

When something like this comes into my world it’s as much a learning opportunity as anything. I’ve heard these units mentioned as stand ins for a Dumblator so I’m looking at the differences and potential performance of each of those units. I have a Dumble style high plate classic build so this was a quick/cheap way to get into an active effects loops and explore some of the discussed benefits of a Dumblator type unit in the passive loop.

There are a bunch of threads here discussing tweaks/mods for the Dumblator so I’m reading through those and trying to learn something from the discussions while testing and contemplating minor circuit changes that I might make to better suit my needs.

Layout, lead dress, etc are all pretty clean as, like you mentioned, there’s not a lot in there.