Orange Drop Cap Failures

GroupDIY Audio Forum

Help Support GroupDIY Audio Forum:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.

Matador

Well-known member
Joined
Feb 25, 2011
Messages
3,365
Location
Bay Area, California
I recently debugged a failing Mesa Boogie preamp, the Studio Preamp, as it had low, weak, distorted outputs. This preamp looks like most Mesa guitar preamps (topology wise), although with an additional tube stage(s) to drive "line level" outputs.

Here is (one of) the output stages:
Screen Shot 2025-02-07 at 1.00.50 PM.png

After pulling the tube, I checked the socket voltages. To my surprise, pin 1 of the tube socket was showing about 50V, whereas the B+ supply (at the top of the 121K) was about 300V. So there was about 2mA flowing through the 121K resistor (even with no tube). I then looked at the center of the output level pot, and it had 20V sitting on it. Clearly the 0.1u coupling capacitor had failed. I removed it, and tested it, and it reads as a dead short (0 ohms) across the two terminals.

After replacing it, the output level was restored, however it was a scratchy, oscillating mess. On a whim, I checked all of the front panel controls, and most had significant DC voltages sitting on them.

To fast forward to the punch line, this preamp has 19 orange drop coupling capacitors in the signal chain. After removing all of them and testing individually:

1) 6 of them (all 0.1u/400V) had failed dead short
2) 7 of them measured significant leakage, like with 10V DC across them, they all flowed over 1mA (showing up to my DMM as ~10K resistors)
3) 2 measured as high value resistors (with 10V, measured as 500K to 1M resistors).
4) 4 measured correctly (out-of-range on the DMM, and rated capacitance on an LCR meter)

I have seen coupling caps fail before when subject to over-voltage stress. However this preamp, even with 20% over line voltages, doesn't approach the VDC rating of these caps. I've replaced individual ones inside an amp, however I've never had an entire amp full of failed orange drops in this manner, to the point where essentially every single one needed to be removed and replaced.

Any idea what could have caused this? The person who asked me to fix it said it was working normally, then just suddenly the output just became low and distorted.
 
I recently debugged a failing Mesa Boogie preamp, the Studio Preamp, as it had low, weak, distorted outputs. This preamp looks like most Mesa guitar preamps (topology wise), although with an additional tube stage(s) to drive "line level" outputs.

Here is (one of) the output stages:
View attachment 145377

After pulling the tube, I checked the socket voltages. To my surprise, pin 1 of the tube socket was showing about 50V, whereas the B+ supply (at the top of the 121K) was about 300V. So there was about 2mA flowing through the 121K resistor (even with no tube). I then looked at the center of the output level pot, and it had 20V sitting on it. Clearly the 0.1u coupling capacitor had failed. I removed it, and tested it, and it reads as a dead short (0 ohms) across the two terminals.

After replacing it, the output level was restored, however it was a scratchy, oscillating mess. On a whim, I checked all of the front panel controls, and most had significant DC voltages sitting on them.

To fast forward to the punch line, this preamp has 19 orange drop coupling capacitors in the signal chain. After removing all of them and testing individually:

1) 6 of them (all 0.1u/400V) had failed dead short
2) 7 of them measured significant leakage, like with 10V DC across them, they all flowed over 1mA (showing up to my DMM as ~10K resistors)
3) 2 measured as high value resistors (with 10V, measured as 500K to 1M resistors).
4) 4 measured correctly (out-of-range on the DMM, and rated capacitance on an LCR meter)

I have seen coupling caps fail before when subject to over-voltage stress. However this preamp, even with 20% over line voltages, doesn't approach the VDC rating of these caps. I've replaced individual ones inside an amp, however I've never had an entire amp full of failed orange drops in this manner, to the point where essentially every single one needed to be removed and replaced.

Any idea what could have caused this? The person who asked me to fix it said it was working normally, then just suddenly the output just became low and distorted.
I have never encountered this phenomenon of simultaneous failure of so many capacitors in the signal chain, especially since they are sized for higher working voltages, maybe even oversized with a good safety factor
🤔
It may have been a whole batch of poor quality capacitors.
 
I’ve definitely come across shorted orange drops in Boogies before. Never all of them, but they can go. I assumed it’s the heat, as they are quite crowded, and I’ve rarely if ever seen them fail like that in other amps. The first time I came across one it took me a bit, because I really didn’t expect a shorted PP.

I may have to rethink the heat thing, as I wouldn’t imagine a preamp getting as hot as a combo. Regardless, yeah, the orange drops are now a first point of inquiry on Mesas for me. Just be glad it wasn’t the Vactrols. They stack those things, and they’re a pain.
 
Are the caps on a PCB? Do they have broken minuses?

I have had broken minuses on Silver Mica that were leaking (silver oxide). The problem was that the lead spacing was wrong and when inserted it broke the seal on the leads.

Duke
Thanks, I’m going to have to mention this to a friend who builds amps and had some silver micas go short on him. He’s all point to point, which I doubt is less stressful on those seals.
 
I would 2nd the 630v cap upgrade. 400v seems like enough but maybe not with those caps/that amp, either bad run, circuit design issues, or sometimes start up voltages can spike higher than idle.
 
Last edited:
Some circuits can give you peak voltage multiplication, in other words you end up with AC voltages that would seem impossible to achieve given the voltage restraints of the power supply.

This was discovered in Jerry Garcia's amps by Dan Healy or Stanley Owsley, I forget which. Jerry was always blowing up his Twin Reverbs. I do not know if they knew the scientific reason for this, it was captured on a scope and the phase inverter caps were changed from 400 to 600 volts, or was it from 600 to 1000? Anyway I use this info when building amps, using 1000 or 1200 volt ratings in the PI spot.
 
Some circuits can give you peak voltage multiplication, in other words you end up with AC voltages that would seem impossible to achieve given the voltage restraints of the power supply.

This was discovered in Jerry Garcia's amps by Dan Healy or Stanley Owsley, I forget which. Jerry was always blowing up his Twin Reverbs. I do not know if they knew the scientific reason for this, it was captured on a scope and the phase inverter caps were changed from 400 to 600 volts, or was it from 600 to 1000? Anyway I use this info when building amps, using 1000 or 1200 volt ratings in the PI spot.
I am not the tube guy here but suspect you might get inductive spikes in some situations. Back in the 60s when I was a tech working on early switching power supplies we saw lots of component killing spikes from interstage transformers.
==
I've heard audiophile legends (mostly dismissed) about tube amps storing energy in transformers that made them capable of more punch. I feel bad about even repeating this. 🤔

JR
 
Did they say Sprague or Mesa on them?
No, none of them carried any labelling either way, just a part number. Visually, they are identical to the new ones being sold through Cornell Dubilier. I ended up replacing everything with CDE 's poly-prop series which is rated to 630WVDC, which are yellow tubular caps of a similar dimension.

https://www.cde.com/resources/catalogs/MPW.pdf

Are the caps on a PCB? Do they have broken minuses?
Yes, definitely on a PCB, however the lead spacing appears correct for these. Nothing is visually amiss on any of the failed caps.

I assumed it’s the heat, as they are quite crowded, and I’ve rarely if ever seen them fail like that in other amps.
I had a similar though, although in a preamp there's really nothing to get particularly hot, as even the preamp tubes get just warm to the touch.
 

Latest posts

Back
Top