• Disclaimer: This Vendor's Forum is a paid-for commercial area. Unlike the rest of diyAudio, the Vendor has complete control of what may or may not be posted in this forum. If you wish to discuss technical matters outside the bounds of what is permitted by the Vendor, please use the non-commercial areas of diyAudio to do so.

need some help- redplate and cooked pcb

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
hi,

i have forever had an issue with fuses popping, but i let a friend borrow it, after many hours of testing without any other issues.

he gave it back to me with a failed channel and he said he didnt notice anything was wrong until it quit working, which was after the capacitor desoldered itself from the board and fell onto the cabinet.

i flipped the amp over to see a 560 ohm resistor had cooked the pcb, and the large cap next to it had bulged and desoldered itself. i replaced the cap(1500@63 volt) and the resistor measured poorly so i replaced it too.

i flipped the amp power on, i ended up with the same channel going red plate.

after turning it off, the capacitor was very hot, along with the resistor, which was not as hot.

can someone please help me troubleshoot this?
 

Attachments

  • 20180905_100146.jpg
    20180905_100146.jpg
    790.6 KB · Views: 216
  • 20180905_100341.jpg
    20180905_100341.jpg
    437.5 KB · Views: 212
There is too much current passing through the tube - it is glowing red because it is severely overheated. The resistor scorched the board because it was overheated.

Among the likely possibilities:

The tube went bad and is acting as a short. However, if it was not bad in the first instance, your friend allowing it to remain in the red plate condition long enough to melt solder and scorch the board would be severely harmful to it. I would replace it.

The coupling capacitor is leaking dc from the preceding stage unto the control grid of the glowing power tube, altering the bias, and allowing the tube to run at an excessive power. The coupling capacitor does several things - one of them is to block the dc from the preceding stage and keep it from appearing on the grid of the power tube. If DC does leak across the capacitor, it will offset some of the negative voltage formed from the drop across the cathode resistor.

The 220K grid resistor has failed;

A combination of all of the above.

What test equipment do you have? I would start with removing the glowing power tube and measuring what voltage, if any, is at the tube socket pin for the control grid. I think that is pin 5, but that is off the top of my head, so I would double check. Also measure its resistance to ground.

With no power tube installed, the dc voltage should be zero. Resistance to ground should be 220K + - the tolerance of the grid resistor.

Others may have better ideas. Good luck!

Win W5JAG
 
Interesting timing. After having had almost the exact same issue back in 2014 but without the time to really look into it and rebuild the amp, I'd left the Tubelab SSE on the sidelines to properly look into the issue at a later date. In the last month or so I've finally had the time. Although it's more of a spiritual successor to the original Tubelab SSE as the PCB, parts and enclosure are all new...

As was the case with your friend, I was not in the room when the amp failed - it tended to be left on and powered down only if I was leaving the house. As far as I am aware the capacitor didn't have an issue, but R17 exhibited the same post mortem. Had the cathode resistor for the second tube (R17) fail, appearing as though it had overheated or arced such that it deformed the PCB and left a large black burn in the same spot as on yours - at the midpoint of it's rectangular white marking on the board - to the point that there was no solder mask left in that location.

For reference, it was spaced 5mm or so off the board and (in a move I now regret) it looks like although I ordered the Xicon cement wirewound resistors at the time, for some reason I soldered a different cement wire wound resistor to the board. Unsure of the brand, just a generic no name 560 ohm white ceramic wirewound. Upon closer inspection of R17 vs R27 noted light brown discoloration of the cement filling on the underside of the former along with a split in said filling. I know I still have the board and resistor that failed, if I can find them I shall upload photos.

Now, what I had assumed was this was simply a catastrophic failure of a sub par ceramic wirewound resistor inside an enclosure that in hindsight was allowed to run far warmer than it should have been. New enclosure is going to receive far better ventilation, throwing in a good quality brand part for R17 and R27 and was considering beefing them up the 7w to 10w as a nervous precaution.

Alas, having seen your issue exhibit the same post-mortem but with more information on the cause and symptoms has me wondering... interested to see what you find, and what measurements I get when I power up the new SSE in the next couple of weeks...
 
90 to 115 volts across the cathode cap is NOT good and will blow the cap and cook the resistor causing it to fail. Sometimes the resistor blows open and all the current tries to flow through the cathode cap causing it to vent, spew its guts, or just explode. After eating this much voltage and getting hot the new cap is likely bad also.

The SSE is a cathode biased amp, so there is no bias supply to fail. This leaves only two possibilities to cause this. A bad tube, or a bad coupling cap. A wrong value, or bad grid leak resistor (the 220K part) could possibly cause runaway, but I have never seen it happen.

They are Svetlana EL34 valves, probably the most reliable in the world.

That depends.....Svetlana tubes sold in the US, or intended for the US market are made in the same St. Petersburg factory as Sovtek, EH and the others. The US importing agent for the "Winged C" tubes folded and Mike Mathews got the rights to use the Svetlana name in the US. From there they went everywhere. Svetlana branded tubes without the "Winged C" logo are probably rebranded Sovteks. Even Winged C tubes are suspect. I purchased some "Winged C" tubes from AES and they were NOT reliable, and tended to red plate even when operated within the specs. You really don't know what you get with the Svetlana brand anymore.

Modern production tubes do tend to have a lower quality vacuum and less purity in the materials than NOS tubes. Operating these tubes at high temperature will further degrade the vacuum and at some point in their future they may go into a runaway condition. NOS tubes are not immune from gas related runaway, they just have a better vacuum and cleaner, purer materials to start from, so they may last a dozen years or more. Some of the new production stuff lives long too. My SSE still has the original EH KT88's in it from 2009, but I have seen some new tubes fail in a year.

A leaky coupling cap will put positive voltage on the grid of the output tube resulting in the same runaway condition as a bad tube. Fortunately it can be tested.

Remove the offending tube, with or without good parts in the cathode circuit, connect a voltmeter's positive lead to pin 5 of the offending tube socket, and the negative lead to ground. Power up the amp, the voltmeter will rise, then return to ZERO. A positive voltage reading indicates a leaking coupling cap. This is RARE with new production modern parts, but common with old stuff, especially those with paper, or paper in oil dielectric. If the cap is found to be leaky, I would change the cap in the other channel too, it came from the same place.

was considering beefing them up the 7w to 10w as a nervous precaution.

A 5 watt resistor should be fine as long as no bad tubes or coupling caps are encountered. I have seen those white sand cast resistors eat enough power to change color and crack, yet still work fine.

If your tube red plated for no obvious reason like a shorted coupling cap, or reverse installed cathode bypass cap, the tube should be suspect and not trusted. A runaway tube will eventually blow a 10 watt resistor and maybe take down a rectifier tube, OPT or power transformer before it gets bad enough to blow the line (mains) fuse. Attempting to find a fuse value that will blow with a runaway tube (1.6 amp slow blow on US power with an Allied 6K7VG power transformer) just leads to frustration because it will blow when there is a lightning strike 10 miles away.
 
i started at the tube socket, and the good side was showing 220k, while the other side wandered. i then clipped my leads directly to the 220k. my 220k and it is indeed bad. what a cheap little part to cause such a catastrophic result.

everything in this amp is brand new, i would hope it is not the coupling caps as they are not exactly the cheapest part in the amp :)
i will do further measurements on the coupling caps tonight, to verify proper functionality.

just a quick question, the 63 volt cap was powered up, but never ran for more than 30-45 seconds while testing(at the excessive voltage, of course). do you still think it is worthwhile to change?

just curious, but shouldn't the fuse have popped before the other items failed?
 
Last edited:
my 220k and it is indeed bad.

Test it after it is removed from the amp. Those little guys rarely die, but a leaky coupling cap can cause the same erroneous readings when tested in the amp. Try the voltage on pin 5 test before installing the tube. If there is any positive voltage that coupling cap is likely bad.

I just noticed the Jupiter paper and wax caps. I know that they are expensive and I have never tried them for that reason, but paper and wax caps are known to be troublesome due to the paper's ability to absorb moisture and retain it. Unless the paper is baked in a low humidity environment during manufacture, there is some residual moisture trapped inside. It all finds its way to one place after it's sealed up inside a container (the cap) then heated from the outside (the amp). Even a few megohms of leakage will cause a red plated tube.

A really bad cap will show on an ohmmeter. It will read a low value initially but should rise to infinity within a few seconds. Any reading at all is a bad cap. This can not be tested with the cap installed. At least one lead must be disconnected from anything. Even if the cap reads infinity on a hand held meter it could still have a small leakage that is beyond the meter's range.

It is possible that the 63 volt cap survived the high voltage, but you stated that it got hot. That means the the oxide barrier broke down and current flowed through the cap. It is possible for the barrier to reform and the cap to heal, but consider what happens if the cap fails. If it goes open or gets weak, the sound suffers. If it gets leaky (the most likely possibility) it will eventually fail to a short, causing a red plate runaway. If it was my amp I would replace the cap, or at least order a spare and swap it out when it comes in. It might save some more frustration down the road.

The red plated tube could possibly be good. The only way to tell would be to try it in the good channel. Wire a voltmeter across the cathode resistor using clip leads. Note the voltage reading with the good tube, then install the suspect tube. Power up and keep an eye on the meter if it goes more than 5 volts higher than the good tube quickly kill the power.

It is difficult to size a fuse such that it will blow on an overload such as this and not blow on a power surge, or even amp turn on. A 1 amp slow blow fuse is probably the correct fuse to use in a 240 volt country. A 3/4 amp fuse will blow needlessly for several reasons, and there isn't much in between.
 
Test it after it is removed from the amp. Those little guys rarely die, but a leaky coupling cap can cause the same erroneous readings when tested in the amp. Try the voltage on pin 5 test before installing the tube. If there is any positive voltage that coupling cap is likely bad.

I just noticed the Jupiter paper and wax caps. I know that they are expensive and I have never tried them for that reason, but paper and wax caps are known to be troublesome due to the paper's ability to absorb moisture and retain it. Unless the paper is baked in a low humidity environment during manufacture, there is some residual moisture trapped inside. It all finds its way to one place after it's sealed up inside a container (the cap) then heated from the outside (the amp). Even a few megohms of leakage will cause a red plated tube.

A really bad cap will show on an ohmmeter. It will read a low value initially but should rise to infinity within a few seconds. Any reading at all is a bad cap. This can not be tested with the cap installed. At least one lead must be disconnected from anything. Even if the cap reads infinity on a hand held meter it could still have a small leakage that is beyond the meter's range.

It is possible that the 63 volt cap survived the high voltage, but you stated that it got hot. That means the the oxide barrier broke down and current flowed through the cap. It is possible for the barrier to reform and the cap to heal, but consider what happens if the cap fails. If it goes open or gets weak, the sound suffers. If it gets leaky (the most likely possibility) it will eventually fail to a short, causing a red plate runaway. If it was my amp I would replace the cap, or at least order a spare and swap it out when it comes in. It might save some more frustration down the road.

The red plated tube could possibly be good. The only way to tell would be to try it in the good channel. Wire a voltmeter across the cathode resistor using clip leads. Note the voltage reading with the good tube, then install the suspect tube. Power up and keep an eye on the meter if it goes more than 5 volts higher than the good tube quickly kill the power.

It is difficult to size a fuse such that it will blow on an overload such as this and not blow on a power surge, or even amp turn on. A 1 amp slow blow fuse is probably the correct fuse to use in a 240 volt country. A 3/4 amp fuse will blow needlessly for several reasons, and there isn't much in between.


bad news for me. one coupling cap hovers around 20mv or so, while the other is letting 420 volt through.

perfect storm for this little amp.
 
I would put the amp together without the offending coupling cap. Then test it to see if there is any red plate. If no red plate happens, you have truly found the culprit. If red plate still happens, then the tube could still be bad, cooked to death by the leaky cap. Either way the supplier might replace the bad cap.

Intel just sent me a free $300 processor chip as a replacement for a 2 year old chip that had died. It does however expressly state "3 year warranty" on the box, so they had to give me a new one......I expected a bunch of excuses, but got none, just a replacement chip.
 
To wrap this up, I contacted Jupiter, and they were alarmed to hear of the failure. They sent me 2 new caps. I replaced the bad ones, which they want back for testing, and also replaced the 1500 63v again. Double checked everyone else and just ran it without issues for 8 hours. Now I just wonder why I'm still popping fuse's on startup..
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.