> Position #1 is killing tubes.
Bad connection between bias source and grid 1 pin. Anywhere along there, including the socket.
Bad connection between bias source and grid 1 pin. Anywhere along there, including the socket.
You'll have to dig in there and hope there's something obvious. Socket damage is definitely something obvious, but it could be something more insidious like a screen stopper resistor that's cracked but still looks fine or a coupling cap that's acting up.
The oil-film colours just suggest the getter layer is only a few wavelengths of light thick on that tube. The others are presumably thicker. Probably the getter was defective or not flashed long enough?
Though that's rather suspect as an explanation as the getter is supposed to be metalic barium, oil-film colours form with a dielectric rather than a metal, so perhaps its barium oxide or nitride already? Perhaps some difference in the glass envelope?
I'd suspect that tube and monitor it carefully.
Though that's rather suspect as an explanation as the getter is supposed to be metalic barium, oil-film colours form with a dielectric rather than a metal, so perhaps its barium oxide or nitride already? Perhaps some difference in the glass envelope?
I'd suspect that tube and monitor it carefully.
Does the amplifier work fine with this tube installed? Does the plate turn red when it is on?
I suggest to double check tightness of pin #5, and if it is fine, install a new tube in this location only, set bias, and carry on.
Last edited:
I have not read all of the posts in this thread however in the first post the tube with the effect is fitted in a location with less ventelation holes around it.
Perhaps it is just that the glass has got a little bit hotter than the others allowing the getter material to migrate slightly.
Perhaps it is just that the glass has got a little bit hotter than the others allowing the getter material to migrate slightly.
Thx
Hi. Amp worked fine with it. I couldn’t tell if it was red-plating, ‘cause stupidly put the cage on (to protect my cat))). Have to check this. Just ordered a new tube. I hope it does not go the same way. I checked and re-tightened all ping’s on all sockets. I wonder if I can spray them with Radio Shack contract cleaner? I checked and re-checked all the voltages around each socket - all are spot-on.
Hi. Amp worked fine with it. I couldn’t tell if it was red-plating, ‘cause stupidly put the cage on (to protect my cat))). Have to check this. Just ordered a new tube. I hope it does not go the same way. I checked and re-tightened all ping’s on all sockets. I wonder if I can spray them with Radio Shack contract cleaner? I checked and re-checked all the voltages around each socket - all are spot-on.
Older Defy 7 sockets
So in my search for older Defy 7 sockets got this far:
1. contacted Jadis dealer. Was told Jadis only supplies white ceramic sockets. From pictures they look like regular ebay Chinese variety. 2. dealer told me they have some original style but in brown, at a mere $39 a piece )))) 3. was able to read part number of the old sockets. They are Omron PLE 08-0 relay sockets, available everywhere at about $5 - $6, and on sale for $25 for 10 pcs )))). Will get them just in case.
Wonder if I should modify this thing with suggested lowering of bias resistors from 474K to 200K, or even lower? Modify to auto-bias? Lower B+ from 485v to about 400v? Install a plate fuse right after 4.7 ohm resistor before the smoothing caps? Or just sell the damn thing?
So in my search for older Defy 7 sockets got this far:
1. contacted Jadis dealer. Was told Jadis only supplies white ceramic sockets. From pictures they look like regular ebay Chinese variety. 2. dealer told me they have some original style but in brown, at a mere $39 a piece )))) 3. was able to read part number of the old sockets. They are Omron PLE 08-0 relay sockets, available everywhere at about $5 - $6, and on sale for $25 for 10 pcs )))). Will get them just in case.
Wonder if I should modify this thing with suggested lowering of bias resistors from 474K to 200K, or even lower? Modify to auto-bias? Lower B+ from 485v to about 400v? Install a plate fuse right after 4.7 ohm resistor before the smoothing caps? Or just sell the damn thing?
for me modify to autobias per cathode, and lowering to 440vdc is a good step, this way when you use cathode resistor bias, the plate voltage will be lowered..
per my experience, the difference between cathode resistor bias and fixed bias is not night and day, you may not even notice...
per my experience, the difference between cathode resistor bias and fixed bias is not night and day, you may not even notice...
Thank you
My knowledge is not that advanced. I will need to do a lot more research and will need a schematic. I have the schematic for the amp.
With respect to the other posters, I think it is premature to suggest modifying the amplifier, or replacing parts, without some concrete evidence of what is wrong. For one thing, modifications will ruin the value of an expensive audiophile amp. I could say sell it and buy an ARC, but that is also a useless response 🙂
If this is a defy 7, the schematic says the cathode of each output tube is fused with a 160mA fuse. Make sure these fuses are installed and are the correct value. If so, they will prevent chernobyl of the the output stage regardless of what happens.
I think the best course here is to put in the new tube, set bias, and just use it for a while under observation. If all is well after a few days, keep using it and the eventually buy some matched tubes once you are absolutely satisfied it is reliable and have saved the money for them (or maybe never?).
Another thing to check is bad solders on the board. I don't say to touch it up if it is good, just inspect all the solder around the suspect tube for cracks or cold/dry joints, and fix if required. This step is optional but may make you sleep better at night.
Yes you can use radio Shack contact cleaner, just a little bit in each hole, and then insert and withdraw the tube a few times, and wipe away the excess (if any) from around the socket with a rag.
The evidence here tells me that one tube went bad, so the owner threw in a used KT88 which was close enough, and got on with life. This gives me the impression that nothing significant is wrong with the amp.
Last edited:
Thank you very much
Amp was abused before me. The fuses were all 200ma and one even 250. Two resistors were burned (1k and 180 ohm) and hastily replaced on the bottom side of the board. These amps are famous for burning resistors while leaving fuses intact. There is an article on general fusing of output stages which suggests putting a fuse on anode supply right after the 4.7 ohm resistor before the smoothing cap. As to the amp’s value - input stages are moded and I don’t want to go in there. They made the input tubes run “hotter” with 250v instead of 130 on plates. I will follow your advice and install new tube (ordered yesterday) and observe. Should I run the amp at lower bias first? About half?
i do not see why fuses needed to be used....a fusible resistor would have been better...
yes, run the amp with the lowest bias you can get away with consistent with good distortions,
there is something about chinese sockets, even their ceramic onse, could be the metals they used...with the "sato" japanese sockets, it takes years and years before any issues involving them are found...
that is why i go for cinch sockets even if used, they are rock steady...
the input tube running at higher plates means more clean output swings...
Replace the fuses with the proper value, and run bias on spec. Don't worry about previous modifications for now. If you cannot source 160mA fuses you can keep the bigger t ones for now.
I agree with Tony that Chinese sockets are worse than other brands, but this amp is made in France and may have good sockets for all I know.
Yes fuses are inferior in some ways to fusible resistors, but they are easier to replace, so I'd just leave well enough alone.
Thank you
So I will leave the input alone.
Cinch is great, except I couldn’t find any PCB mounts by them. I will order some new original ones.
Jadis recommends to set bias at 4 to 5 v measuring between fuses and ground. Which divided by 180 ohm cathode resistor gives between 22 and 27 ma. I set it at 2 v for now.
Amp
I have found 160 ma fuses, but also have a ton of 200 ma.
The sockets are these:
PLE08-0 Omron Automation and Safety | Relays | DigiKey
I have found 160 ma fuses, but also have a ton of 200 ma.
The sockets are these:
PLE08-0 Omron Automation and Safety | Relays | DigiKey
2V is way too low, you will definitely get crossover distortion. The Jadis value is actually quite low already, I would set to 4.5V and leave it be.
While it has been 20+ years since I serviced one (as you know), each tube has an individual bias resistor and if you reduce them to reasonable values for the 6550, the driver stage will not be able to drive the output stage properly.
Some of these amplifiers had fuses in the cathode circuit, presumably yours doesn't.
Whatever charm these amplifiers have is in the funky circuit design and output transformers, I would not modify them other than perhaps to reduce the raw AC supply to no more than 117 using a buck transformer.
I would make a jig to measure grid current and check each tube under the existing operating conditions for grid current. Select those with lowest grid current for use in the amplifier. (Get some extras)
FWIW most modern KT88 use the same internal structures as modern 6550 (there are probably exceptions) and there really isn't a large difference between them unlike MO KT88 and a GE 6550 4 - 5 decades ago..
The SV6550C if you can find NOS are probably about as good as you are going to find for these amps now.