I grafted one of those GLY513 ASC 600 uF capacitors into the power supply of my diy SE amp based on Mikael Abdellah's KT88 SE amp.
Capacitor Reference: http://www.diyaudio.com/forums/showthread.php?postid=1887003#post1887003
The original power supply was CLC at 50uf, 5H, 235 uF. The 600 uF ASC capacitor replaced the 235 uF electrolytic capacitor. PSUD II didn’t show any potential problems with the 600 uF ASC cap substitution.
After the installation of the ASC capacitor, I started off the evening playing a few CDs with no problems. But on about the fourth CD, I saw a brilliant white flash out of the periphery of my vision. I turned my attention to the amp and saw the flash (arcing) again through one of the plate alignment holes on one of the EH KT-88’s. It’s hard to tell for sure but the arcing seemed to be coming from the middle of the screen or grid. The arcing occurred a couple more times in synch with some bass notes. I quickly turned down the volume and the arcing stopped. I never heard any crackling or distortion through the speakers when the tube arced. I turned the volume back up to almost the same level and listened to a different CD for about an another hour with no further occurrence of the arcing.
The EH KT-88’s are over a year old (out of warranty) but are low hour usage (less than 25 hours). I checked the bias at each EH KT-88 and they were both identical and steady at 81 ma. B+ is 390 volts. Plate dissipation is a conservative 28 Watts. I turned the lights out and I could not see any hot spots on the screens or grids. There’s no visual gas glow and with the steady bias, I’m assuming that the KT-88 that arced is not gassy. The KT-88’s are triode connected through a 100 ohm, 5W resistor connected plate to screen. I have a 220k on the KT-88 grid to ground and 100 ohm CC on the grid of the KT-88. OPTs are transcendar 5k into 8 ohms.
I had never seen the arcing with the original power supply. Any ideas on the cause of the arcing? Was it the ASC capacitor and if so, why? Or is it due to some other cause that coincidently showed up with the capacitor change? Could it have been due to some transient oscillation? Do I need to increase the value of my screen resistor? Do I need to lower the value of the 220k grid to ground resistor? Is the KT-88 on its way out?
Capacitor Reference: http://www.diyaudio.com/forums/showthread.php?postid=1887003#post1887003
The original power supply was CLC at 50uf, 5H, 235 uF. The 600 uF ASC capacitor replaced the 235 uF electrolytic capacitor. PSUD II didn’t show any potential problems with the 600 uF ASC cap substitution.
After the installation of the ASC capacitor, I started off the evening playing a few CDs with no problems. But on about the fourth CD, I saw a brilliant white flash out of the periphery of my vision. I turned my attention to the amp and saw the flash (arcing) again through one of the plate alignment holes on one of the EH KT-88’s. It’s hard to tell for sure but the arcing seemed to be coming from the middle of the screen or grid. The arcing occurred a couple more times in synch with some bass notes. I quickly turned down the volume and the arcing stopped. I never heard any crackling or distortion through the speakers when the tube arced. I turned the volume back up to almost the same level and listened to a different CD for about an another hour with no further occurrence of the arcing.
The EH KT-88’s are over a year old (out of warranty) but are low hour usage (less than 25 hours). I checked the bias at each EH KT-88 and they were both identical and steady at 81 ma. B+ is 390 volts. Plate dissipation is a conservative 28 Watts. I turned the lights out and I could not see any hot spots on the screens or grids. There’s no visual gas glow and with the steady bias, I’m assuming that the KT-88 that arced is not gassy. The KT-88’s are triode connected through a 100 ohm, 5W resistor connected plate to screen. I have a 220k on the KT-88 grid to ground and 100 ohm CC on the grid of the KT-88. OPTs are transcendar 5k into 8 ohms.
I had never seen the arcing with the original power supply. Any ideas on the cause of the arcing? Was it the ASC capacitor and if so, why? Or is it due to some other cause that coincidently showed up with the capacitor change? Could it have been due to some transient oscillation? Do I need to increase the value of my screen resistor? Do I need to lower the value of the 220k grid to ground resistor? Is the KT-88 on its way out?
Thanks Russ for the comments.
I ran the amp today and played the same song again that caused the arcing but this time the KT-88 behaved itself. The song was only Jack Johnson’s “Upside Down”. Not like I was playing the “1812 Overture” or anything like that.
I thought I was operating the KT-88’s conservatively. I guess I need to go back and review the KT-88 screen characteristics.
I ran the amp today and played the same song again that caused the arcing but this time the KT-88 behaved itself. The song was only Jack Johnson’s “Upside Down”. Not like I was playing the “1812 Overture” or anything like that.
I thought I was operating the KT-88’s conservatively. I guess I need to go back and review the KT-88 screen characteristics.
It's an SE amplifier therefore class B operation does not enter into it.
I fail to see any very likely relationship between the replacement of the existing capacitor with the new one wrt to your arcing tube, true the voltage will sag somewhat less short term than with the 235uF capacitor, but the supply voltage did not increase significantly - again since this is class A operation I would expect no significant changes in amplifier power capability or current draw.
Where the cap is likely to provide some significant benefit is in a lower source impedance at low frequencies and some reduction in ripple present at the output.
The most likely explanation is that when working on the amplifier you dislodged some internal contamination in that tube which resulted in a temporary short between the plate and the screen grid or suppressor assemblies in the output tube. (Possibly you were clipping the amplifier resulting in some additional voltage stress as primary inductance in the transformer generates a rising EMF as it attempts to maintain the normal operating current level and it arced to the suppressor. The other possibility is you melted some part of the screen grid, and that is what you saw. All of this is just a SWAG..
)
EH tubes are OK, but I recently had a EH KT88 self destruct (major arc over, and I mean major!) in one of my voltage regulators so I am not inordinately fond of them. (delta plate to cathode of ~300V and 80mA load current)
FWIW I would err on the side of caution and just replace the tube.
I fail to see any very likely relationship between the replacement of the existing capacitor with the new one wrt to your arcing tube, true the voltage will sag somewhat less short term than with the 235uF capacitor, but the supply voltage did not increase significantly - again since this is class A operation I would expect no significant changes in amplifier power capability or current draw.
Where the cap is likely to provide some significant benefit is in a lower source impedance at low frequencies and some reduction in ripple present at the output.
The most likely explanation is that when working on the amplifier you dislodged some internal contamination in that tube which resulted in a temporary short between the plate and the screen grid or suppressor assemblies in the output tube. (Possibly you were clipping the amplifier resulting in some additional voltage stress as primary inductance in the transformer generates a rising EMF as it attempts to maintain the normal operating current level and it arced to the suppressor. The other possibility is you melted some part of the screen grid, and that is what you saw. All of this is just a SWAG..
) EH tubes are OK, but I recently had a EH KT88 self destruct (major arc over, and I mean major!) in one of my voltage regulators so I am not inordinately fond of them. (delta plate to cathode of ~300V and 80mA load current)
FWIW I would err on the side of caution and just replace the tube.
Thanks, Kevin.
Your contamination theory seems plausible since I haven’t been able to recreate the problem. I’ve convinced myself that it is the tube and I have already taken the tube out of service. It bothered me that I made a change and then something bad happened. I now believe the events were purely coincidental.
Your contamination theory seems plausible since I haven’t been able to recreate the problem. I’ve convinced myself that it is the tube and I have already taken the tube out of service. It bothered me that I made a change and then something bad happened. I now believe the events were purely coincidental.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.