Mighty Monarch of The Air

[vts1134]

Got a good supplier for a 500pf 20,000v door knob capacitor?

[VintagePC]

Digi-key has a few... for example:
http://search.digikey.com/scripts/Dk...me=445-3890-ND
http://search.digikey.com/scripts/Dk...me=445-2235-ND


... but they won't be cheap :S

[old_coot88]

Well sunnava gun, so it was the doorknob all along, and now the filament winding is popping through. Some kind soul oughta send him a hunk of HV wire to replace that with, along with a doorknob.

[vts1134]

So as I sit here and ponder I can't help but think that the door knob cap ISN'T shorted to ground. If that cap was shorted to ground it would certainly account for our problem, but could just a leaky cap cause our problem? If not could it be the ground side of the cap some how isn't grounded? Or some other problem other than the cap itself being bad. I just know that some times in my work when I think I've found the answer from trouble shooting I just end up "fixing" things that aren't broken and still not finding the real problem.

[old_coot88]

OK, did the CRT light up with a raster after you removed the doorknob? If so, it would pretty well prove the doorknob was at fault. Probably arcing internally since it measured open circuit on your ohms test.

Earlier you mentioned a "new arc" that's developed. I (probably erroneously) interpreted it to mean there's a hole in the insulation of the filament winding that's allowing a DC arc. That would be evinced by a loud, sharp snapping sound. If instead the "arc" is a bluish, lightly hissing corona discharge from the 'tire' to the filament loop, you may be able to insulate it with some silicone. Is it possible to move the filament loop further down on the frame, away from the 'tire'?

[vts1134]

I have a new theory. The old 1X2 tube was fine. The new 1X2 is bad. The "new" arc I described from the "tire" to the filament loop was always there and I never noticed it. Every time there was a discharge from the tire to the filament loop it was echoed inside the 1X2, making me think the tube itself was bad. I've removed the flyback and there is no insulation missing from the filament loop at all. The "tire" has an area where it looks slightly melty, but only slightly. I'm going to get it nice and insulated with the afore mentioned sensor safe silicone, attempt to get the filament loop as far away from it as possible and reinstall with the old 1X2.

One thing that struck me as odd was there was already three or four pieces of electrical tape squares stuck between the "tire" and the filament loop. Looks like the problem could have been there before when the set was in use.
This theory makes absolutely the most sense to me so far.

[vts1134]

Originally Posted by old_coot88

OK, did the CRT light up with a raster after you removed the doorknob? If so, it would pretty well prove the doorknob was at fault. Probably arcing internally since it measured open circuit on your ohms test.

The CRT lit up with a raster before I removed the doorknob, and after as well.

[Phil Nelson]

A capacitor can leak internally when voltage is applied, yet test as "not shorted" using an ohmmeter. The voltage applied by a general-purpose multimeter is negligible compared to the actual working voltage. That's why they made dedicated capacitor testers. Mine can check a cap at any voltage up to 500v. I haven't heard of any cap checkers that can test in the KV range, however. For a doorknob like this, the test is substituting a known-good cap and observing what happens. Which was no problem in Ye Olden Dayes when the repairman would just fish one out of a bin on the shelf and pop it in.

If it looks like you need a new doorknob, Surplus Sales of Nebraska has one at 500pf/20KV for a low, low price of $35:

http://www.surplussales.com/Capacito...cap_trans.html

Click on the "Available in these styles" link to see whether that type would replace what you have.

It's not a cheap part, so I'd proceed with whatever testing is needed to isolate that as your trouble source.

TVs can have more than problem at a time, of course. If you also want to replace the 1X2 filament loop, I believe that needs to have HV-rated insulation.

The photo shows some sensor-safe RTV that I used on a Philco flyback.

Phil Nelson
Phil's Old Radios
http://antiqueradio.org/index.html

[vts1134]

Originally Posted by vts1134

I have a new theory. The old 1X2 tube was fine. The new 1X2 is bad. The "new" arc I described from the "tire" to the filament loop was always there and I never noticed it. Every time there was a discharge from the tire to the filament loop it was echoed inside the 1X2, making me think the tube itself was bad. I've removed the flyback and there is no insulation missing from the filament loop at all. The "tire" has an area where it looks slightly melty, but only slightly. I'm going to get it nice and insulated with the afore mentioned sensor safe silicone, attempt to get the filament loop as far away from it as possible and reinstall with the old 1X2.

One thing that struck me as odd was there was already three or four pieces of electrical tape squares stuck between the "tire" and the filament loop. Looks like the problem could have been there before when the set was in use.
This theory makes absolutely the most sense to me so far.

FAIL!
No more discharge from the fly to the filament winding but problem is not solved. Every time I get over 95v I get sparks on the inside of the old 1X2, and sparks plus a blue glow on the new 1X2.
Time for a new $35 capacitor.

[old_coot88]

Originally Posted by vts1134

FAIL!
No more discharge from the fly to the filament winding but problem is not solved. Every time I get over 95v I get sparks on the inside of the old 1X2, and sparks plus a blue glow on the new 1X2.

This is with the doorknob removed, right? What's happening to R89, the 470K resistor? Is it getting hot, or remaining cold? If R89 remains cold, there may be a leakage path from the 1X2 socket itself to ground that's dragging down the HV. If R89 is getting hot, it may be that the 1X2 (in this case either one) is gassy and passing unrectified pulse that's hammering the resistor.

Time for a new $35 capacitor.

A new cap isn't going to help until you get the current problem sorted out. All this cap does is smooth the rectified pulse coming out of the 1X2.

One other thing- when you re-installed the flyback, where you soldered the filament loop back to the 1X2 socket, are there any whiskers or sharp points to the joints? In soldering any HV connections it's very important to build a nice 'ball' of solder on each joint to eliminate any points.

[jeyurkon]

He previously said that with the doorknob removed there was no arcing within the 1X2 or glow. So, the doorknob could be breaking down internally causing the arcing within the 1X2.

The arcing from the filament loop is another issue.

John

Originally Posted by old_coot88

This is with the doorknob removed, right? What's happening to R89, the 470K resistor? Is it getting hot, or remaining cold? If R89 remains cold, there may be a leakage path from the 1X2 socket itself to ground that's dragging down the HV. If R89 is getting hot, it may be that the 1X2 (in this case either one) is gassy and passing unrectified pulse that's hammering the resistor.

A new cap isn't going to help until you get the current problem sorted out. All this cap does is smooth the rectified pulse coming out of the 1X2.

One other thing- when you re-installed the flyback, where you soldered the filament loop back to the 1X2 socket, are there any whiskers or sharp points to the joints? In soldering any HV connections it's very important to build a nice 'ball' of solder on each joint to eliminate any points.

[vts1134]

Originally Posted by old_coot88

What's happening to R89, the 470K resistor? Is it getting hot, or remaining cold? If R89 remains cold, there may be a leakage path from the 1X2 socket itself to ground that's dragging down the HV. If R89 is getting hot, it may be that the 1X2 (in this case either one) is gassy and passing unrectified pulse that's hammering the resistor.

Stupid question, but I'm not afraid to look stupid to get it right. "Hot" and "cold" aren't euphemisms right? If they are not euphemisms then would a simple finger test be recommended for a thermometer (after discharging the HV of course)? How "hot" is hot and how "cold" can it run? Also how long should the set heat up before testing temp?

Originally Posted by old_coot88

This is with the doorknob removed, right?

My results of:
Over 95V I get sparking on the inside of the old 1X2, and sparking plus blue glow on the new 1X2 were with the doorknob installed.
Previously when I thought I had removed the doorknob from the equation I had infact not. Tonight I removed, completely, the doorknob from the set. The results with the doorknob removed:
Blue glow on the new 1X2 (although not as strong as when the capacitor is installed). No glow on the old 1X2. After leaving the voltage around 100V for 70 seconds or so both tubes made a pffffssst noise. I fired the set down immediately after hearing this.

The only things that are connected post 1X2 at this point are the filament loop, two resistors, and the anode connection on the picture tube. I've tested the set without the anode connection to the picture tube and it failed. Both resistors measure within 20%. The filament loop is solid and double insulated where is loops through the flyback. While the doorknob was removed I took the time to make sure all solder points on the base of the 1X2 are smooth. I'm thinking that the only two things that remain are the 1X2 itself, and the flyback output to the 1X2. I'm going to try another 1X2. As far as the flyback output is concerned, I will obtain a high voltage probe to measure the output. Does any one know what the voltage should read out of the flyback? Hopefully this will pinpoint the exact problem.

[Phil Nelson]

In my unscientific mind, "hot" would be too hot to touch.

Regarding flyback output, the data sheet for that CRT says the typical anode voltage is 12KV. Recommended max = 14KV, min = 10KV.

http://scottbecker.net/tube/sheets/093/1/16GP4.pdf

That would be measuring at the end of the HV anode lead. Since the CRT lit up before, I'm guessing you have several KV, at least.

I wonder if the 1X2 socket could be faulty? Have you carefully inspected it using a strong light, and looking for any cracks, black carbon trails, etc.? If you leave the doorknob disconnected and use a listening tube, perhaps you could tell whether the Pfffft sound comes from the socket.

Phil Nelson

[vts1134]

Maybe I should just replace the tube socket was well. I'll do all of these things one at a time to pinpoint the exact problem.

All learning experiences for the next one .

[old_coot88]

Originally Posted by vts1134

Stupid question, but I'm not afraid to look stupid to get it right. "Hot" and "cold" aren't euphemisms right? If they are not euphemisms then would a simple finger test be recommended for a thermometer (after discharging the HV of course)? How "hot" is hot and how "cold" can it run? Also how long should the set heat up before testing temp?

Well, assuming it's a half-watt, "hot" would mean blistering and probably smoking after a minute or two. Since it's holding value within tolerance and showing no sign of overheating it would be 'cold'.

As far as the flyback output is concerned, I will obtain a high voltage probe to measure the output. Does any one know what the voltage should read out of the flyback? Hopefully this will pinpoint the exact problem.

A HV probe is DC only and will not measure the direct output of the fly, which is pulsed RF. The probe will measure anything 'downstream' of the 1X2 however, which is DC. Since you've already gotten a hefty 'air arc' off the fly and have seen a raster of full width and adequate brightness, this indicates the fly is performing up to par.

Tonight I removed, completely, the doorknob from the set. The results with the doorknob removed:
Blue glow on the new 1X2 (although not as strong as when the capacitor is installed). No glow on the old 1X2. After leaving the voltage around 100V for 70 seconds or so both tubes made a pffffssst noise. I fired the set down immediately after hearing this.

This suggests to me a leakage path from the tube socket itself to ground that's killing the HV and thus leaving resistor R89 cold and unmolested. Is there any way you can temporarily remove the mountings of the socket so it 'floats' free of any connection to ground? This would prove whether or not a leakage path through the mounting(s) exists. The "pffssst" noise certainly sounds like such a leakage condition.

The only things that are connected post 1X2 at this point are the filament loop, two resistors, and the anode connection on the picture tube. I've tested the set without the anode connection to the picture tube and it failed. Both resistors measure within 20%. The filament loop is solid and double insulated where is loops through the flyback. While the doorknob was removed I took the time to make sure all solder points on the base of the 1X2 are smooth. I'm thinking that the only two things that remain are the 1X2 itself, and the flyback output to the 1X2. I'm going to try another 1X2.

I would definitely try 'floating' the socket as the next troubleshooting step if it's doable.