bent neck CRT pins...

[stj]

Re: bent neck CRT pins...

if one of the diodes or caps in the HV ladder circuit breaks down under load it will pull the input to ground.
can you lift the ground at pin10 and put a lamp inline.

[redwire]

Re: bent neck CRT pins...

If the CRT is biased-on hard, it could overload the HV after it warms up?
I would be tempted to run it with the cathode (-1,900V) pin 2 disconnected, but anode connected. I think that would not allow the CRT to conduct, but the HV should come up, unless some short internal to the CRT.
Assuming the CRT end-connector on, for filaments etc. in all this.

Mildly interesting, note 8 minute ramp on U1890A/Q1980 from R1991.

[eccerr0r]

Re: bent neck CRT pins...

if one of the diodes or caps in the HV ladder circuit breaks down under load it will pull the input to ground.
can you lift the ground at pin10 and put a lamp inline.

That would be a neat experiment, alas I would end up having to remove the whole multiplier from the board (or cut a hole into the board) to disconnect pin 10 as it's underneath the module :-(
The module has 7 pins underneath: 4 are the corners (pin 13 and 15 are two of the four) and the other three (9, 10, and 12) are sprinkled in the middle. Don't ask me why pin 14 is along the sides of the module however, the pin numbering makes no sense

I was staring at my junk box a bit. I have 66 megohm resistor. If I had 20KV across it, it would dissipate about 6 watts of power... not sure where this resistor came from but I figure it was from the CRT days, and no wonder why this resistor is physically so large... I think it was for bleeding HV.

[eccerr0r]

Re: bent neck CRT pins...

New observation: I think this thing is frying my LM1458's ... It's coming up dead short once again even without the anode connected... I have a feeling perhaps the initial guess was right: HV is arcing internally against cathode/grid circuitry and leaking out killing the LM1458?

Hmm... sigh. I wonder if it's worth to fix this. The short period of time I had this thing seemingly working, it did detect and showed a trace when in1 was connected to the calibration output, but it did seemed sinusoidal instead of square. I just wonder how much I should sink into this thing before giving up on this... are these really reliable scopes and will I just run into issue after issue? (Then again what can one expect from a 30 year old scope... for it to work as well as my digital-less ~40 year old Vu-Data ~20MHz scope?)

[eccerr0r]

Re: bent neck CRT pins...

Hmm..Suppose I buy a working PSU. What can I do to it to protect it from a bad CRT if in case the CRT did damage the PSU? I know the grid and cathode are not shorted (else the protection neon lights won't come on) but is there any other protections?

Still wondering if I should get HVPS + CRT or just HVPS by itself...

Dang these unfinished projects... must get this completed...

[redwire]

Re: bent neck CRT pins...

Fixing a Tek scope is no walk in the park. Don't let the beast defeat you!
I would say the HV is overloaded.
The error amp U1956 (LM1458) can't really fry, it has series output and input resistors for protection. Even an arc internal to the HV module would just hit R1944 and the 1458 has internal clamp diodes.
U1956A is an integrator (no DC feedback) part of the main oscillator I think. Waveform #70 shows it's a 40kHz sine-wave oscillator for the flyback, low EMI and noise, so not a flyback circuit really.
You might have a loose connection somewhere or an intermittent part, sometimes that throws me off when things change.

Gently check the ferrite on the flyback T1970 and make sure there are no cracks or loose clamps. Any extra air gap would cause Q1981 to get hot.

The CRT (rear) socket look for any arcing or carbon tracking, on its PCB and inside the socket between pins, with a magnifier. This can resemble a shorted CRT.

Does the HV come up with the CRT socket unplugged? It should regulate at -1.9kV with no load and crackle on the 2nd anode (+14kV).
Does the HV go down as the CRT warms up? Then the CRT could be biased wrong and drawing too much current.

Next I would be disconnecting rails, like the -300V etc to try narrow down where the overload is. Just offering some ideas here.

[eccerr0r]

Re: bent neck CRT pins...

Thanks, yeah, I think pretty much everything these days are difficult, though mostly is my fault (though I did not bend the pin to begin with!). I think I did crack the tube, but now I just wonder how powering up a gassy tube would cause the psu to fry if it really did. As far as I can tell it's something between the "flyback" (yeah, this thing definitely isn't a true flyback as there's no horizontal retrace that's activating it, but it does depend on the switching of the transistor... as it's not heatsinked it better be saturating.)

The ferrite looks OK. The only thing that's loose is the external winding for the filament.

I can't seem to find any forms of arcing but I did hear something as the device failed. But no more as the PSU is completely overloaded now and blows fuses.

I think I've come to the end of the road on this. I think the unbent pins likely caused the demise of the CRT and it dominoed to the HVPSU.

[redwire]

Re: bent neck CRT pins...

I've only seen CRT's fail shorted or leaky on the aquadag. Usually arc/carbon track on the inside.

The tube will hiss for a long time if you trashed the pins and started a leak. Like a can of pop but vacuum. I'm not sure what the mystery noise was.
Maybe air got in and the aquadag died?

You could try a megger or hipot tester on some of the CRT's pins, but not heater/cathode/control grid, they are all low voltage compared to 1st, 2nd anode, focus and deflection etc.

[stj]

Re: bent neck CRT pins...

i already pointed out that if you lose the vac, the phospor will look bad on the front.

[R_J]

Re: bent neck CRT pins...

You will only lose the phospher if the air rushes in fast, ie break the neck off, if its a slow leak the phospher will remain untouched. I would connect the heater to a filiment transformer with a 6.3 volt secondary and see if it lights and most of all stays lit. if there's air in the tube the heater will most likely burn out.
during crt rebuilding air is let into the tube slowly and then the old gun assembly is cut off and a new one is installed and the tube evacuated, if the vacuum is released to fast it can damage the phospher and the tube is no longer useable.

[eccerr0r]

Re: bent neck CRT pins...

I don't know how fast the air leaked in, it must have been really slow because I did not hear any hissing. I just heard glass cracking, I assume it was glass by the pin cracking.

I mainly am coming to this conclusion because the getters look bad right now, starting to clear up which is a telltale sign

I was wondering whether I should have coated the interface between the wire and the glass with either epoxy or cyanoacrylate before trying to bend it back so that if it leaks it would suck in the glue first, if it's a minor leak... But I did not...

[R_J]

Re: bent neck CRT pins...

Where did you come up with the getter being able to show that there is air in the tube?
Here I found this:
In vacuum or gas-filled tubes, it is a small, ring or cup-shaped device containing a powdered metal that reacts strongly to oxygen. When the tube is sealed, the getter is fired (heated) to further evacuate a vacuum tube or to remove impurities from the gas. Firing causes the getter material to oxidize and absorb any free oxygen in the tube, which would otherwise oxidize the active electrodes and decrease the tube's life. Tubes with graphite-based electrodes do not use getters, because the graphite itself absorbs oxygen. In a vacuum tube, the location of the getter is sometimes identified by a shiny silver deposit behind the glass, which is residue from the firing. The metal used in the getter depends on tube manufacturer, type and composition. Barium is an excellent oxygen absorber and is used in tubes with a fired getter. Zirconium and titanium, also used as getter material, oxidize from the tube's self-generated heat and do not require firing.


The only real way to check the condition of the tube is electronically.

I have seen bent crt pins where the the glass around them cracks a bit but did not cause the tube to be bad. the glass can crack a bit when the pin is stressed but usually only only at the surface where the glass meets the pin.

Here is some additional information reguarding tubes http://www.vacuumtubes.net/How_Vacuum_Tubes_Work.htm

[eccerr0r]

Re: bent neck CRT pins...

Hmm... I was still staring at the getter spots along the side of the neck and am seeing a bit of whiteness in it where it should be silvery - which seems to imply that the seal has failed. I'm looking into removing the tube now to confirm (need to remove from the shield to look at the spots straight on) but currently I have to look into the shield from a very steep angle (say, 5 degrees from parallel to the spot) with a flashlight and apparently seeing the not so healthy getter spots.)

[R_J]

Re: bent neck CRT pins...

I give up

[Sparkey55]

Re: bent neck CRT pins...

Hmm... I was still staring at the getter spots along the side of the neck and am seeing a bit of whiteness in it where it should be silvery - which seems to imply that the seal has failed. I'm looking into removing the tube now to confirm (need to remove from the shield to look at the spots straight on) but currently I have to look into the shield from a very steep angle (say, 5 degrees from parallel to the spot) with a flashlight and apparently seeing the not so healthy getter spots.)

Like someone else suggested here, you need to test the crt heater filament. The safest method would be to connect four 1.5 volt batteries in series, D cells would work to give you 6 VDC. If the heater does not light up a dull red or bright orange then it is dead. IF the filament checks good then leave the HV anode cable off but connect the focus lead to the crt then watch the tube in the dark for any corona or purple flashes.

[eccerr0r]

Re: bent neck CRT pins...

How long would it normally take to "kill" it, assuming the vacuum is gone? I guess I do have a variable PSU that I can connect to it, but at least right now the filament is intact based on resistance. Oddly enough I can't see the glow of the filament due to the shield and the socket covering the back end so I have to infer by resistance.

[Sparkey55]

Re: bent neck CRT pins...

How long would it normally take to "kill" it, assuming the vacuum is gone? I guess I do have a variable PSU that I can connect to it, but at least right now the filament is intact based on resistance. Oddly enough I can't see the glow of the filament due to the shield and the socket covering the back end so I have to infer by resistance.

If the crt vacume is lost the heater will burn to an open circuit when power is supplied to it, just like when a light bulb is cracked. Disconnect the socket from the crt neck so you can see the filament glow when applying external 5 to 6 volts to the heater pins on the crt. DO NOT go over 7 volts on the heater at any time. If you can not see the filament then try to measure the current draw with a meter. Do not have the scope powered on while doing this test.

[Sparkey55]

Re: bent neck CRT pins...

If the crt vacume is lost the heater will burn to an open circuit when power is supplied to it, just like when a light bulb is cracked. Disconnect the socket from the crt neck so you can see the filament glow when applying external 5 to 6 volts to the heater pins on the crt. DO NOT go over 7 volts on the heater at any time. If you can not see the filament then try to measure the current draw with a meter. Do not have the scope powered on while doing this test.

If the crt can not pass this simple heater filament test then it makes no sense to go any further on testing the tube or HVPS. Will you perform the test as suggested?

[eccerr0r]

Re: bent neck CRT pins...

I tried passing 5.6VDC with a LM317T regulator. Current started around 134mA but went down to 125mA and is still slowly going down, then again regular light bulbs do this too. I tested for 2 minutes....

[momaka]

Re: bent neck CRT pins...

If the crt can not pass this simple heater filament test then it makes no sense to go any further on testing the tube or HVPS. Will you perform the test as suggested?

+1

Unless you somehow plan on evacuating the tube again.

I've been actually wanting to try that with my dead GDM-FW900. But for CRTs, you need a really good vacuum pump: usually dual diaphragm and capable of evacuating down to less than 100 microns. Also, even if you had the equipment for that, you'd have to get creative with where to make a hole on the tube to pull the vacuum from and then stop air into coming back in (unless you plan to run the pump continuously with the CRT). Something like this experiment:
https://www.youtube.com/watch?v=jlV-2gcO0_0