Burning out an intermittant heater-cathode short in a CRT

My B&K CRT analyser / restorer removed shorts by discharging a cap across the affected electrodes. I seem to recall that g1-k shorts were actually more common, due to 'burnt out' emissive material falling off the cathode and lodging in the g1-k gap. The location of the short was indicated by neon lamps, the intensity of the glow giving an indication of the resistance of the path. When the zapping took place, I think that the CRT gun was being run by the analyser as a basic triode. Some quite spectacular fireworks used to occur in the neck, as the contaminating material was vapourised by the discharge across it from the cap. It was very effective at removing inter-electrode leakage paths. The B&K was a fine instrument in its day. I lent mine to someone some years back, and it was never returned. I can't even remember who it was to ...

CRT heaters are not particularly fragile, and are insulated from the cathode cylinder, by a heat conducting coating. If you are discharging between h and k, then the heater pins should be shorted together to avoid one of those 'unfortunate accidents' ... And of course, make sure that the base connector is removed from the tube !!

Arfa

You can't clear a heater to cathode short, no matter what you do. No commercial CRT restorer even attempts it. If you try to discharge a current through the short, you will kill the filament. The only solution is to provice a "floating" heater source that won't drag the cathode low when the contact is made. The heater voltage will actually rise to cathode voltage (above ground), but will remain 6.3 between the filament pins. One method is to buy an isolation transformer from supply houses, which is basically a toroid with two windings on it. The other is to manually wrap some wire around the flyback core, and feed to the filament. In either case, the CRT board must be modified to float the filament pins completely away from ground. Very common procedure done at repair shops.

Hmm, there seems to be some disagreement over whether a H-K short can be fixed. Will the filament also be killed if you tie its pins together? That way, a current path other than through the entire filament always exists.

Floating might work, but not in this case, because the problem exists in two guns. Additionally, it would decrease image quality (at least when the short exists; it might clear up when the short disappears when it's warm), and the reason I want to fix this monitor is because of it's high quality; a quality that not even other T766 models have/had.

Another thing; if I measure the filament voltage, it's 5V. Is that normal deviation from the 6.3 which is common? And, when isolating, is it useful to use some kind of regulation, like with a couple of diodes and an LM317, to make sure the voltage is exactly what it should be?

The short that exists, could that be because of degraded insulation material, or is external debris more likely?

Floating the heater in a monitor is not usually possible because they run the heaters on a regulated DC supply instead of using AC from the flyback. DC obviously won't work with a CRT heater isolation transformer. I have seen other Sony monitor CRTs with 5v heaters, so it may be fine. If your monitor does use AC heaters, then you need a true RMS meter to measure this voltage.

I've seen more H-K shorts than anything else. K-G1 shorts can usually be blown out, but H-K shorts are harder to remove reliably. The problem with fixing a H-K short is that the heater is coated with insulation. If this insulation cracks, then the heater will short to the cathode. Attempting to blow out the short can blow the heater. Also, even if the short goes away, the heater still has damaged insulation and could short again at any time.

A CRT with two shorted guns is probably past hope. If you really want to fix the monitor, start looking for a compatible replacement CRT. Almost any 19" Trinitron monitor CRT should work. Andy Cuffe

snipped-for-privacy@gmail.com

I did isolate the heater on a Sony monitor once years ago, it was a DC heater but I disconnected the plug and then wound a couple turns of wire around the flyback core and used that instead, it worked great for a day, but the cathodes were shot so I added another turn and it was bright again for an hour or so, then it got unusably dim again, added another turn and the heater burned out. The picture did look fine while it worked though.

Also when the guns only short intermittently? And, when it's warmed up, the problem is gone.

The problem with replacing the CRT, is that I'm almost completely sure that the image quality will be lower. When I bought this monitor, it required three trips to the service desk to have it calibrated properly. Out of the box, it sucked. I've also seen other T766 models, which all sucked. If I am going to replace the CRT, I'm going to let an Eizo qualified repair company do it, because they can do a better job of it. Problem is, Sony doesn't manufacture the tubes anymore, so it's possible it can't be repaired. And even if it can, as I said, quality will probably be lower than what I'm used to.

I would look for another monitor and get that calibrated if need be. CRT monitors are cheap these days, plenty of really nice ones are getting replaced by LCDs.

My experience with second-hand monitors is not too good. Also, the Eizo's on the second hand market are all older models. On large Dutch advertising sites, I can't find a single T766. Most of em are older shadowmask models, and it's the trinitron I like.

Eizo states that they keep spare parts for, I believe, 5 years, but I don't know if that includes the CRT. Some might call me nuts for investing money in a CRT monitor, but the fact that to this day, I have yet to see a better monitor, justifies me in it :). The only problem is, how can I get that lazy repair guy calibrate the monitor properly for me, especially when his shift ends in 5 minutes, and not stuff in some old CRT they have lying around... On their website they advertise with how well they repair monitors (warm up period, calibrate of focus and colors, etc), but I've experienced it first hand, it at one time, I got it returned with the colors calibrated to make just about everything green... I wonder how they do such a sloppy calibration.

Just to confirm something: the CRT repair FAQ says this about K-G1 shorts:

connected together, this will affect not only the color of the guilty cathode but the others as well. The result may be a very bright overloaded *negative* picture with little, none, or messed up colors.

But [1] says that cathode to G1 can produce similar symptoms as heater-cathode shorts. I ask, because I noticed that G1 is connected to ground on the CRT socket board. I can imagine that one cathode can very well short to G1 without the others being affected. Is that a right assement?

Can I disconnect G1 to test if it's a cathode-G1 short? There are two pins on the CRT to which G1 connects (which I don't understand BTW).

And, what is the use of G1 being at ground potential...?

BTW, [1] also states a 25% success rate of blowing out cathode-heater shorts with the flyback...

[1]

I just did a test. In off state, I can't measure any short. But when I heat up the heater with an external power supply, I can measure 1K resistance between G1 and any of the cathodes (R, G and B). As the heater heats up, the resistance decreases. The strange thing is, even tough they all read 1K, the continuity meter only read complete short on red (that is, "0". The others read "0.2"). I still couldn't measure anything between heater and cathodes.

So, as far as I can see, it's a cathode-G1 short after all (that is, that 1K resistance should be there, right?). The only thing that stands in the way of this conclusion, is the comment from the repair FAQ above.

Any comments on this are appreciated.

Have you tried running a degaussing coil over the shorted area (set powered down). I recall someone telling me ages ago that he had dislodged a wedged particle causing a short by doing so. Degauss and tap at the same time maybe. Jango.

No, I haven't. But I also don't have one... And, I can remember reading something about that being dangerous for trinitron tubes; or perhaps that was only about using it near the aperature grill mask (since that's extra fragile).

k to g1 is by far the most common leakage path as far as I recall - and I have to say it's been a while since I have been into CRT based repairs at this level, so what John said above about h-k shorts may well be true with 'modern' tubes, although I'm sure that I remember being able to remove h-k's in some circumstances with my B&K 465. As far as what you are measuring, there should be no reading between g1 and any cathode. If the grid connection is indeed grounded, then the statement about all guns being affected equally, is not valid in this case, and individual gun conditions could be affected by individual shorts. What you have to remember, is that it is irrelevant what *actual* potentials the individual electrodes are connected to, as long as the correct *differentials* are maintained. Hence, if the gun requires the grid to be say -40v to achieve cut-off, it doesn't matter whether the cathode is at +40v and the grid at ground, or the grid at

+20v and the cathode at +60v, the differential of -40v is maintained, and the bias conditions remain the same.

You questioned earlier what constituted a "firm rap". I really can't tell you in terms of anything that you could take as 'definitive instructions'. It's just one of those things that's a 'feel' that you gain over years of doing it. When I was directly involved with CRT equipment repairs, it was something you 'just did' on a daily basis, and you never actually broke a neck as a result ... Let's say that if you rap your fingertips on the bench just hard enough to get that sort of 'buzz' in them immediately after, that's probably about the most force that you want to be hitting the neck with. I would normally be using an old box spanner that I have, which has a wooden handle, and is a bit 'kinder' to the glass. As far as putting the monitor on its side or upside down when you do it, it's just a case of propping or holding it in position whilst you do it. It sometimes allows particles to dislodge under the effects of gravity coming at them from a different direction.

Other than this, use the capacitor discharge method to try and vapourise the particles. You might want to try this with the heater on, as you say that the resistance of the short goes down as the heater warms up. This is probably because the particle(s) causing the short get 'squeezed' tighter in the k-g1 gap, as the cathode cylinder expands. A better contact between the two electrodes, via the contamination, is likely to result in a higher discharge current from the cap, and a better chance of vapourising the material, rather than just dislodging it.

Arfa

The current required to clear a short needs to pass through both elements, the cathode and the heater. Even with the heater pins shorted, current will pass from the cathode to at least a portion of the heater, which is subject to fail at the point of contact.

That's true. If you have contact between the filament and more than one cathode, the tube is garbage.

Additionally, it would decrease image quality (at least when the short

Not true. In the vast majority of tubes I've isolated over the years, there had been no picture issues when the short occurs. The only way that you could tell the short was still there was by measuring the DC component of the filament above ground, and actually watch it vary from virtually nothing to a couple hundred volts.

If the filament is AC supplied from the flyback, it's at the scan rate, not 60hz. A true RMS meter is needed to accurately measure the voltage. I use a Fluke 87 for measuring high frequency AC. Non RMS DMMs usually read scan derived filament voltage below 5 volts. If the filaments are DC and regulated, they will read exactly correct with any DMM.

John

Thanks for your elaborate response.

I just had a revelation today, which also sheds light not only on this issue, but also with earlier issues I had with this monitor; that being that it turned brighter and brighter over time, resulting in a very washed out picture with visible retrace lines, when at default settings. The OSD menu allowed me to decrease cut-off and gain, but at some point that wasn't enough anymore, and I needed to decrease G2 value on the flyback. And now I think I know what's going on.

I think a leakage path between the cathodes and G1 has slowly been forming, which slowly increased beam current over the years, and therefore brightness. Now, I'm at a point where there is not just leakage, but a dead short now and again. Tapping therefore, will not solve the issue.

Another problem I had, was that the automatic color calibrator didn't correct for the washed out picture. And I suspect that is because the cathodes pulled the extra current directly from ground, and therefore the circuit which measures the beam current was oblivious to it. However, this is just speculation.

I think I will need a proper CRT rejuvenator to clear the leakage path between cathode and G1. I could do it myself, by using the G2 voltage for example, but googling reveals that modern CRTs are too fragile for this, and that professional rejuvenators are matched precisely for modern tubes.

It also appears that cathode-G1 shorts are one of the easier things to fix, so I think I can still avoid buying a TFT :). The only thing is, that it would be better to let a professional servicer do it, but I dread the likely outcome of it coming back with a calibration that is way off.

It's not really relevant anymore, because I think my problem is not H-K after all, but K-G1 (see my other post, about my "revelation" :)), but I have a scope, so I could test it. But a DC test on the supply yields 4.9V, and an AC test 0, so I think it's safe to say it's 5V DC.

One more thing. I may just decide to try to fix it myself (I was planning to use the capacitor discharge method, starting with a few uF at about 230V or so). But I have a question:

Most pin-outs of CRTs I've seen, contain only one G1 pin. But the G1 connection on this neck board connects to two pins on the CRT. And, with the socket removed, I can also measure 0 Ohms resistance between those pins. Is it normal that there are two G1 pins? When zapping, should I connect them together?

Probably for no other reason than there being a 'spare' pin. If they read zero to one another, then it's pretty safe to assume that they are both connected to the same place. Can you not find a data sheet for the tube on the 'net somewhere, or a schematic for the whole monitor ?

might be a good place to look for one. As far as your calibration fears go, I would suggest that when these shorts are removed, it is going to want a good set-up anyway ...

Arfa