HV tracking across flexible PCB strips
mtapley at swri.edu
Mon May 11 12:02:52 CDT 2015
On May 11, 2015, at 11:55 AM, Jules Richardson <jules.richardson99 at gmail.com> wrote:
> On 05/11/2015 11:22 AM, tony duell wrote:
>>> However, there's a flexible plastic strip running between the CRT face and
>>> the chassis which carries three lines (it's a storage tube rather than a
>>> conventional CRT) and I'm still getting periodic arcing across these lines
>>> which of course upsets the display's operation. As far as I can tell,
>>> there's no 'sandwich' (and hence glue) involved - it's just a single
>>> plastic strip with conductive traces drawn onto it.
>> What does this strip connect to? Is it part of the CRT (or connected to the CRT), does
>> it connect to a coil around the CRT, or what?
> The CRT end is soldered to three terminals which protrude from the front of the CRT - one is presumably the normal anode connection, one related to the storage mesh which I gather that such tubes have, and I'm not sure about the third (but possibly to do with erase functionality? I've not found an explanation of how that typically works)
>> I am wondering if there is something else leaking or flashing over which is causing
>> excessive voltage to appear here.
> Hmm, yes it could be that the HV is just too high and so it arcs over periodically (and by "periodically", it actually spends more time arcing than not) - although there's no obvious sign (change in brightness etc.) in terms of the displayed image immediately prior to it flashing over.
Not an expert at all, Tony’s advice is worth more than mine. That said, here are a couple of thoughts.
Surface contamination in a geometry like that is a pretty good suspect. Any hydrocarbons (coughnicotinecough) that deposited on the plastic between the traces could lower the effective resistance a long way; worse, once the first arc takes place, it’s not unlikely it burns a carbon filament across the contaminant which is even lower resistance. That might be a reasonably hard thing to clean off and could be narrow enough it’s hard to spot.
Geometry is a big deal in this case; wherever the sharpest point or bend radius is is a good place to look. However, I would definitely clean the whole thing pretty carefully (from the CRT terminals all the way to whatever the connections are at the other end of the strip) as part of the debug process.
One other possibility is relative humidity; dry air insulates a lot better, so if there is a way to turn down the thermostat for a while and let the air conditioner dry out the room (and hence unit) under test, that could help.
More information about the cctalk