Fixing a Toshiba scan converter

Has anyone done one ? I forget, but the ones in the fdx or hdx series. I suspect they're all the same.

They were on backorder for quite awhile and we finally got it and the set is fixed and gone, and the money is probably spent, but I was looking at the old one.

There are several SMD caps on the module that check bad and some are not even recognized by the cap checker. Some of these are near two identical ICs that each have two crystals nearby. I reckon these are digital clock"ing" chips and with the symptom I think it may be repairable.

In this particular specimen there appears to be no capacitor piss, or at least the resultant corrosion.

The symptom is that there is constant piecrusting in the main image only, the PIP is good and the OSD is good. It is not a sweep problem. It is because the set does not run 480i even on an NTSC input, and the PIP uses a seperate scan converter for obvious reasons.

The problem used to be only when the set was cold.

Two plus two have at least come up to three here. Thus my plan; a friend has a Toshiba with exactly the same problem, so I made a deal with the shop, I will replace all the bad caps on company time, the only way to see if this worked is for me to go install it in my friend's set. These things are $280 list, so we agreed, if it works I'll bring $100.

Now a question, who owns my buddy's old board which will be abandoned to me ?

I don't have the tools at home anymore to fix it, the boss sold me a board which was a pull, they pulled it, cust didn't want it, it is theirs until I give them money.

I will append this later with more info, and later when I know if the fix worked. I'd like your input, especially if you've fixed one, but I also post this to share the info.

For those of you who want raw data : I found at least 6 marginal capacitors, 3 bad and 2 that weren't even recognized as caps by the checker. The symptom used to go away when warmed up.

2+2= ?

JURB

Reply to
ZZactly
Loading thread data ...

I've done at least ten of these boards for various video\piecrusting issues. I believe the 10uf caps were the culprits, but in any case there are two at the top rail that need to be replaced. Not one came back.

John

Reply to
John-Del

Thanks, that's encouraging.

So now I took the ESR checker to it again and am now replacing 29 of them. Most of the 10s are bad and I think all of the 22s. I wonder how the thing ran so long, I mean at this point some of those caps could be so non critical application wise they could probably be removed, but who knows which ones.

I have some theories about how to service stuff like this in the future as the bottom and top lines put the financial squeeze on us.

To start, think about this; just what makes them use a 22 instead of a

10 ?

  1. If HF needs to be bypassed they won't use those shitty performance caps unless they have a smaller cap as an HF shunt.

  2. The capacity tolerance makes it so they almost never use SMD caps for any timing.

The only exception to #2 might be that they are trying to manage startup and shutdown. I mean they want certain sources to come up or go down faster than others. I have indeed seen Mits PIP modules that seemed identical except for SMD cap values. Of course that doesn't mean they were in an identical chassis, prehaps it has something to do with the set having a different power supply. Choosing the wrong values could possibly hurt reliability by causing unwanted surge currents during starup or shutdown.I will tread carefully when messing with other people's designs, but I will indeed tread.

Not in rebuttal to #1, but there is another factor. Even with HF help from a ceramic or other decent cap the need to filter the current fluctuations on the actual wires from the power supply, and foil traces. The lower frequency components of this hash do not need to be supressed completely, so they can get away with the cheapo caps.

Might take a little experimentation to figure out the best way to do these.

JURB

Reply to
ZZactly

Just replace all of the 10uf ones. That is the standard rebuild on these

========================== Jeff Stielau Shoreline Electronics Repair

344 East Main Street Clinton,CT 06413 860-399-1861 860-664-3535 (fax) snipped-for-privacy@snet.net ========================
Reply to
Shoreline Electronics

ElectronDepot website is not affiliated with any of the manufacturers or service providers discussed here. All logos and trade names are the property of their respective owners.