ICP and DeOxit story

[Steve Guttag]

I, recently, ran into an issue where a different problem lead to a new problem. I had a Barco DP2K-20C that was throwing Enigma communication errors, seemingly at random. Reseating the SDI, and even the Enigma itself wouldn't stop it (including cleaning the contacts with IPA). These were all done by the client.

The projector was old enough that every Barco tech's favorite replacement was in order...signal backplane. Normally, the signal backplane problem show up in the ICP connections with red or green (rarely blue) noise or haze on the screen. But I've even had them cause a cover sensor (microswitch) to not get its message to the CCB.

Fine, so I change that. As per my standard practice, if I'm going to pull the boards, particularly on a signal backplane change, I apply a fresh coat of DeOxit (D5). I believe in the light-touch on DeOxit in that I spray it onto my Kimwipe or Lens Tissue (whatever I have that day) and then apply it by wiping that onto the gold fingers, in the direction of the gold fingers...let it cure (haze over) and reinstall. Upon powering up, the Enigma communication issue was gone...but...I picked a new and different problem...ICP version mismatch, error connecting to the red satellite; signature fail.

This cannot be. We didn't have that error before and we have a new connector on the signal backplane and the ICP was JUST cleaned. Perhaps a "reseat" nope. Same error. Okay...perhaps I put dirt or something on the connector, somehow when I cleaned it the first time. As the shampoo commercials say...rinse and repeat. Clean the edge connector again with DeOxit. This time reseat it 3 times...and...nope. Same error.

Well, what's changed? The Signal Backplane. Could it be that I got a bum unit? Hmmm...I reseated that board 5-times and it finally came up clean. Why? Reseated it again (I don't want it to be flaky) and the problem returned. Reseated it about another 5 times, problem gone. Maybe it is a bad signal backplane and it is now flexing just so, so now there is a good (enough) connection? I don't have another signal backplane with me so I'll need to return.

Days later (and it has been running) I return with another signal backplane. Before I start the procedure and even before loosening the screws for the ICP, I grab its two knobs and try to wiggle it against the force of the two little screws...bam...the error pops up. I'm thinking...yeah, it has to be the signal backplane...we never had this issue before and the slightest flex causes it to flake out.

Fine...I swap the signal backplane again to another new one...and...the exact same error comes up! Now, mind you, I've cleaned the ICP my normal way, with DeOxit twice with seemingly zero difference. But it is clear, this isn't a signal backplane issue. It could be a cold solder joint on the ICP that is responding to the flexing and how the board is fastened on each insertion. That would suck as finding that needle in a haystack would be very difficult since that sort of failure could also be a cracked trace, including on an internal layer.

So...I decided to get medieval on my contact cleaning (and they did look normal to me with none of them having an overt oxidation/dirt appearance). No more Mr. Nice Guy. I went cross-grained on the fingers with IPA, then a generous helping of D5. Wiped that off...then a generous helping of G5 (supposed to be specially formulated for gold contact surfaces) and again...aggressive cleaning, with pressure, cross grained. Then I pulled out G100 (all juice) and painted that on. Then gently wiped in the direction of the fingers for a nice even/thin coat and let cure a few mintues.

The result...solid. The problem was gone. I powered down, reseated...problem gone...reseated again...problem gone. It's solid now.

The moral of the story? I don't know but, apparently, kid-glove use of DeOxit is not always the right procedure...though it has worked for me in all previous instances. Even worse, had I merely swapped the ICP, that would have fixed it too but to the tune of many thousands of dollars for what amounted to a contaminated gold finger that didn't respond to the normal methods. It's a first for me.

[Armand Daiguillon]

Based on this - would you not recommend de-oxit in the future ? Is there something else you'd think is better going forward ?

[Steve Guttag]

No. DeOxit's reputation is well earned. Typically, if I clean a board, I never hear of problems from it again. This was the odd outlier that is the exception to the rule. I think, clearly, there was something on there that didn't jump out at me by mere visual inspection. I would suspect that if looked at under a microscope, the problem would be been more apparent. Normally, one isn't "aggressive" when handling such a board or gold fingers. The ICP is a pain due to low-signal levels (can't overcome much of any form of contact resistance) and its density (look at the number of simultaneous connections that thing has to make via an edge connector. Their (TI's) choice of connection was unfortunate. Had they used connectors like that are used on the fan controller, I suspect their problems would have been near zero but to get the density of connections, the board would have take up a bit more space.

[Leo Enticknap]

I wonder if the underlying cause of the original issue was actually a bad backplane, and that a reaction took place between its contacts and the contacts on the ICP that left a bigger problem on the ICP than a gentle wipe with a DeOxit-soaked Kimwipe was able to remove. My technique during maintenance C or a backplane replacement is to clean the contacts with an alcohol prep pad first, squeezing gently between thumb and forefinger, leave them for 2-3 minutes for the alcohol to evaporate, apply DeOxit Gold from a bottle/brush, seat the board in its rails but don't push it home, then push it home a few minutes later, after having let the DeOxit cure. But if there was serious contamination on the board contacts, that likely wouldn't remove it. In Steve's case, maybe more aggressive pressure on the contacts is what did.

[Dave Macaulay]

I will loosen the screws holding the ICP board to the the front plate, pry the board away from the panel, and tighten the screws. This gets a half millimeter maybe more insertion into the backplane, has helped with intermittent errors.

[Steve Guttag]

Dave...that was step-1 YEARS ago...but it is always worth reminding for the odd ball ones that haven't had that done. The advice applies to all three OEMs too.

[Marco Giustini]

I don't think that your gentle approach with D5 was the problem here, I would have done the same.
As you say, if you looked with a microscope you'd probably have seen the problem. I think it was a Derbauer video on YT where they were analysing GPU power connectors under a very powerful microscope and you could see when a gold-plated connector was plugged onto a non-gold plater housing, there was a butter-like silver compound deposited on the golden fingers, coming from non gold-plated housing. Chances are the new backplane had some contamination which spread on the gold finger and the "gentle action" was not enough to remove it.

Is the solution to be more wild on those connectors? Not sure.

I'm still puzzled that those projectors commonly require a new backplane while my 10 years old PC does not.

Coincidentally, I've got my first Deoxit bottles delivered today: D5 and G5. The expectation bar is high!

[Steve Guttag]

How often did you need to "reseat" cards in a Dolby CP100, 50, 200, 55, 65?

Why does Barco need 100:1 more Signal Backplanes than the other two? Mind you, all three OEMs will get ICP issues, sooner or later, but Barco gets them at least 10:1 faster/more often...which lead to Signal Backplane replacements.

[Leo Enticknap]

I'm still puzzled that those projectors commonly require a new backplane while my 10 years old PC does not.

Your PC does not contain up to a 6.5kW lamp spewing heat and UV into its case.

I'd have to look at my records to be positive, but I'm pretty sure that all my Series 2 Barco backplane replacements have been on lamp machines, and none on the laser phosphor variants (DPxK-XXBLP and CLP). Even then, the ratio of lamp to laser Series 2 projectors I service regularly (though Barco calls BLPs and CLPs Series 3) is around 7 or 8 to 1, so that observation is likely not statistically significant. But I do wonder if that design is not as good as that of its Christie and NEC equivalents in managing the temperature inside the card cage while the lamp is on. One related point is that in the Barco, the (heat-producing) SMPS is immediately underneath the card cage, whereas on most Series 2 NECs and Christies, the low voltage power supply modules are further away from it.

[Marco Giustini]

How often did you need to "reseat" cards in a Dolby CP100, 50, 200, 55, 65?

That was before my time in cinemas, do you mean it was often needed or not? I remember having to re-seat a cat 249 in a CP65 multiple times to get rid of some noise but eventually it was a broken joint on the card. Once that was fixed, there was no need to "re-seat" that card anymore

Yes, I appreciate that the many high-speed fans might play a role here. Heat, I'm not sure about as the card cage gets fresh air from the outside. The SMPS is at the bottom indeed but you have air floating across it, I doubt the heat would be able to spread across the card cage.

If vibrations were a problem though, then why don't manufacturers implement anti-vibrations counter-measures? It would be simple to mount the cage on some rubber or similar dampeners.

[Marcel Birgelen]

"Originally posted by Marcel Birgelen"

Uhuh... It's really interesting how I get quoted without even being here! It's like magic.

Meanwhile, the topic sounds a bit like one from those "Frank's story thread". I've spent some trees, 5 swimming pools of water and 2.5 dolphins on the following AI story, so you better like it:

The Silent Guardian and the Elixir of Life - A DeOxit story

In a dimly lit workshop, tucked away behind shelves of oscilloscopes and soldering irons, stood an old Texas Instruments ICP. Its once-bright LEDs had faded, its buttons stiff with age, and its connectors tarnished by time. It had been decades since it last hummed to life. Now, it sat silent, a relic of a bygone era of engineering brilliance.

The workshop’s owner, an aging electronics enthusiast named Elias, had rescued it from a forgotten corner of a flea market. He remembered the ICP’s glory days— but now, it was just another dusty artifact, its potential locked away by corrosion and neglect.

One evening, as Elias wiped the dust from its casing, he noticed something peculiar: the ICP’s connectors gleamed faintly under the workshop light, as if whispering for help. He sighed. "You deserve better," he murmured, reaching for a small bottle of DeOxit—the legendary contact cleaner, known in the electronics world as the "elixir of life."

With the precision of a surgeon, Elias sprayed a tiny amount of DeOxit onto a cotton swab and gently cleaned each connector. The tarnish dissolved like morning dew under the sun, revealing the shiny metal beneath. He repeated the process for the buttons, the switches, and even the edge connectors. The ICP, once dull and lifeless, now sparkled as if it had been reborn.

Elias plugged it into a power source, his heart pounding with anticipation. He pressed the power button. For a moment, nothing happened. Then—a flicker. The LEDs blinked weakly, like a tired traveler waking from a long slumber. He pressed the button again, and this time, the ICP roared to life, its display glowing with the familiar prompt:

READY.

Tears welled in Elias’s eyes. The ICP wasn’t just a tool anymore; it was a guardian of memories, a bridge between the past and the future.

That night, the workshop hummed with energy. The ICP, once silent, now sang its song of ones and zeros, a symphony of revival. And the DeOxit? It had done more than clean—it had unlocked a story, a testament to the enduring power of passion and precision.

Epilogue
Elias never looked at his tools the same way again. To him, they were no longer just objects—they were partners in creation, each with a story to tell. And the ICP? It became a legend in his workshop, a reminder that even the most forgotten things can be brought back to life with a little care, a little patience, and a bottle of DeOxit.

The end.

[Steve Guttag]

"Originally posted by Marcel Birgelen"

Uhuh... It's really interesting how I get quoted without even being here! It's like magic.

I'll take the "Merican" approach..."y'all look alike." In Leo's defense (who didn't start in these parts) your profile pictures do like identical!

I suspect that real explanation is that there are those that just use the Quote button in the reply so nobody is referenced and those that use the Quote button in the posting, which often means editing/deleting the quote down and then there are those that know enough about the formatting to add in the quote reference and Leo probably just snagged the wrong person but I strongly suspect it was a manual endeavor.

[Leo Enticknap]

That's exactly it - sorry.

[Marcel Birgelen]

Well... Excuses accepted... grudgingly. . Barely integrated into American Culture and already playing the Yank card... the country that made you had manners, you know? :P

As why I look so much like Marco... Maybe you don't want to know how I look in real life and I'm all just doing you a favor! Or maybe I'm just too lazy to upload a proper profile picture? I'm just new here, you know... Don't know what Marco's excuse is, but he may use mine, for free... at least this time.

As for the ICP and PC analogy... consider the ICP to be a GeForce RTX 5090 and the backplane your humble PCI-e interface. If you have ever had the distinctive pleasure of trying to make them meet in blissful electric unity, you know what I'm saying...

[Marco Giustini]

I don't know what you're talking about, Marcel, mine is an actual picture!