Barco DP2K-10S – Blue Blotch Bottom Right (Prism Contamination or Blue DMD?)

[Marcel Birgelen]

Torque on some components might be a critical factor, especially since this stuff goes through quite some heat cycles. Too much torque might not outright damage stuff immediately, but the next time stuff gets hot during production, insufficient torque might result in sloppy component alignment, conversion issues, etc.

I would be less concerned about the DMD contacts. I've plugged in processors into mainboards countless of times and the amount of contacts there is similar. I've never really had any big trouble with contacts there, as long as you keep stuff reasonably clean.

I still don't think that a light engine really is a field-serviceable item, but with the right equipment and proper knowledge, I see no problem in doing this yourself. The risk level on those older models also has gone down, especially if you're already in an all-or-nothing situation, where the price of a new or revised light engine and service hours might be more than the remaining value of the entire machine.

It would be nice if there were more practical means of getting light engines serviced. Common issues (that are growing):

• stuck pixels (either on or off).

I guess that fixing stuck pixels on a DMD without outright replacing the DMD is a bit of a lucky shot. Even if you'd be to remove the glass protection in front of the mirror substrate, I'm not sure how you could effectively do anything with tools available to the normal human, other than to give it a "good shake" in the hope it would somehow have fixed the issue. And as soon as as you enter this territory, every spec of dust will be a lethal enemy.

I suspect that when you have a big blotch on the screen, some probably invisible specks of dust might be the lessen of the two evils. The optical path will focus on the DMD, the protection glass is further away so only large contamination will be seen.

A tiny spec of dust will probably not visibly register on the image, but you don't need a lot of contamination in order for stuff to show up at these kind of magnification levels. Also, at those kind of light levels, you want to keep the "foreign material" really to a minimum, as stuff could burn in or gas out and do other nasty stuff.

[Marco Giustini]

I've plugged in processors into mainboards countless of times and the amount of contacts there is similar. I've never really had any big trouble with contacts there, as long as you keep stuff reasonably clean.

I've never had to re-seat my CPU or GPU or RAM - yet, I've had to re-seat an ICP countless times. Or replace an entire backplane! ?
Fair enough, my PC doesn't have 15 shaky fans inside.

[Marcel Birgelen]

I blame that one on the DCI backplane design. I've seen countless of other modular backplane designs with interface cards that work great, but I've also seen some that were inherently broken. I remember a particular type of modular switch by Extreme, in which wiggling around interface cards was the norm...

I'm not sure who provided the backplane design in use for DCI equipment, but I guess they based it some kind of industry standard. Apparently, it allows for too much tolerance. If you want a PC comparison: It's a bit like those PCI(e) cards, especially if you're trying to fit a giant, oversized, 3-slot Nvidia GPU in a humble PCIe slot. If that card doesn't work, 9 times out of 10 it's also because of a bad PCIe connection...

[Steve Guttag]

With the ICP, you have a multitude of factors working against it. There is an inordinate number of gold fingers on the board (it's got two sections and they are double-sided). You then have extremely low current signals that will be negatively influenced by any contamination to either the board or its mating socket. You then place this precarious set up in a projector that has no only a lack of a controlled environment but also pump whatever environment it has on it. You then, particularly in some parts of the world, have it cycle from high humidity to low humidity. Have you ever noticed that the need to reseat an ICP goes way up when the humidity swings? Around my part of the world, Mid-to-late November the humidity drops and ICP reseats goes up.

Adding into this is that everyone is using different PCB connectors to mate with the ICP. Here I think we get to more of the crux of the problem. The ICP was made by TI for ALL THREE projector companies. The only difference is the faceplate. The only software difference are the unique secure passwords used by the projector itself (not your passwords but how their projector's PIB/CCB/FMT board talks to it). Yet, Barco has what, a 100:1 differential on board reseats with similar differentials on signal backplane replacements? That is more than coincidence. The connectors must be inferior. Combine that with the projector's inability to self-heal at all. I'll see an NEC, self-reboot, on start up, if the ICP doesn't talk with the satellite boards (Formatters) but Barco won't do that. Then again, NEC can't deal with even the slightest power hiccup and takes a day and a half to do simple operations like move the lens.

Life has gotten a little better in the S4 world but still I've reseated ICP-D (or ICMP-X) and I've changed at least two signal backplanes. So, perfect, they are not.

I've had to reseat ICPs on all three OEM projectors but it is a significant difference between Barco and the other two.

And least you think DeOxit is the cure for it all. It isn't. It helps and can prolong things but it isn't any magic elixir. Recently, I was doing a regular PM call in prep for the season, plus their ICMP was showing a low-battery (measured 2.850, FWIW). I did a full C-Maintenance, including applying a fresh application (sparingly, of course) of DeOxit to the CCB and ICMP...let the cure/haze. Everything went together fine. Everything tested fine. Within a week, a reseat was needed to cure a flashing blue on screen (something they didn't have before). But hey, the low-battery warning was gone!

My personal opinion is there is way too much need to "reseat" things and to "reboot" things. It is a sad state that those procedures are considered standard "repair" when what they are is a cover up of poor design. This stuff should self-heal on these sorts of things and electrical connections should be designed to not need such attention. And don't think I'm singling out Barco (though their edge connectors suck)...I've been burned by all three, just differently.

[Marcel Birgelen]

With the ICP, you have a multitude of factors working against it. There is an inordinate number of gold fingers on the board (it's got two sections and they are double-sided). You then have extremely low current signals that will be negatively influenced by any contamination to either the board or its mating socket. You then place this precarious set up in a projector that has no only a lack of a controlled environment but also pump whatever environment it has on it. You then, particularly in some parts of the world, have it cycle from high humidity to low humidity. Have you ever noticed that the need to reseat an ICP goes way up when the humidity swings? Around my part of the world, Mid-to-late November the humidity drops and ICP reseats goes up.

It's essentially the same design as PCI(e) interface cards in your PC. They're also the most problematic, have a tendency to become loose during transport and only need to wiggle them out a little bit to become problematic.

When I look at other high-bandwidth backplane designs, e.g. that of big core switches or routers, which also operate at high bandwidth rates with relatively low voltages, I never see this kind of on-PCB connectors, but always a bunch of separate, high-density connectors, often with extra metal guidance pins on the sides of the connector.

The TI design was clearly a way to cheapen out on the costs for the interface cards' PCBs, namely the ICP, because it omits the need for extra connectors to be mounted on the interface cards themselves. Also, the connectors on the backplane for such "on PCB connectors" are cheap and abundant. Given the price tag of those cards, there really was no need to cheapen out on this. Maybe they should've looked at the Telco industry, they've been building "NEBS compliant" equipment since the 70s, where temperature swings, salty air, vibrations and humidity are all accounted for.

[Carsten Kurz]

Keep in mind, these series 2 cards have much longer edge connectors than the typical PCI(e) or similar cards. The wider these cards are, the more issues you get under temperature cycles.

[Sihong Cheng]

Hello everyone, I am a cinema technician from Indonesia, I work in an independent cinema, so I have minimal fellow technicians.
I am currently operating a Barco DP2K-10S . Recently, I noticed a permanent blue blotch appearing at the bottom right corner of the projected image.

The blotch is always blue, regardless of the color being projected.

It is most visible on a full white test pattern, but it can also be seen (less strongly) on all colors.

The position is fixed and does not move.

The appearance is more like a blurred spot/cloud, not sharp pixels or lines.


What I have already tried:

Cleaned the porthole glass.

Removed and reseated the lens.

Swapped lamp house with another unit.


In all cases, the blotch remains the same.

My questions:

1. Do you think this is prism contamination rather than a failing blue DMD?


2. Has anyone here successfully cleaned or refurbished the prism block or dmd chip in a DP2K-10S light engine?


3. Would replacing the blue DMD chip solve this, or is it more likely that the entire optical engine needs servicing/replacement?

20250920_232452.jpghttps://drive.google.com/file/d/1FA1ka9Zn6daO6hj9jH8tyvpH-VYRveBY/view?usp=drivesdk

I attach sample pictures of the screen output for reference (white test field, with the blue blotch visible).

Thank you in advance for your advice and experience.

Best regards,

This is not a problem with DMD. It is caused by dust on the light-collecting rods within the optical path system. Perhaps there was dust on the light-collecting rods during your cleaning process of the optical path.