Film-Tech Forum ARCHIVE: NC900 light engine rebuilds

This is topic NC900 light engine rebuilds in forum Digital Cinema Forum at Film-Tech Forum ARCHIVE.

To visit this topic, use this URL:
https://ft-forum.com/ft/cgi-bin/ubb/ultimatebb.cgi?ubb=get_topic;f=16;t=003551

Posted by Kyle Mitchell (Member # 8059) on 01-10-2019, 01:37 PM:

I have two with many blue pixels. What are the non-warranty options?

Posted by Mark Gulbrandsen (Member # 72) on 01-11-2019, 08:32 AM:

The only option on NC-900 is to send them back to the factory over seas. Its a 6 month wait... I suggest doing it through MIT out in Orange County.

Mark

Posted by Steve Guttag (Member # 268) on 01-11-2019, 09:13 AM:

I may be getting fuzzy on this but can't just the Blue Formatter be changed (chip and all)? On the 900, I'm pretty sure that the formatters are a part (each color) and there is no adjustment for them (as-in, they have to build them right, there are no convergence adjustments.)

Posted by Leo Enticknap (Member # 534) on 01-11-2019, 09:26 AM:

There are no Barco and Christie-style physical convergence adjusters, but on some NECs at least there are electronic ones. Never had to do it on a 900, so don't know if there are on that.

Kyle: if you're interested in returning yours for repair through us (MIT in Orange County), please give Jim Stewart a call on 714-751-7998, ext. 228, and he'd be happy to help.

Posted by Steve Guttag (Member # 268) on 01-11-2019, 12:21 PM:

All DLP series 2s have "electronic" convergence...that is a function of the ICP and it can merely move by 1-pixel increments (handy if you discover the problem going into a screening but don't have time to do a full convergence...which on an NEC is somewhat annoying (NC1200/2000 because they don't leave the screws/adjusters and often things move as you snug up the screws).

The NC900C-A blue formatter assy is 79TY1241. I can't say definitively that the formatter includes the DMD but I'd think it would have to since critical alignment has to happen there in order for the non-adjustable board (and it doesn't adjust) to work. They'd need a precise jig that guarantees all of them are identical. Since they likely cannot control TI's tolerances on the packages, how they mount the DMD would be where slop would be introduced.

Posted by Mark Gulbrandsen (Member # 72) on 01-11-2019, 03:47 PM:

On the larger projectors the DMD and formatter board are separate parts. I have a 900 engine headed my way soon to experiment cleaning...

Mark

Posted by Marco Giustini (Member # 4544) on 01-12-2019, 06:23 AM:

Have TI fixed that issue where using the electronic convergence adjustment causes an issue with some internal signal processing? (a CSC issue if not mistaken?). At the time TI was recommending NOT to use electronic convergence.

Posted by Leo Enticknap (Member # 534) on 01-13-2019, 09:23 PM:

From the release notes of ICP firmware version 4.5, under the heading "Known issues":

quote: TI Release Notes
There is a known issue using electronic convergence with DMDs that have a DDR interface, (the 1.2DC2K and 1.38DC4K DMDs). When electronic convergence is used to move one or more colors in the horizontal direction, the computation for CSC-P7 color correction is incorrect. This could manifest itself as incorrect color particularly noticeable at the edges of objects in the picture.

I don't know whether this issue is also present in earlier versions, or if it was actually (and presumably unintentionally) introduced into 4.5. It's been over two years since 4.5 was published, so I'm not sure if TI has any plans to release a new version of the ICP firmware that fixes it.

Posted by Steve Guttag (Member # 268) on 01-13-2019, 09:37 PM:

The problem predates ICP 4.5.

Oddly, Christie starts their electronic convergence at -3 for everything. NEC and Barco start at "0" for everything. Why Christie does this, I have no idea.

I've only used electronic convergence as a quick fix when there is insufficient time to do a mechanical convergence with the indication to the customer that a return call will be needed.

Posted by Kyle Mitchell (Member # 8059) on 01-14-2019, 12:40 AM:

The formatter board does not include the DMD and swapping out the board does not change the situation. I tried that months ago when the first one occurred, and now I've got one at another location.

Posted by Steve Guttag (Member # 268) on 01-14-2019, 01:51 AM:

What you need are the DMDs. If they don't include them, they must have them as parts. I figured that the formatter was the only way to get them since alignment would be critical. NEC is pretty good about providing torque specifications if you end up changing them yourself. Torquing those right is key.

Posted by Mark Gulbrandsen (Member # 72) on 01-21-2019, 10:50 AM:

Must also have the right tool kit to do proper torquing. I have one similar to his although mine came as a complete kit. Amazon is the cheapest source for this one. The SK will do all the hardware properly on an NEC light engine.

https://sktools.com/shop/skt0587-1-4-drive-2-36-in-lbs-adjustable-torque-screwdriver.html

Posted by Greg Routenburg (Member # 1742) on 01-21-2019, 02:58 PM:

From what I recall of my NEC training, the DMD on the 900/1000s and the larger machines is secured to the prism separately from the formatter board. When they do the factory alignment, it must be done as an assembled unit. The formatters can be changed in the field though. I've done it a handful of times. That won't solve a stuck pixel though unless it's a connectivity issue between the formatter and the DMD socket.

The other thing that could be happening is a degradation of the formatter cables unless the captan covers have been installed. That will cause single pixel errors to occur but they usually become greater in numbers rather quickly as the machine vibrations rattle off the damaged insulation from the cables.

I can't speak for how NEC does their factory alignment but I know that Christie "Flies" their DMDs into place on the prism using some sort of robotic arm and then glues them in place. As I understand it, the Formatter with DMD are attached to the flight machine and fed an alignment signal to the formatter. The whole unit is then precisely aligned to a lit prism. When I asked if I could see it on a factory tour, the fellow walking me through said that it was well above his pay grade to let me anywhere near that room and only a very select few have ever seen the process in action. It was still nifty of him to tell me that much.

Posted by Marco Giustini (Member # 4544) on 01-21-2019, 06:49 PM:

I don’t think Christie glue their chips in place. Sony does that and convergence is indeed not adjustable mechanically - but never really experienced a convergence issue with a Sony - or at least it’s not a common issue.

Older Christie’s are not easy to adjust - as series ‘1.5’ Barco’s - later models are spring loaded as Barco does - but without easily accessible knobs.

Posted by Carsten Kurz (Member # 5396) on 01-22-2019, 06:17 AM:

That would have been my question - how would you do mechanical convergence if you do a one time pick and place and glue process? Also includes the question how difficult it is to replace a single DMD with stuck pixels practically?

- Carsten

Posted by Greg Routenburg (Member # 1742) on 01-22-2019, 12:42 PM:

As I understand it, the Christie DMDs and their formatters are mounted to a frame that is then glued to the prism. The frame has adjustment screws built in to allow for convergence adjustments to be made. The only expedient way to have a Christie LE fixed if you have a DMD or Formatter failure is to purchase a replacement LE and submit the defective one for a core credit. They take the defective units in, clean & rebuild them and then sell them as refurbished. You may still be able to send your LE in for repair but you're dark until it comes back. You also need to wait for a manufacturing slot to come available before it gets looked at. It's not usually that long but you're really at the mercy of their order load at the time you send it in for repair.

Posted by Steve Guttag (Member # 268) on 01-22-2019, 04:02 PM:

Greg,

I'm 99% certain on the NEC, the DMD is NOT glued to the prism. I've removed a formatter on an NC1200 in recent memory (within the last 4 months) due to contamination being blown in by the cooling fan) and the DMD was on the formatter board. There was a neoprene (could be another material) gasket that would seal it up to the prism but it definitely wasn't glued there.

And, I seem to recall from my NC900 training that the DMD was non-adjustable and was also connected to the formatter board. Which makes sense, the formatter is what drives it.

Posted by Greg Routenburg (Member # 1742) on 01-27-2019, 11:39 AM:

Steve,

The last formatter that I ordered for an NEC projector came without a DMD. There were gold plated pads on the PCB where spring loaded pins on the back of the DMD socket seated against the formatter. I can't speak to how the DMD itself was held to the prism or aligned in the first place. It could very well be a pressure fit as you describe. When I spoke of gluing, it was the Christie LE.

NEC doesn't have the DMDs listed in the parts list and when I've had the "Blue DMD Starfield" failure on a 900C, I was advised that a factory rebuilt prism would be require because the DMDs are not field serviceable.

I hope that clarifies things a little bit.

Posted by Steve Guttag (Member # 268) on 01-27-2019, 06:12 PM:

Yeah, somewhat. So when you changed the formatter...was the DMD still on the prism? Or did you transfer the DMD from the old formatter to the new?

And if you transferred the DMD...that would debunk changing the DMDs as a non-field item.

Posted by Greg Routenburg (Member # 1742) on 01-27-2019, 11:28 PM:

So when I changed the formatter, the DMD stayed on the prism. I honestly made no attempt to try and remove it. Don’t have the budget to be messing with them too much if ya know what I mean. I just swapped the boards, torqued the screws that secure the formatter to the NEC specified values and it worked.

Posted by Mark Gulbrandsen (Member # 72) on 01-28-2019, 08:11 AM:

NEC has a convergence mount the DMD's are mounted in and there is also a seal between DMD face and prism to keep dust out. Christie has the dust blower system built into their DMD mounts.

Mark

Posted by Steve Guttag (Member # 268) on 01-29-2019, 07:03 AM:

That would actually make sense then, given no adjustments. So yeah, NEC has to get the DMD alignment exact, at the factory, on the prism. So yeah, in the event of a bad DMD...a new prism would be required.