View Full Version : Nec XG: Blanking Pedestal Tips


Bjoern Roy
07-10-07, 08:57 AM
Hi folks,

in Kenny's recent thread (http://www.avsforum.com/avs-vb/showthread.php?t=870425) about black level problems on his XG, my first thought was 'blanking pedestal level', because this is the ugly step child of the XG series. I have had to correct this problem on dozens of XGs, so this isn't a isolated problem. Actually, i haven't seen a single XG which had this set properly no matter how many hours on the chassis.

Independent of the fact whether this is Kenny’s problem, I think blanking pedestal levels is something not discussed in-depth here yet, so let me give you guys some pointers.


Indicators of the issue:

1.) Common case
You have some funny vertical stripes in your picture which only show up in dark images and a black image. These stripes seem to move when you adjust left and/or right blanking!

Delimitation to other issues:
- Stripes also show up in white field? On the left side mostly? -> Might be: Swing-in of signal. Cause: Projector, Source or a combination of both
- Stripes do not move when adjusting left and right blanking. Heck they even stay in the same place when change width of image through image size! -> Burn in :(

2.) Extreme case
You cannot get your black level low enough. There is always a considerable glow on the tubes after (the hideously painful ;) )G2 calibration. If you watch closely, you might even see some sort of inward bound edges at the top and bottom of the screen. That’s your retrace! Why the heck is this visible here?

Note: The problem having this discussion here is, how much glow is acceptable? What do I mean with ‘considerable’? After proper G2 and black level calibration, there WILL be a glow. Otherwise you crush your blacks. How much is normal? How much is ‘just barely visible’ etc. Don’t have an answer here :(

Solution: Calibrate the blanking pedestal with a scope

The problem with the blanking pedestal is, in my opinion, that the service manual isn't 'real world' enough in its recommendation how to set it. If i remember correctly, it says something like 'Set the blanking pedestal so that it matches the video black +/- 0.xxV'
A) For any sane videophile, this should be '-' only. You do not want to have blanking brighter than your video black, because the blanking level is also the voltage that drives the raster retrace and thus pollutes the image. (See 2 above!)

B) The given range is WAY to small from my experience. I start with at least 10 times the tolerance in mind. The blanking signal just has a LOT of contamination in it, especially in the close proximity of the actual blanking border. Have seen some wild peaks there on some units.

So what i do is set the video level relative to the blanking level so that the video level is just barely above all contamination that i see in the 'visible blanking area'.

Huh? What i mean by that, is that i don't care for peaks in the blanking which are going to be left and right of the picture once you actually USE the blanking. The critical peaks are those which are visible in the retrace!

Definition: ‘visible blanking area’. The retrace part of the blanking signal which is visible behind the actual video signal.

For this to work, i do this adjustment with the target resolution (e.g. 1080p60) and target porches, because these influence which part of the blanking actually the ‘visible blanking area’ is.

Note: How do I correlate the scope with what I see on screen? By increasing brightness and moving left and right blanking around, i get a feeling for which peaks on the scope correlate to which bands on screen. Why do we need to correlate them? Well, the scope is unavoidable for the actual adjustment. But we need the screen for the determination which peaks are in the ‘visible blanking area’

Never mind setting these to the ‘same’ level. Just hide all the ciritical contamination individually on all 3 tubes. These ‘tend’ do be similar, though. Don’t sweat it if they aren’t. Also don’t sweat if your blanking level is WAY above the tolerance mentioned in the service manual. Its written a) without real world signal contamination in mind or b) for use with signals which use NTSC setup pedestal ;)

Since you are changing the balance between the blanking and the video black level with these adjustments, you will need to make sure that you get your desired 700mV peak-to-peak again afterwards. But this is another discussion.

If this is way over your head, consider having someone calibrate your XG. Its one hell of a machine, but a real bitch to master.

Best regards
Bjoern

P.S:
http://www.tfw2005.com/boards/images/smilies/this_thread_is_useless_without_pics.gif

Yeah, i know. Some drawings, or pictures of the tube or scope would be awesome. Maybe in a future XG install. Am not doing nearly as much calibrations these days due to time constraints, though.

Mark_A_W
07-10-07, 06:33 PM
Great post Bjoern - should be a sticky!

(Might see if we can get it added to the advanced procedures on the "other forum"?)

If I fiddle with mine I'll takes some pics, but it's not likely either, I have no time.

Mark

benny
07-12-07, 03:03 AM
Its one hell of a machine, but a real bitch to master.


Another thanks for this tip from downunder :)

I've noticed those anomalies outide the active image area ( and within to some extent ) and put it down to the way NEC's video chain works. You've given us all an avenue for further improvement now so it's time to drag the 'ol CRO out again and have a rethink about those base level video setup's.

I'm still mostly pleased with the image my XG-1350 is producing after all the countless tweaks I've done but I reckon this one might be just about the last needed. It's such a pain in the a$$ though to have to re-do everything from scratch in the video chain! I know ... I'll practice on Mark_A_W's first to get the technique right :D

Boil the billy Mark .... time to play again.

Cheers again Bjoern :)

Russ

Mark_A_W
07-12-07, 03:30 AM
Bewdy Mate:)

Just wait till I've bought a new video card - HD2600XT have been released, but haven't hit our shores yet.

I want to sort out a new 1080i96hz/100hz res that doesn't cause raster ringing on the LH edge.


When I got my XG I thought it was unusable due to the red stripe about 200mm in from the RH edge - that turned out to be the blanking/pedestal level. They are much more finicky with the rasters than PG Xtras. More capable, but more picky.

If you haven't got any of the issues listed above, then yours will be set correctly. Someone tweaked all sorts of things on my XG chassis - including disabling the Oscillator board CPU, which is why it was "broken" when I bought it.

Hmm...better get me a bigger kettle...

pasey25
07-13-07, 01:56 AM
Well, since mine is sitting waiting to be setup again from scratch, maybe i can temp you up north Russ ;)

All my travel has quieted down again. time to fire up the beast again, revel in some LC goodness!

kschmit2
07-13-07, 02:49 AM
Bewdy Mate:)

Just wait till I've bought a new video card - HD2600XT have been released, but haven't hit our shores yet.

I want to sort out a new 1080i96hz/100hz res that doesn't cause raster ringing on the LH edge.





I have ordered that card as well. Hope to get it early next week.

mp20748
07-13-07, 03:15 AM
Very interested thread indeed. A less known or thought about demon being discussed here, and for exactly the right reasons.

Watching and maintaining pedestals, is a very important thing when doing mods. Especially with HDTV sources and signals. The tri-level sync, or HDTV smpte sync system does strange things to the pedestals in CRT projectors. Not sure why, but I've read articles on the difficulty of controlling the black level with modern the modern HDTV system. That also explains the many black crush problems that used to be quite common on earlier transcoding. Digital component is perfect for transmission and storage, but hell when converting back to RGB.

Proper pedestals are not a problem on the NEC's only. It's a problem on all the CRT's. I would say it's good that NEC provided a means to adjust for it. That was very thoughtful of them.

carpfisher
07-13-07, 04:47 AM
The tri-level sync, or HDTV smpte sync system does strange things to the pedestals in CRT projectors.

Interesting, as it was designed to be transparent to the average dc level of the signal, so as it has equal amounts of +/-, it evens out.....

Mark_A_W
07-13-07, 04:50 AM
I have ordered that card as well. Hope to get it early next week.

Let me know what res you end up using Kai. Does it have Powerstrip support yet? The 2900 does.

mp20748
07-13-07, 07:36 AM
Interesting, as it was designed to be transparent to the average dc level of the signal, so as it has equal amounts of +/-, it evens out.....

I was under very similar theory before I designed and built my own transcoder (MP-5).

After much R&D, it produced exceptional low end performance. In fact, it's low end performance was far better than everything we compared it to. And that included both the DVI units and the commercial transcoders.

It had 3 pots inside for precise adjustment of the pedestals of Y, Pr and Pb. I had to put the pots there because the pedestals would not follow the rule. And even after precise calibration, some HDTV signals would still push the pedestals away from where they were calibrated.

Others have also added pedestal pots, and because the timings played out so differently, some even had H shift pots, that were also called "brightness" controls.