Sign in with
Sign up | Sign in
Your question

CRT slightly dimmer on one side

Last response: in Computer Peripherals
March 24, 2003 4:32:53 AM

I was just over at a friends house... she has a Sony G400 19" CRT monitor.

Her monitor looks perfect... great convergence, great color, great focus... everything real nice... except that the right half of the screen seems to be slightly dimmer than the left half. When I say "slightly", I mean it's barely noticeable, but you can tell if you really try to see it. If she didn't mention it to me, I wouldn't have noticed (it's subtle)... but after she mentioned it, I could see what she is talking about.

It seems to start at about 50-65% the way across the monitor... meaning that it affect the RIGHT 35-50% of the display. It seems to be consistent from the top to the bottom.

We tried all the adjustments, and nothing made a difference. Tried degaussing it a few times, and selected "Image Restoration" from the Options OSD menu (whatever that does, I don't know).

The monitor is fairly bright overall, so she runs it a brightness of 10 (out of 100) and a contrast of 85 (out of 100). 91.3kHz/85Hz at 1280x1024 are the frequencies and resolution.

The monitor is running on a true sine-wave UPS.

There are no obvious sources of magnetic interference nearby. The only things even remotely close are a tall floor lamp (with a dimmer control) behind the monitor, a desk lamp to the left, and an electrical outlet two feet to the right (with only the floor lamp plugged in). There is a large old stereo speaker about four feet to the right. We turned off the lamps, turned off the switch to the electrical outlet (which I THINK controls both recepticles, but I'm not sure), and turned the stereo off. The only other thing nearby is the computer, and it's about two feet away, and to the left. So, I don't think it's any sort of local magnetic interference.

Sorry for all the details... just trying to answer most of the obvious questions in advance.

Any ideas what is causing this issue? It's not a big deal... and surely not worth shipping the unit back to Sony since it could receive damage in transit. But it would be nice to correct or at least know what could be causing it.

Thanks in advance!
March 24, 2003 3:41:45 PM

This is common on CRT monitors. A typical CRT specification for brightness uniformity is 0. Meaning the brightness can vary up to 0 over the screen. Most CRT monitors Default to 100 candelas, meaning it could be 100 candelas in the center and 70 candelas on the sides. This would be very noticeable difference.

This is one of those specs like convergence that many manufactures will not tell you about.

Jim Witkowski
Chief hardware Engineer

<A HREF="" target="_new"></A>
March 26, 2003 7:46:45 AM

Is it easily adjustable? Does Sony have a secret "technicians on screen menu" like some other displays?

BTW, I'm not sure what you mean by "vary up to 0"... it looks like a typo, but maybe that's what you meant?

Also, her screen is only slightly dim on one side... it's not bright in the center with both sides dim... but maybe you were just using that as an example.

I'm surpised Teq hasn't replied yet... like you, he seems to have quite a bit of knowledge on these issues!
Related resources
March 26, 2003 2:26:24 PM

Sorry I meant to say 30 percent not 0. Yes the variation can also be side to side.

<A HREF="" target="_new"></A>
<P ID="edit"><FONT SIZE=-1><EM>Edited by gosharks on 03/26/03 08:27 AM.</EM></FONT></P>
March 27, 2003 5:34:05 AM

Thanks! I thought maybe my mind was going (still possible!).

Is there a way to easily adjust it (maybe a secret technicians menu)?
March 29, 2003 10:21:13 AM

Yeah they sure don't tell you about that... Worse still, if you look at test screens of pure red, green and blue you will probably find out it's only one of the colours that's whacky.

Most often this is the result of a deflection yoke that's not mounted on center, causing you to have to adjust the monitor into a non-linear portion of the CRT's capabilities.

If it's under warranty, get it fixed, any good tech should be able to re-mount the yoke to minimize the problem.

--->It ain't better if it don't work<---
March 31, 2003 7:59:13 PM


Care to explain your theory on the deflection yoke not being mounted properly. If it where the deflection yoke the geometry would be off, not colors.

Jim Witkowski
Chief Hardware Engineer

<A HREF="" target="_new"></A>
March 31, 2003 9:21:14 PM

Hmmm... can't give you the theory behind it, at least not in a short message... but, if you ever get the chance to remount a deflection yoke you can certainly see the effect I've described.

In simplest terms this has to do with the shadow mask or aperture grill geometry. When the yoke is off center the beam travels an indirect path to the side of the screen affecting the way it passes through the masks, landing off center of the phosphor dots, causing a loss of energy.

There are 4 crucial aspects of (re) mounting a yoke.

1) Placement on the neck.
Get it too far forward and the image will be wildly distored along it's center looking something like a magnifying glass. Get it too far back and it begins to go circular on the edges. Obviously because you are changing deflection geometry, purity and convergence may end up being beyond adjustment if you don't get it right.

2) Vertical centering.
Get it too high and the image bows upwards and becomes impossible to converge. Get it too low and the image bows downward and again is impossible to converge. While errors in vertical positioning are less impacting on purity, in extreme cases you can get the wrong colour being displayed.

3) Horizontal centering.
Too far left produces a dark patch on the right side of the screen that might even be the wrong colour. Too far right... same problem on the left side of the screen. This is wildly impacting on purity, (eg.) blue patches will appear in the corners of a red test screen and variations in brightness of pure colours become very apparent.

4) Rotation.
This is pretty obvious, turn the yoke and watch the image turn with it. Needless to say the yoke needs to be rotated so the image is straight inside the display area of the tube.

These settings are crucial. If you want a good result, tolerances are a couple of millimeters at best.

Next time you scrap a monitor, cut the yoke loose and play with it... You can easily observe these effects by working with pure colours while moving the yoke around on a picture tube. Try it some time, you'll see what I mean.

--->It ain't better if it don't work<---
March 31, 2003 11:15:27 PM

You do not need to tell me how to mount a yoke and what it does. I would like you to explain why you think this would only affect one color and not all three. If the yoke caused the original poster problem I would expect all three colors to be off or a convergence issue not one like you indicate. Odds are the yoke / mounting has nothing to do with the purity problem the original poster reported.

Chief Hardware Engineer

<A HREF="" target="_new"></A>
March 31, 2003 11:45:24 PM

I see... I know you are very knowledgeable and obviously a good tech, so forgive me when I ask if you have actually removed and remounted a Yoke?

(I don't mean lifting it off the rubber wedges then plunking it back in place... I mean completely realigning the entire magnetic system with corrector wands, squaring magnets, convergence rings and all.)

The first and most sensitive part of the process is getting the colour purity right. This can only be done with pure colour test screens and a lot of patience because almost everything affects the luminance and colour purity of the display. Even minor misplacements --and I'm talking millimeters here-- can have very noticeable effects on colour purity.

Think about what you are doing... aiming a beam of electrons at a pin-prick sized hole, 7 inches away so that it hits a .1mm dot of coloured phosphor about an eigth of an inch on the other side of the hole. Misalignments as small as a degree or two will move the beam off the center of the phosphor causing it to hit the black area between dots reducing the brightness of the phosphor... larger angular errors can result in the beam hitting the wrong dot altogether. When you move the yoke you change the angle at which the electron beam passes through the mask to the phosphor... It's a simple matter of geometry.

A quick way to see this in action is to degauss your screen... see how the colours go all stupid while it's working? That's because interference from the magnetic field of the degaussing coil is changing the angle of the electon beams.

You are correct that convergence and focus are also affected by yoke misalignments but purity will be totally screwed up long before that happens.

--->It ain't better if it don't work<---
April 1, 2003 3:12:01 PM

You seem to be missing my point. Look at the original posters problem and tell us how you think the yoke could only affect one side of the screen. If it where a yoke problem I would expect it to affect the entire screen. Isn’t the variation in the phosphors (brightness uniformity spec of the CRT) itself more likely to be the problem in this case?

Jim Witkowski
Chief Hardware Engineer

<A HREF="" target="_new"></A>
April 1, 2003 4:17:10 PM

It certainly could be a variation in the CRT itself. But, given the precision with which CRTs are made and the sloppy way monitors are assembled (using jigs rather than visual adjustment) I think it far more likely to be an assembly error than a CRT manufacturing error.

In fact, if you are looking at luminance variation specs for a monitor rather than for the CRT device itself, it is likely you are looking at a spec that results from "acceptable" variations in yoke placement during assembly. The problem is that because of variations in the windings of the yoke itself, the magnetic center does not always coincide with the physical center of the yoke... Using jigs during assembly, as most manufacturer's do, fails to take this into account.

A slightly misplaced yoke will not affect the entire screen. One side will dim first. It's like shining a flashlight down a large pipe... angular change will affect the illumination of the sides of the pipe long before you lose illumination in the center.

I've re-mounted/replaced quite a few yokes in my day, mostly because of convergence problems, so I have seen this effect first hand many times.
You really do have to try this to see what I'm talking about...

Get a pure red screen, full screen on a monitor, loosen the yoke, remove the wedges, and move the front of it horizontally and watch what happens... repeat the test with pure green and pure blue. Even a couple of millimeters of motion produces clearly noticeable effects on luminance and at some point (probably near the limits of motion) purity goes for a crap and you will actually get the wrong colours being displayed.

--->It ain't better if it don't work<---
April 3, 2003 5:54:56 AM

Many thanks to both Teq and Jim for your comments. This post will be a tad long, so I'll break it up into three parts.

Here's what I've gathered so far: maybe it's a "variation" issue (Jim's opinion) or maybe it's a "yoke not adjusted perfectly" issue" (Teq's opinion).

As far as the yoke thing go (isn't that the center of an egg :-) ), Teq mentioned that it could be from having to adjust the monitor using the panel controls to such an extreme that it the electrons were hitting at a too steep of an angle (did I get that right?). So if I got the settings from her monitor(when the monitor is set to look nearly perfect), would that help to narrow it down to see if that's it? To save a step, here they are (all out of 100):
Vertical Centering: 56
Horizontal Centering: 51
Vertical Size: 46
Horizontal Size: 79

I can also get you the geometry ones if you want (they range from 31 to 56). And the convergence options (horizontal and vertical) are both set at a PERFECT 50.

As you can see, the centering is nearly perfect (only required extremely tiny adjustments with the controls). The horizontal size is a bit high, but that's uniform, so I don't think it matters (but I don't know squat compared to you guys).

These values, to the uneducated mind (mine), seem to indicate that a misaligned yoke is unlikely. But maybe I'm not understanding something.

I did notice that the on just a couple of images that consist of entirely a solid color (perhaps dithered, I'm not sure), the moire pattern is extremely noticeable. It usually is really noticeable when you highlight an image in a web browser (which probably does some sort of dithering to get the highlighted effect). It's not noticeable on normal images at all (so it's not a big deal). I tried the moire adjustement, but it makes everything blurry. Any ideas?

Whatever the cause of the problem of the brightness issue (the real issue), is there anything to do to adjust it myself? I know some monitors have "secret" menus for technicians. Does the Sony G-400 have any menus like that?

If it's a brightness variation issue, is there any way to adjust that at all? (I would guess that it would require a new tube, which would not be a realistic possibility.)

If it's a yoke issue, are there any adjustments that can be made that would help (besides sending it in to Sony to have a professional readjust the yoke)? I'm thinking perhaps through the standard user controls on the screen (which are actually more like a gamepad control), or through an "on screen display" menu (besides the obvious... it is adjusted correctly for everything else).

The monitor is under warranty, but I believe Sony requires you to pack it up and send it to them (I'm not sure, but somehow I doubt onsite service is provided by Sony under warranty). Sending it in probably wouldn't be worth it in this case because I think the risk of messing up other adjustments during shipment would outweigh the benefits of possibly fixing a slight problem (plus the hassle of shipping a 50-60 pound huge box). Besides the small flaw discussed in this thread, this monitor is the most perfectly adjusted monitor I've ever seen!
April 3, 2003 7:22:42 PM

First let me point out that either Jim and I could be correct in this case. Fact is we could both be right. The debate was not over your monitor but rather over the technical aspects of monitor adjustment.

I have seen CRTs with luminance flaws but they are far less common than a mis-mounted deflection yoke and in some cases I've even been able to "cheat" and deliberately misadjust the yoke to compensate.

Sadly, we are not discussing anything you can adjust or tweak without pulling the monitor all apart and doing some serious work inside. If Jim is right it'd probably be cheaper to buy a new monitor. If I'm right the monitor has to be completely disassembled and re-adjusted visually, which is going to carry a rather hefty labour bill.

The deflection yoke is a very complex coil of wires spun around plastic and metal forms that is attached to the back of the picture tube (CRT) inside the monitor. It's held in place by a combination of clamping rings and rubber wedges that are supposed to keep it stable. What it does is generate a magnetic field that moves a flying dot across the screen in lines to form the image. As I mentioned in other messages, this is tricky business requiring a surprising degree of precision to work right.

(In fact eliminating this one finicky part played a key role in the push to produce cost-effective LCD displays.)

There is literally nothing you are going to do about a misplaced yoke or a bad CRT from the user menus.

As for the moire lines... try differing resolutions and refresh rates, see if you can find a reasonable combination that minimizes them. Once you do that, most monitors have "moire cancellers" you can adjust to further minimize the interference.

--->It ain't better if it don't work<---