24" Widescreen CRT (FW900) From Ebay arrived,Comments.

spacediver said:
Actually been looking into signal generators. From what I gather, WinDAS won't ask me to to generate anything beyond 1920x1200 @ 85hz, which means that I need a programmable signal generator that can do about 300 Mhz, and is compatible with CRTs.

But it seems that different signal generators have different test patterns - which test patterns will I need for WinDAS?
Click to expand...

I can't remember honestly. Just run WinDAS and check for yourself. Focus needs the MEME pattern, which should be on most generators. Crosshatch will need both White and RB (magenta). Keep in mind that there are two different levels of convergence adjustment on Dynamic Convergence - Red with respect to Blue (use Gun_RB to do this), and Green with respect to Magenta (white crosshatch should do this). But you can't really do convergence until you get the geometry correct.
 
Thanks, that's good to know jblt.

I think the grayscale pattern is a strip of 16 vertical bars, going up in increments of approximately 7% signal from black to white. I'm gonna pay more attention during the G2 adjustment instructions in the white point balance procedure. If I'm not mistaken, it's asking to crush the first two levels together (which is what you do when you're calibrating for RGB 16-235). If so, that doesn't seem to make sense, as you'd end up with crushed blacks when working with RGB 0-255.
 
Also discovered that a few of those supported generators are actually PC based video generators - that is, they're essentially specialized video cards.
 
jbltecnicspro said:
Dynamic Convergence adjustments on WinDAS requires the FW-900's Mode 5 - 1920x1200 at 85hz with both horizontal and vertical polarity set to negative.
Click to expand...

At least on my video card (Geforce GTX 660), I can set both polarities to negative and it works fine. I may just stick with the video card solution. However, I would caution that anyone doing this ensure that all video card LUTs are at default settings (for example, digital vibrance in Nvidia should be at 50%), and that the software you're using to display the test patterns is not utilizing ICC profiles for gamut correction, etc.
 
spacediver said:
Also discovered that a few of those supported generators are actually PC based video generators - that is, they're essentially specialized video cards.
Click to expand...

That is correct. There is an ISA version of the Quantum Data 801 GG generator. But honestly - like everything else, I would get the external version.
 
spacediver said:
At least on my video card (Geforce GTX 660), I can set both polarities to negative and it works fine. I may just stick with the video card solution. However, I would caution that anyone doing this ensure that all video card LUTs are at default settings (for example, digital vibrance in Nvidia should be at 50%), and that the software you're using to display the test patterns is not utilizing ICC profiles for gamut correction, etc.
Click to expand...

I wouldn't do it on the video card. At least with my video cards, none of them could fully synchronize to Mode 5. They would always be slightly off and force the monitor into a different mode. This also happened with my P991 Trinitron. The only real way to get it is through the signal generator, I'm afraid.
 
Unless I'm mistaken, isn't mode 5 the recommended mode for operation (1920x1200, 85 hz, etc.) If so, that's what I run my unit at all the time.

What mode are you running in now?
 
spacediver said:
Unless I'm mistaken, isn't mode 5 the recommended mode for operation (1920x1200, 85 hz, etc.) If so, that's what I run my unit at all the time.

What mode are you running in now?
Click to expand...

Is there a way to tell just using the monitor? I can't tell right now without WinDAS. You may think you run your monitor in Mode 5, but it may not be. Maybe my version of WinDAS is jacked... ?
 
spacediver said:
Also, from what I can tell, the P991 only has 3 modes. The FW900 has 9, and the CPD G520 has 4.
Click to expand...

No. P991, according to WinDAS, definitely has more than just three modes. I can't remember the exact number (sorry for my terrible memory, lol) but I know for a fact that there were more than just three modes to configure in WinDAS. But you can set the default geometry settings for each mode that the monitor has - and the P991 definitely has more than just three modes to calibrate.

EDIT: If you want - I can check tomorrow and verify about the mode count. But you're right I think - the FW900 had 9 modes to calibrate in WinDAS. P991 definitely had more than 3 - I think it was like 5 or 6. Right now I've had a little too much champagne (I haven't had a drink in seemingly ages, so it seems that I'm a bit of a lightweight right now :D), so I can't really set all of this up. But I'd be happy to do it tomorrow for you.
 
jbltecnicspro said:
No. P991, according to WinDAS, definitely has more than just three modes. I can't remember the exact number (sorry for my terrible memory, lol) but I know for a fact that there were more than just three modes to configure in WinDAS. But you can set the default geometry settings for each mode that the monitor has - and the P991 definitely has more than just three modes to calibrate.
Click to expand...

I was going on the assumption that the P991 is a rebranded G400 - and the service manual for the G400 only lists 3 modes. Perhaps the P991 is not actually a G400, or more likely, the service manual only lists some of the modes.

You sure you selected the correct model in the windas setup?

jbltecnicspro said:
EDIT: If you want - I can check tomorrow and verify about the mode count. But you're right I think - the FW900 had 9 modes to calibrate in WinDAS. P991 definitely had more than 3 - I think it was like 5 or 6. Right now I've had a little too much champagne (I haven't had a drink in seemingly ages, so it seems that I'm a bit of a lightweight right now :D), so I can't really set all of this up. But I'd be happy to do it tomorrow for you.
Click to expand...

Don't trouble yourself, but if you do end up calibrating it in the future, let us know.
 
jbltecnicspro said:
Is there a way to tell just using the monitor? I can't tell right now without WinDAS. You may think you run your monitor in Mode 5, but it may not be. Maybe my version of WinDAS is jacked... ?
Click to expand...

Well, in the service manual, the parameters for Mode 5 seem to match my timings in Nvidia control panel. Hard to say exactly, because some of the terms and units seem to be different. But I imagine that if WinDAS is requesting a specific mode, it is one that the monitor supports as a default standard.
 
I looked back through the thread and noticed you said you had your FW900 in mode 13. What makes you think it was in Mode 13? Where did you select Mode 13?

I find that by simply selecting a resolution and refresh rate, and allowing the timing to be automatic, it'll set everything up in a way that works. You can also switch around a few things, and use some different presets, like "reduced blanking", etc.

But nowhere do you actually select a "mode". You using Nvidia?
 
spacediver said:
I looked back through the thread and noticed you said you had your FW900 in mode 13. What makes you think it was in Mode 13? Where did you select Mode 13?

I find that by simply selecting a resolution and refresh rate, and allowing the timing to be automatic, it'll set everything up in a way that works. You can also switch around a few things, and use some different presets, like "reduced blanking", etc.

But nowhere do you actually select a "mode". You using Nvidia?
Click to expand...

I didn't select it. I did as you did with my GeForce GTX-560. I set the timings like WinDAS says, and it told me that instead of running in Mode 5 - it ran in Mode 13. :confused:
 
WinDAS did. Before you do anything, it will tell you what mode the display is in (this way you can be sure it's set correctly).
 
Ah, good to know! I never noticed that, but will look out for it next time I run it. I haven't yet tested the dynamic convergence or geometry yet, maybe that's when windas tells you the mode you're in.
 
spacediver said:
Ah, good to know! I never noticed that, but will look out for it next time I run it. I haven't yet tested the dynamic convergence or geometry yet, maybe that's when windas tells you the mode you're in.
Click to expand...

Each time before a procedure, it will give you the specs to set your signal generator. In this window (I could screenshot it later), it will have a MODE: at the bottom corner. This tells you the current mode of the display.
 
ah, understood - thank jbl, appreciate your help :)

I'm gonna give windas another shot tomorrow - will report back on my results!
 
spacediver said:
Also discovered that a few of those supported generators are actually PC based video generators - that is, they're essentially specialized video cards.
Click to expand...

Those are not video cards. They are signal generators in a PC board form factor for the ISA bus.

Unkle Vito!
 
Last edited:
I had a couple questions:


1:Is the grayscale pattern that WinDAS is asking for similar to the following:

23sezbq.png



2: If it is the correct type of pattern, why is it asking the operator to crush the first two levels together? That should only be done when working with RGB 16-235. I guess if you're connecting your FW900 directly to a video source, then that would be appropriate, since black level would be encoded at RGB level 16 (7.5 IRE). But if you're working in a PC RGB environment with 0-255, then if you crush those two bars together, you'd end up with crushed blacks, since everything from 0-16 would be black.

How do you deal with this issue?
 
I successfully ran through a complete white balance procedure on the P275. Managed to hit all the targets almost dead on. I think it took close to 2 hours! After I was done, I ran through a grayscale test in HCFR - the delta E errors were way off, but I think this may be because the monitor hadn't rewarmed up after I saved the dat file. Didn't have time tonight to wait, so will check results soon.

edit: I also kept records of each of the settings I used, in case I want to repeat the process in the near future and have a starting point reference. I recommend this as it can save a lot of hassle :)
 
So I recently dug up my old FW900 monitor; turns out the blurring problem is gone. Maybe all it needed was some time to dry up.

However, I have another problem; whenever I try to configure it via WinDAS I get a message saying "ECS syntax error, NG!! NG!! NG!! NG!!" And I have the right model selected, the right drivers installed. This has never happened to me before but somehow it's happening and I even put the RS2332 cable in right?

What do I do, any suggestions?
 
I find my copy of windas somewhat temperamental. When I try to initiate a procedure sometimes it just crashes. But once it gets going, it works fine for me (so far).

I'm using the version found here and I'm running XP.
 
Kermie said:
So I recently dug up my old FW900 monitor; turns out the blurring problem is gone. Maybe all it needed was some time to dry up.

However, I have another problem; whenever I try to configure it via WinDAS I get a message saying "ECS syntax error, NG!! NG!! NG!! NG!!" And I have the right model selected, the right drivers installed. This has never happened to me before but somehow it's happening and I even put the RS2332 cable in right?

What do I do, any suggestions?
Click to expand...

How long has your blurring problem been going on? It could be a flyback transformer issue. Even if it's gone right now, it could definitely come back.
 
LAGRUNAUER said:
Performing a hardware calibration white point balance adjustment in WinDAS is not a walk in the park, and you must have the proper instrumentation and the proper software to do that.

If you want to do it right, besides having a decent laboratoty grade colorimeter probe, you must have a video signal generator capable of displaying the correct patters at the right dot clock, bandwith and frecuency. We use the Sencore CP5000 and program to calibrate CRTs and LCDs, Sencore CR7000 for CRT emission and gun integrity tests; Sencore VP403 and Quantum Data GX Series Video Signal Generators for generating patterns.
Click to expand...

According to the quantum data manual, the GX model supports up to 150 mhz.

Mode 2 on the CPD -G520 (1600x1200@85hz) requires about 230 Mhz (which the GF model can handle, but not the GX).

Mode 5 on the GDM-FW900 (1920x1200@85hz) requires around 283 Mhz.


So would the Quantum Data GX be useless for a CPD G520 and GDM-FW900?


What about the Sencore VP403? What is the highest pixel clock it supports?
 
also, looking through the test patterns of the Sencore CP403, which one would you use when WinDAS asks for "black". Would you just use the pluge pattern and aim the probe at the black portion of the screen?
 
jbltecnicspro said:
How long has your blurring problem been going on? It could be a flyback transformer issue. Even if it's gone right now, it could definitely come back.
Click to expand...

A couple of months; I retired it because of this issue but it seems to have fixed itself. I'll keep an eye out to see if it turns up again.
 
spacediver said:
According to the quantum data manual, the GX model supports up to 150 mhz.

Mode 2 on the CPD -G520 (1600x1200@85hz) requires about 230 Mhz (which the GF model can handle, but not the GX).

Mode 5 on the GDM-FW900 (1920x1200@85hz) requires around 283 Mhz.


So would the Quantum Data GX be useless for a CPD G520 and GDM-FW900?


What about the Sencore VP403? What is the highest pixel clock it supports?
Click to expand...

The Quantum Data GF will support Frequency Ranges of 3.9975 MHz to 250 MHz. The QD 802R will support 3.9975 MHz to 400 MHz (we have a few of them).

We use Sencore VP403 generators (1.5-250kHZ horizontal timing) for field testing only while on service calls and they help us to evaluate the units before we issue work orders for calibration/repair.

For actual calibration and adjustments, we have top of the line RS & Textronix generators.

Hope this helps...

Sincerely,

Unkle Vito!
 
spacediver said:
I successfully ran through a complete white balance procedure on the P275. Managed to hit all the targets almost dead on. I think it took close to 2 hours! After I was done, I ran through a grayscale test in HCFR - the delta E errors were way off, but I think this may be because the monitor hadn't rewarmed up after I saved the dat file. Didn't have time tonight to wait, so will check results soon.

edit: I also kept records of each of the settings I used, in case I want to repeat the process in the near future and have a starting point reference. I recommend this as it can save a lot of hassle :)
Click to expand...

On the GDM-FW900, the entire process of performing a white balance adjustment correctly and without cutting corners can takes anywhere from 4 to 6 hours. Two (2) of those hours are used on letting the monitor warm-up (even though WinDAS/WinCATs states 60 minutes, we've obtained better results with the unit warmed up 120 minutes); and another 40 minutes to perform a "test run" to check how far the units are off targets. In addition to this, we run diagnostics on the guns, shorts and emissions to further check if the units can even be calibrated.

We use Sencore OTC-1000 and the Chroma 5 pods along with the program that comes with the instrumentation, and set the sensitivity of the instruments to "high", then we let the instruments take the measurements and readings until NO fluctuation is detected. That is very critical on performing this process! And that takes time, usually 20-40 minutes per parameter. There are four (4) parameter readings per reference point, so that will be twelve (12) readings. So do the math...

That is the correct way to do this process in order to achieve Delta E of +/- 0.05% which are my standards of very monitor that leaves my lab.

I don't expect the same for the enthusiasts trying out and venturing into the world of WinDAS, but if after performing white balance adjustments, the values obtained are way off specs and/or Delta Es above 5% of all reference point targets, then the process was done incorrectly and the results obtained are garbage.

Hope this helps...

Sincerely,

Unkle Vito!
 
Last edited:
LAGRUNAUER said:
The Quantum Data GF will support Frequency Ranges of 3.9975 MHz to 250 MHz. The QD 802R will support 3.9975 MHz to 400 MHz (we have a few of them).

We use Sencore VP403 generators (1.5-250kHZ horizontal timing) for field testing only while on service calls and they help us to evaluate the units before we issue work orders for calibration/repair.

For actual calibration and adjustments, we have top of the line RS & Textronix generators.

Hope this helps...

Sincerely,

Unkle Vito!
Click to expand...

thanks, that's useful to know. That QD 802R looks real nice, and it allows for custom images.
 
LAGRUNAUER said:
On the GDM-FW900, the entire process of performing a white balance adjustment correctly and without cutting corners can takes anywhere from 4 to 6 hours. Two (2) of those hours are used on letting the monitor warm-up (even though WinDAS/WinCATs states 60 minutes, we've obtained better results with the unit warmed up 120 minutes); and another 40 minutes to perform a "test run" to check how far the units are off targets. In addition to this, we run diagnostics on the guns, shorts and emissions to further check if the units can even be calibrated.

We use Sencore OTC-1000 and the Chroma 5 pods along with the program that comes with the instrumentation, and set the sensitivity of the instruments to "high", then we let the instruments take the measurements and readings until NO fluctuation is detected. That is very critical on performing this process! And that takes time, usually 20-40 minutes per parameter. There are four (4) parameter readings per reference point, so that will be twelve (12) readings. So do the math...

That is the correct way to do this process in order to achieve Delta E of +/- 0.05% which are my standards of very monitor that leaves my lab.

I don't expect the same for the enthusiasts trying out and venturing into the world of WinDAS, but if after performing white balance adjustments, the values obtained are way off specs and/or Delta Es above 5% of all reference point targets, then the process was done incorrectly and the results obtained are garbage.

Hope this helps...

Sincerely,

Unkle Vito!
Click to expand...

I did another WPB on the P275 today. Let it warm up for 1.5 hours, and then went to work. For the 12 parameters it took me about 5-7 min each, and I hit the targets pretty close. However, when I ran through a check on HCFR afterwards, the delta E's weren't as low as I expected, although they were much better than my previous attempt (I think the settings didn't save last time I did it).

When I have time, I will wait for the readings to stabilize before proceeding with each step, as you recommend.

When you say you get delta E's of under 0.05% do you mean delta xy? Neither are percentage units - delta E is the three dimensional geometric distance between target and actual, in a perceptually normalized color space, and I think delta xy is the projection of that distance onto the 2D xy color space, ignoring perceptual normalization and ignoring luminance. Are you able to get your delta E's below 2, from 10%-100% gray levels?

To be honest, I found the geometry calibration more daunting than the WPB! Measuring the distance with a measuring tape is NO EASY TASK, because of parallax errors. In other words, the distance you measure depends on where your eye is. You have to make sure your eye is really aligned with each reference point before reading off the tape. Very tricky, especially when you're reading off the lower end of the monitor, (which in my case is sitting on my bed at the time, so I had to put my face right against the mattress!).

I've also been making test patterns as I go. When I've made an entire list, I will share them all, in all the relevant resolutions.

I have also found that the mode that WinDAS reports me as using does not match the mode it is requesting, despite everything looking fine, and despite the matching parameters. I was even able to get the 1770x1440 pattern for high freq geometry adjustment. I got the correct polarities, etc. But WinDAS said I was using mode 256. I'm not going to worry about that right now, however.
 
Last edited:
You must log in or register to reply here.
Back
Top