The default value for DRIVE_MAX on my P1130 is 220 and I'm positive nobody touched it before.
There are a lot of myths concerning what can kill or not a tube but the only thing that's really true is a very intense white color aka high brightness aka screen voltage from the flyback transformer.
So, again, NO, a high Contrast will not kill a tube as long as the brightness is in check.
As I mentioned in a previous post, it is wiser and safer to push the contrast to 100 if you want high contrast instead of meddling with various Windas registers that you don't know exactly what voltages they affect. It's a rule of common sense.
Driving the tube at it's highest refresh rate will again NOT damage the tube. These things are electronically protected so as to give you an "Out of Range" message when you try a refresh rate that could be potentially dangerous. The only disadvantage will be that on most Trinitron tubes the image will not be that focused on a resolution's highest refresh rate.
This behavior is not present on Shadow Mask's and I've used in the past 5 years several such CRTs with contrast all the way to 100 and highest refresh possible and they look just as good as new ones (I actually have a CRT warehouse where I can test unused CRTs with the ones that I have).
Well, after an hour is trying to get the 3 wires into the service port and hooked up onto the FW900, everything went as according to plan.
The WinDAS cable works great.
My settings, according to WinDAS, even in the dimmed mode, was G2 at 150 and DRIVE_MAX at 255.
Now I want to know, is it normal that my DRIVE_MAX was at 255 on its own? Did someone else made this change? Because I thought the default would be around 210.
I bumped to G2 to 160, kept the DRIVE_MAX at 255. Everything is like new now. Very bright whites and great contrast. Front brightness setting is about 40 and 85 for contrast for perfect rich color.
Should I turn the DRIVE_MAX down a bit? I think I will drop the G2 to 155. Either way this doesn't make any sense that a 10+ bump in G2 made so much difference in the "dim screen" problem. Very weird.
From what I've read these registers differ across models. Here's the start of the F500R (G1 chassis) register list and their values after a factory restore:
Compare to your model and see if the register names are different.
Mike
I don't have REG1-6 except for REG5 which is at 0. Some values are labeled like G2 and PIN but other values are either labeled SKIP or REG--(-). All the regs are in random places on the list. However the regs do have a relationship to the item numbers like REG130 is number 131 on the list and REG22 is number 23.
Maybe someone with the F99 chassis can help me find the DRIVE_MAX and DRIVE_MIN values?
Well, after an hour is trying to get the 3 wires into the service port and hooked up onto the FW900, everything went as according to plan.
The WinDAS cable works great.
My settings, according to WinDAS, even in the dimmed mode, was G2 at 150 and DRIVE_MAX at 255.
Now I want to know, is it normal that my DRIVE_MAX was at 255 on its own? Did someone else made this change? Because I thought the default would be around 210.
I bumped to G2 to 160, kept the DRIVE_MAX at 255. Everything is like new now. Very bright whites and great contrast. Front brightness setting is about 40 and 85 for contrast for perfect rich color.
Should I turn the DRIVE_MAX down a bit? I think I will drop the G2 to 155. Either way this doesn't make any sense that a 10+ bump in G2 made so much difference in the "dim screen" problem. Very weird.
From what you're saying the monitor looks in good condition. The higher the original G2 value the better.
Leave your configuration as it is, don't change anything anymore. To give you an ideea, here's my P1130 settings:
From what you're saying the monitor looks in good condition. The higher the original G2 value the better.
Leave your configuration as it is, don't change anything anymore. To give you an ideea, here's my P1130 settings:
Well, I had to stop using a p991 because it started making a "snap" noise and the picture will bloom in and out. I've seen this happen before on other CRT's. If the problem gets severe enough it can kill the tube. Anybody have an idea where this snapping noise could be coming from?
I connected my Dell P991 to an older laptop with a PL2303 USB-TTL cable, selected "Save to file" in WinDAS, and was able to transfer to ~95% before it failed with an ECS error and "Bus is NG". Now my monitor doesn't even turn on- it just makes a slight charging noise but the CRT never powers up. The green power LED turns on for about a second and then shuts off. Disconnecting the VGA and TTL cables does nothing to help.
I didn't even try to write any data to the monitor. Is it fried? What did I do wrong? The TX/RD cables were swapped at first but the +5/GND cables were connected properly from the start.
I connected my Dell P991 to an older laptop with a PL2303 USB-TTL cable, selected "Save to file" in WinDAS, and was able to transfer to ~95% before it failed with an ECS error and "Bus is NG". Now my monitor doesn't even turn on- it just makes a slight charging noise but the CRT never powers up. The green power LED turns on for about a second and then shuts off. Disconnecting the VGA and TTL cables does nothing to help.
I didn't even try to write any data to the monitor. Is it fried? What did I do wrong? The TX/RD cables were swapped at first but the +5/GND cables were connected properly from the start.
I had the same thing happen to me except my computer froze when I was in the middle of writing new data into my monitor. What has happened is the monitor didn't get all of the data needed to power up so the monitor refuses to turn on. You will need to get another monitor and hook it up to your laptop and with the power switch on the P991 turned on try to reload the data to your monitor and see if it works the next time.
I had the same thing happen to me except my computer froze when I was in the middle of writing new data into my monitor. What has happened is the monitor didn't get all of the data needed to power up so the monitor refuses to turn on. You will need to get another monitor and hook it up to your laptop and with the power switch on the P991 turned on try to reload the data to your monitor and see if it works the next time.
Success!
I let the monitor sit unplugged for about an hour and reseated the four programming wires on both ends. With this I was finally able to read to 100% and the monitor sprang back to life.
I adjusted G2 down by 8 and then discovered the convergence procedure in WinDAS- I made some rather drastic adjustments to the bottom half of the screen to get rid of a red tinge in the bottom left.
It looks beautiful, esp. for $15! I'll have to set it next to my HP LP2475w (currently in another city) and compare.
Thanks!
Edit: I am a little confused over this design choice though. Why would Sony design a display that renders itself inoperable if you don't finish reading the monitor's settings?
Edit: I am a little confused over this design choice though. Why would Sony design a display that renders itself inoperable if you don't finish reading the monitor's settings?[/QUOTE]
I have never really liked Sony's displays because they break so easily. I am using these Trinitron monitors because I can't find a flatscreen that beats it and also because I have the necessary tools to fix them (brightness, convergence, etc.). I think Sony designed these monitors the way they are right now because they know people like these monitors and will pay good money to get them fixed.
I just upgraded to a new system running Windows 7 64-bit. The msflxgrd.ocx will not register because it is not 64-bit. Is there an alternative? Or do I have to use a parallel installation of XP?
Just picked up a P1110 from an engineer friend which tried to fix it a few years back but gave up on it. I paid him $24 for the beast. I'm currently using a 19" Nokia 920C which is great, but it isn't flat, and the ABL kicks in prematurely (contrast gets lower when transitioning from dark to bright). I'm 20, i'm a student, and i fix LCDs in my spare time, but till now i never found one that beats the good ol' tubes (IPS panels are so darn expensive).
I've battled this beast for a good 4 hours till i got it to work so i'll give a few pointers.
Sometimes WinDAS will not connect to the monitor unless you try to connect immediately after you switch the monitor on.
I have found that my monitor loses the 5v standby rail at a certain point, making WinDAS unable to save the file completely. The solution was to draw a wire from my power supply's 5v line and wire it up to the 5v of the TTL cable.
My P1110 will not load the file immediately after save - i need to turn it off then back on. Otherwise the WinDAS software would hang after the 2nd bar and the monitor would be DEAD. Still, after the WinDAS hang, ending the process, unplugging the power then plugging it back in, and opening it up and selecting Load File again would write the file properly and the monitor would wake up. Phew.
It won't work under 64-bit. You can use an XP installation inside VMWare like i did, or better (since it's free), Windows XP Mode for Windows 7.
Hope this helps. This P1110 was insanely bright with retrace lines visible - i found the G2 value at 156, and i lowered it gradually till i got to 86. Background still isn't pitch black and has a slight greenish tint, but right now i don't even have a spare VGA cable available so i'm only basing my observations on the "no signal" screen. I'll buy a VGA cable tomorrow and see how it looks when connected to the computer.
I also have an IBM P275 at my grandparents' place (same thing basically), and it too has a slight greenish tint until it warms up so i am assuming this to be normal. The P275 i bought 2 years ago IIRC. It has normal brightness, but there are some slight ripples in the picture. Probably dried up capacitors on the signal card, but till now i haven't bothered with it, it looks okay at 1400x1050. Hope i can push this P1110 higher than that, because i run my 19" Nokia at 1440x1080 85Hz and it looks mighty fine. And it is a regular aperture grille.
This P1110 was insanely bright with retrace lines visible - i found the G2 value at 156, and i lowered it gradually till i got to 86. Background still isn't pitch black and has a slight greenish tint, but right now i don't even have a spare VGA cable available so i'm only basing my observations on the "no signal" screen. I'll buy a VGA cable tomorrow and see how it looks when connected to the computer.
I also have an IBM P275 at my grandparents' place (same thing basically), and it too has a slight greenish tint until it warms up so i am assuming this to be normal. The P275 i bought 2 years ago IIRC. It has normal brightness, but there are some slight ripples in the picture.
The greenish tint may be fixed by doing Color Return through the OSD. When I got my P991's, one of them had this slight greenish tint to it. I did a color return and although it did get rid of the green tint the brightness went sky high! (and it was already bright before) If you do the color return there is the chance that the brightness will be way up again. Since you seemed to have a lot of trouble just trying to adjust the screen I don't know if you want to try it. If it doesn't bother you then I would leave it.
Ripples on a CRT monitor. I don't think I have seen that before. My P991's have these skinny lines about 1/2 a cm apart running vertically from top to bottom on the screens. Some lines are more noticeable than others. These lines are most noticeable on the color grey. I have no idea what these lines are or what is even causing them. I don't know if this is what you are referring to or not.
The greenish tint may be fixed by doing Color Return through the OSD.
I did that on the IBM last year i think... Didn't change much. Could simply be an aged tube, as apart from the greenish tint the screen is also slightly brighter, and slightly shifted to the right and to the top until it warms up (about 30 minutes).
As for the P1110, i lowered G2 even more, to 76, and now the greenish tint is also gone. With the lights off, the contrast was incredible.
Ripples on a CRT monitor. I don't think I have seen that before.
It's like tiny vertical waves running thru the picture. Most noticeable on small text and at window edges. They are also present in the OSD so i'm sure it's a signal card issue and has nothing to do with the cable. I'll have to trace the VGA signal path and replace all capacitors it goes thru, that should fix it.
My P991's have these skinny lines about 1/2 a cm apart running vertically from top to bottom on the screens. Some lines are more noticeable than others.
Could be moire - try a moire adjustment pattern. I like Nokia Monitor Test but it's 16-bit so you'll still need a virtual machine of some sort to run it on a 64-bit OS.
I'm off to buy me a VGA cable now. Btw, i also got WinDAS working on my win98 laptop (that was the original plan but it didn't work at first), i had to copy the ocx file in the activex folder into the main WinDAS folder and then i could access the model list. I actually did the final G2 change on the laptop. The monitor still needs 5v supplied externally tho, but that's no biggie as even this ancient laptop (which i bought for DOS games) has an USB port i can draw 5v from.
That didn't seem to do the trick. I'm not worried about it. I use my monitors mostly for gaming so the lines aren't really noticeable. It's when I am reading text like on the internet is when I notice the lines. There is a post earlier in this thread that better describes what I am seeing. I don't know what page it is on though. I'll find it and post it.
I'm back to report... I've gotten the VGA cable and had a stab at tweaking the P1110. I have also found a service manual which has proven useful.
I did the color return. It got brighter again. But, the G2 value hasn't changed! Anyway, it just got only a little brighter so i don't mind. The greenish tint stayed the same. I went into Expert color menu and lowered the green bias, that took care of it. I then adjusted the drive levels for gamma 2.2 using these patterns: http://www.photoscientia.co.uk/Gamma.htm
Btw, a more technical explanation: The "bias" controls should have actually be called "cutoff". They adjust black level of each individual gun, while the gains adjust white level. The proper way to set them is lower the contrast to barely visible and input a greyscale pattern. Adjust the bias controls till the gray is neutral. For the gains, return contrast to maximum, calibrate black level (brightness) properly, and adjust using your favorite gamma patterns and desired target gamma... since few of us have access to a color meter so you can properly set the white point, gamma patterns will do. Make the gains as high as possible while keeping your desired gamma.
I also tweaked the convergence a little using the pots on the deflection. Specifically, i adjusted the vertical balance, because the bottom side would always be worse than the top. It looks very good now. I'm still not 100% satisfied but that was just a quick adjustment so it's okay.
And now the negatives... It's nowhere as sharply focused as my Nokia 920C. Someone even confirmed it on here - as refresh rate increases, a Trinitron gets blurry. This doesn't happen with regular tubes. And yes, he is right.
I'll definitely have to take off the antiglare film, as not only is it scratched, it also has this habit of going blue when it is cleaned with regular glass cleaning products, and you have to wipe it with the cloth till it's completely dry... which is annoying to say the least. Also, the monitor is really #$&* reflective. Even tho the anti-glare film is supposed to PREVENT reflections... If i take it off and i'll be greeted with a laptop-style glossy screen, i'll go crazy.
And the deal killer - it's got the exact same bright shadow on the right of dark stuff that my IBM P275 has. No wonder - it's the same platform. It's not that obvious but still enough to make stuff appear that isn't there, making graphics editing impossible. I looked it up and everything seems to point at the video amplifier, either bad capacitors or simply parasitic capacitance caused by dust. I have another monitor, an Eizo F67 19", that exhibits the same issue, and that one isn't even a Trinitron. I took that one apart now and will see if cleaning that area makes any difference.
Don't know what is wrong but every now and then my memory settings for different resolutions reset. This is the 5th time it has done this now. Although this time I was able to see what was up with the right side of my screen.
For a while I have noticed this brighter area which is along the right edge of my P991 screen. I didn't know what to think of it or what had caused it. All I know is it showed up one day. After the memory settings reset, I found out what was causing the right edge to be brighter. It turns out there is a pinkish line that is for some reason being displayed. I can't see it unless I shrink the screen. My guess is when I stretch the screen either the electron beam or the light is being reflected off the edge of the tube and is showing up as just a bright area. I attached a picture of what the line looks like. However my camera is not picking up the bright area when I stretch the screen.
"I too have an issue with bright monitors (though nowhere near as severe as some in this thread), but I have a cable being shipped to me at the moment so I can fix that without too much hassle - I hope.
The question I have is regarded some faint vertical lines (not the two horizontal ones). They appear across the image as:
Code:
| || | | || | | || |
</PRE>
(Fantastic ascii art, I know).
Imagine the two || together are a thicker line. They are faint and run across the whole screen. They do this on two P991 monitors so it's not a specific thing. It's not too noticable except in when playing games, and then it gets annoying.
What causes this? Is it fixable?
The monitors look "OK" at 20 brightness, 90 contrast. I'm hoping playing with the G2 will allow me to get some better definition."
This is what I am seeing, too. However, they don't appear in that pattern. Instead each line is about 1/2 cm apart. I would like to know what causes this. Like I said I am not trying to fix it.
This is what I am seeing, too. However, they don't appear in that pattern. Instead each line is about 1/2 cm apart. I would like to know what causes this. Like I said I am not trying to fix it.
Are the lines ALWAYS in the same pattern? If so, this is consistent with either a bad signal cable, a bad signal source (try another computer), or cold joints inside the monitor. Give it a healthy smack on the back and see if it changes any. If it does, it's time to do some soldering.
As for the pink line, i'm quite sure that one is coming from your video card and not from the monitor. Again, try it on another computer, or try another video card.
Are the lines ALWAYS in the same pattern? If so, this is consistent with either a bad signal cable, a bad signal source (try another computer), or cold joints inside the monitor. Give it a healthy smack on the back and see if it changes any. If it does, it's time to do some soldering.
As for the pink line, i'm quite sure that one is coming from your video card and not from the monitor. Again, try it on another computer, or try another video card.
The pink line is coming from the video card. I'll decide if I should return it since I just unboxed it 2 weeks ago.
As for the lines, they change with the movement of the grille inside the tube. I think the cause of the issue is the grille. Only way to fix that is replace the tube.
Btw, you did mention about using XP mode on a 64-bit machine. I tried that a month ago and everything was fine except for some reason it isn't detecting my adapter. Are there supposed to be separate drivers installed inside the XP operating system? If so, it is not seeing my adapter in its device manager. Did you have any issues with yours?
I've just finished checking all capacitors in the Eizo with an ESR meter. All of them are good. By the looks of it i'll probably know everything about CRT monitors and then some till i get to the bottom of this issue...
The good news: The Dell has gotten better in this regard. I'm currently looking for the bright shadows and i can't see them unless i REALLY try to.
I've just finished checking all capacitors in the Eizo with an ESR meter. All of them are good. By the looks of it i'll probably know everything about CRT monitors and then some till i get to the bottom of this issue...
The good news: The Dell has gotten better in this regard. I'm currently looking for the bright shadows and i can't see them unless i REALLY try to.
Who knows? Maybe you wiggled something. Have you tried your own method of smacking the back of the monitor to see if it changes any?
Also if you have a good camera could you get a screenshot of what the ripples look like?
Will post pics when i get the chance, like i said the IBM is in a different place, different city actually.
As far as slapping it goes, since i was the one who came up with it, it's obvious that i tried it, to no effect. Anyway, i've looked around more and the bright shadows are caused by ringing on the signal, which is due to improper terminations (bad resistors) or bad wiring *somewhere* in the signal path. Unfortunately the VGA signal is very complex, which makes my 100MHz analog oscilloscope basically useless in determining anything helpful. It'll take a while... but i will get to the bottom of this eventually. The ripples on the IBM on the other hand, are most likely caused by bad caps.
I'll also inspect the reason why these monitors start up brighter and slightly offset and then gradually come down to a normal picture. My P1110 takes several hours to 100% stabilize the position... I believe this is directly linked to the G2 issue which causes the abnormally bright screen, but these things don't quite like G2 being measured directly because it is a high impedance signal, so they'll shut down if you attempt measurement, because the internal impedance of the multimeter brings the G2 voltage down. I'll try to devise a solution for that.
Btw, anyone noticed that small monitors always tend to work better than bigger ones? Have here a Sampo Alphascan 718, which is a 17" CRT. Crystal clear at 1280x960 / 72Hz.
Guess what? I'll be getting another 21" Trinitron in working condition. It's an IBM, not sure if it's the P275 or not... Anyway, can't beat the price: FREE.
This is perfect - i can tinker away at one of them while i use the other one as my main monitor. Expect lots of fixes in the next month.
Does anyone have working Windas? Seems all the dl links are dead...
I'm working on Sony F520 which has the usual G2 problem. However I'd like to ask if anyone knows if the "dynamic contrast/brightness" feature is something you could turn off from Windas or with some other trick. What I mean is the feature that adjusts brightness dynamically based on what I'm viewing, (web page vs movie for example).
I have it. I'm really really sleepy right now but i promise to post it tomorrow mkay?
As for your "dynamic contrast" problem, that's the ABL (automatic beam limiter) circuit kicking in and reducing the HV supply of the tube to a safe value. When you fix the G2 issue it'll likely go away as well.
Just discovered some more options that can be messed with.
Setting CONT_MIN to 1 will ensure the maximum possible range for the contrast control. I like my monitor very dim at night, and with the default value it didn't go low enough.
Setting CON_MAX_BRT_MAX to the same value as CON_MAX_BR_CENT will not decrease the contrast anymore when you crank the brightness up, at the expense of slightly worse focus. Great for watching movies.
BRIGHTNESS_CENT and BRIGHTNESS_MAX are self explaining i believe.
Btw, have 100% confirmed that the bright screen before warmup is NOT a tube problem - i've had the monitor on all day and now after messing with WinDAS the screen was darker for a minute or so from when WinDAS shut it down to when it came back up - ie just like it should be. It probably traces to the same resistor people change to fix the G2 problem without using WinDAS. I'll try and take this beast apart next weekend and replace that resistor to see if the high initial brightness will be cured.
In May 2006 I bought on eBay a used SGI GDM-5411 monitor (22'' flat square Trinitron CRT). Already the vendor turned out to be a fraud; claiming that I would have rejected the parcel, he refused to deliver. So I had to inform the police to get it. When the monitor arrived, reinforcements of the plastic swivel pod were cracked off and the case was grey instead of blue, but it worked. Unfortunately black was never as dark as it should, and when the monitor or room was cold, the picture turned pale, which I also could not really fix by graphics card setting without loosing black details. So I used the "KOMPLEX" (expert) mode to set grey and white balance manually according to a downloaded screen test pattern. But as we all know, when a Trinitron monitor gets older, it turns even brighter, so now black became light grey with visible diagonal retrace lines. The lines first showed up on the "no signal" mode and low DOS or Amiga 50Hz (with scandoubler) resolutions, but finally also appeared on Windows 98SE. When I turn the cold monitor on, also screen size first changes with brightness (especially in 50Hz Amiga mode) and occasionally the picture quickly zooms a bit with a soft popping noise; apparently condensed air humidity after venting the room caused a high voltage discharge somewhere. (Yes, CRTs do this occasionally when turned on after a while. I have seen that with about a dozen of well working TVs and monitors. Its likely dust and humidity.)
Strange is that during the "colour return" test algorithm the picture temporary steps dark enough to hide the grey and retrace lines, how ever it finishes with adjusting everything too bright again. And in hot summer the problem seems to be less severe. So I concluded that the CRT can not be that faulty and downloaded the service manual. Finally I websearched for a solution and found this huge thread (DELL P1110 monitor too bright) on the Icrontic forum. (Yes, I fully read it to the end.)
I have survived 6 semesters of electrical engineering, so some statements I read on this forum really makes me shudder...
- high voltage won't kill you
Especially the claim that the stored high voltage from a monitor plugged out of mains for hours will kill you is ridiculous, so stop spreading this false urban legend! I do not say it is impossible; people have already died from a baseball hitting their heart in a wrong phase causing fibrillation, or from taser guns those were sold as "non lethal", so never say never. But the chance of dying from the stored "high voltage" is like panic of getting smashed to death by moving the heavy monitor on a desk.
Picture tubes indeed work with 20 to 30kV, but this is not at all like the instantly deadly 20kV high voltage electric locomotives work with. While touching a train power line or 3rd rail indeed has the vigour to burn humans to ash, the current of CRT high voltage is way too low, and because it is DC it is also unlikely to stop a heart. True is that when disconnected from mains, this voltage stays electrostatically stored inside the CRT between the outer and inner conductive layer on the glas like in a capacitor (websearch for "Leyden jar"), but it stays only inside the CRT and the thick cable with the big suction cup that runs to the flyback transformer. Unless you intend to remove or destroy any of these parts (only necessary for replacing CRT or flyback or recycling the whole thing) you will never get in touch with it. While I definitely will not recommend anybody to try that as a courage test, even if you touch it, it would only hurt badly and leave a burnt spot on your finger - as dangerous as sticking fingers into a candle flame. You may know, a flyback is called flyback because the jolt makes you "fly back" some meters. So the worst that may happen is that you jerk so badly that you smash the CRT neck (or touch other dangerous voltages if you are stupid enough to have it plugged into mains). Very hypothetically this might also make the CRT implode and stab you to death by flying shards, but this would be purely a freak accident since modern CRTs are designed not to implode into thousand glass shrapnel even when a TV falls from a truck.
(If you really want to safely discharge a CRT before working, disconnect the monitor from mains, connect a well grounded cable to the metal frame of the CRT (which is the 2nd pole of the stored voltage) and a cable connected to it with a flat metal contact (with well insulated plastic handle) under the suction cup until you hear the crackling of high voltage sparks.)
With flyback transformers the danger is not so much the DC high voltage that comes out of it (safety circuits would surely turn it of, else it can be about 30mA), but the current of the AC voltage that goes into into it (some 100V AC at some 100mA) when the monitor is on. This would be about as dangerous as mains voltage and should not be touched. But any toaster or table lamp may kill with such a voltage while nobody seems to fear to change a light bulb.
The only dangerous voltage that stays stored inside an unplugged monitor (or PC or big amplifier or what ever) is rectified mains voltage in big electrolytic capacitors of the switching power supply. When these caps are large enough, the DC current from these may theoretically kill, but power supplies are designed with safety resistors those discharge the caps within 15 to 30 minutes. (Decades ago I got a jolt on a 400V charged electrolytic cap in a tube radio with defective resistor. I only felt a sting and it left a black spot on my finger.)
High voltages without much current are harmless. Ask YouTube about "violet ray". These electromagnetic healing devices (as painkiller, skin disinfection, burn warts away etc.) contain a small tesla coil that outputs through a glow bulb lightning bolts at radio frequency with even 50kV(!) those you can touch with bare hands. So stop the hysteria that CRTs are so much more dangeorus because they have "highvoltage" in them. (Did you know that ordinary CFL and fluorescent lamps use thousands of volts also?)
- xray protection
The only thing you really should not do with the high voltage is cranking it up higher than intended by design to increase brightness of a worn out CRT, because this exponentially increases the amount of xray (bremsstrahlung), which goes through walls (also to your neighbour or sleeping childs), can cause cancer and crashes the immune system by preventing sleep due to irritation of the melatonin gland. (The light researcher John Ott warned a lot about this.)
With my monitor the HV adjustment pot is RV901 (HV ADJ).
Quote from the service manual:
When replacing or repairing the shown below table, the
c)Beam Current Protector Circuit Check
following operational checks must be performed as a
(1st Protector): D Board
safety precaution against X-rays emissions from the unit.
...
a) HV Regulator Circuit Check
1) Enter black crosshatch signal (black on white back-
ground), and check that high voltage is in the speci-
fied range.
[Specification]: 27.00 ± 0.10 kV
2) Check that the voltage of D912 cathode on the D
board is 27.0 V or more.
b) HV Protector Circuit Check
1)Enter black crosshatch signal (black on white back-
ground).
2)Apply the specified voltage to the D912 cathode on
the D board, and check that high voltage is 0.1 kV or
less.
[Specification]: 31.90 + 0.00/ñ 0.05 V
I worry that even the too bright Trinitron picture may output unnecessary xray, because xray is at least proportional with beam current.
- my RS232 adapter
That some people here fried their serial ports and ICs by plugging in the adapter cable wrongways without prior measuring is simply stupid. At least you should always measure with a multimeter that +Vs and GND is correct. (Confusing TX and RX won't harm. TX means "data out" RX means "data in", so they always need to be connected to the opposite.) GND can be easily identified with a resistance meter or continuity tester even when the monitor is plugged out, because GND is connected with the metal shielding. Also doing a firmware or flash rom upgrade using a "breadboard" and sticking a bunch of loose individual wires somehow into a jack is like transporting gelignite bottles in shoe boxes stacked 1m high upon a wheelbarrow; one little shake and everything goes boom and can't boot any more.
So I soldered a 2 transistor circuit for the RS232/TTL converter like this one:
I added an old diskette drive power cable. For the RS232 cable I did a deep look in my bag of dead mice and dug out an old serial mouse cable (with screw mount). Because pin 4,6 and 8 need to be connected, I pried the plug front out by screw driver (which tore the leads from it) to solder wires on its back. The plug was potted with hard hotglue that I had to pry out and pull the cable through the plug cover to re-solder it. Because these monitor apparently tend to make a fun of switching its +5V output off in the least suitable moment (much like old Thinkpad laptops needed a full battery during BIOS upgrades), I used a 100 Ohm resistor followed by a zener diode against GND to turn the +12V from the PC serial port into +5V. I remember that the 12V output is high ohmed, so for the LEDs I used 1 kOhm instead of 220 Ohm to reduce power consumption.
I haven't used the adapter yet. I have to check if I can install WinDAS on my IBM Thinkpad 760XD laptop (Win98SE,100MB RAM,166MHz) to see WinDAS menus even when the monitor crashes by failed uploads. I hope the monitor won't fall onto me or break my spine when I make the hole to access the ECS port. The thing weights about as much as me and was a hell job to move at its place.
- eastereggs/service mode
In the service manual stands:
If the (power) indicator is green
1. Remove any plugs from the video input 1 and 2 connectors, or turn off the connected computer(s).
2. Press the 'power' button twice to turn the monitor off and then on.
3. Press the '->' button for 2 seconds before the monitor enters power saving mode.
If all four colour bars appear (white, red, green, blue), the monitor is working properly. Reconnect the video input cables and check the condition of your computer(s).
(Picture shows monitor type and serial number and "WHITE", "RED", "GREEN", "BLUE" each written on a such coloured stripe.)
If the color bars do not appear, there is a potential monitor failure. Inform your service representative of the monitor's condition.
If the 'power' indicator is flashing orange
Press the 'power' button twice to turn the monitor off and then on.
If the 'power' indicator lights up green, the monitor is working properly.
If the 'power' indicator is still flashing, there is a potential monitor failure. Count the number of seconds between orange flashes of the (power) indicator and inform your service representative of the monitor's condition. Be sure to note the model name and serial number of your monitor.
This all does not sound too spectacular. But I found a few strange additional features.
1. The so-called service mode can also be entered by simply switching the VGA port switch to the unused port and holding '->'. The led starts flashing alternately green and orange and I see the following message:
MODEL:GDM-5411
SER NO:2432556
MANUFACTURED:2001-02
(and 4 colour stripes)
2. When I hold the '<-' button, I get a white upright solid white rectangle on a black background, that looks like a crude screen test for something.
3. When I hold the 'menu' button, the whole screen becomes an orange tinted white surface, which slowly grows dimmer. This may be another variant of the "colour return" function.
The up and down buttons seem to do nothing. To avoid trouble, I didn't dare to mess with 'reset' or 'center' button. The service mode is exited by switching back to the VGA port that has a signal. Unfortunately I found nothing sophisticated hidden like advanced convergence or G2 level settings.
- bright warmup fix found?
I read here about the well known resistor fix on the G2 input of the CRT. A look into the schematics reveals that on the A board (CRT neck) the IC406 "cuttoff amp" is connected to something that they call "drift correct". This strange thing consists of the NPN transistor Q406 (2SC2412K) that has its emitter on the shielding GND and its collector on the GND line of IC406. Between emitter and base is R405 (680 Ohm), and between base and collector R404 (2.2k). I guess that the whole thing works as a temperature sensor that changes the GND level of the video cuttoff amp and this way shifts the output voltage of all 3 RGB electron guns when temperature changes. I can well imagine that the transistor (or its resistors?) wears out over time and so makes the Trinitron go brighter when cold or even in general. So messing around with G2 voltage (by resistor or WinDAS) may be the wrong place to fix the symptom.
The IC406 additionally has 3 capacitors C120, C220, C320 (each 0.1uF 250V) on pins 4 (R_BASE), 6 (G_BASE) and 8 (B_BASE) those seem to set reference voltages and may go bad, but I rather suspect the transistor.
- high frame rates & Trinitron blur
I read that at high horizontal frequencies Trinitron CRTs blur more than shadow mask CRTs. I yet always suspected my antique video cards (Riva TNT2 32MB passed through a 3Dfx Voodoo 1) and my homemade VGA switch box (modified "TV Mouse" CVBS converter with RGB spiral cable on Amiga side), but if other people experience this also, it would be plausible that the Trinitron wire mask has to do with it. I imagine that the thin wires may have higher inductance or simply higher resistance than a sheet metal mask and so become the less "visible" to the electron beam (which interacts by a capacitive pulse) the faster it moves.
It may sound funny for some people here, but I anyway run my SGI GDM-5411 only at 64kHz (1280x1024@60Hz), not least to save the flyback. I grew up with 50Hz frame rate (still daily use PAL on my severely modified Amiga 500) and have no problem with screen flicker (unless flu tube lighting pests the room with a beat frequency of additional flicker, so I solely use incan).
Flyback transformers heat up more and die much faster by high horizontal frequencies. (You can even see the energy waste on a socket power meter.) My first Nokia 417TV burnt out after few years by overclocking (it was only rated for 800 lines modes, but on Linux I successfully used 1024 and more). So my new Nokia 417TV is solely used as videophile PAL TV and hasn't degraded much despite it runs roughly 14h per day. Flatscreens jerk way too much and make me seasick and nothing beats a CRT with fast scrolling classic 2D games.
- add PAL/NTSC modes through WinDAS?
Is it possible to reprogram the SGI GDM-5411 with WinDAS to display true PAL resolution (50Hz interlaced = 25Hz) or NTSC (60Hz interlaced = 30Hz) without a scandoubler, TV card or other digital picture degraders?
It would be nice to play also lightgun or Sega 3D Glasses games on it, which is only possible wirh true unaltered PAL output. I still have the old CVBS converter attached (which had worked with a horrible TOEI monitor, NEC MultiSync 3d and Nokia 417TV), but the SGI shows only "out of range" on black background when I deactivate the Amiga scandoubler. Also people with arcade PCBs would certainly like to have true PAL/NTSC modes. Is this only a firmware problem, or is the hardware of late generation CRT monitors generally incapable to process RGB signals with such low frame rates? May also be that they considered that a too fine beam would look ugly on low resolution (like staring through window blinds), but this would be fixable by widening the vertical focus of the Trinitron tube in low resolution modes (but certainly hard to hack due to some kV voltage).
MAY THE SOFTWARE BE WITH YOU!
*============================================================================*
I CYBERYOGI Christian Oliver(=CO=) Windler I
I (teachmaster of LOGOLOGIE - the first cyberage-religion!) I
I ! I
*=============================ABANDON=THE=BRUTALITY==========================*
{http://weltenschule.de/e_index.html}
I read here about the well known resistor fix on the G2 input of the CRT. A look into the schematics reveals that on the A board (CRT neck) the IC406 "cuttoff amp" is connected to something that they call "drift correct". This strange thing consists of the NPN transistor Q406 (2SC2412K) that has its emitter on the shielding GND and its collector on the GND line of IC406. Between emitter and base is R405 (680 Ohm), and between base and collector R404 (2.2k). I guess that the whole thing works as a temperature sensor that changes the GND level of the video cuttoff amp and this way shifts the output voltage of all 3 RGB electron guns when temperature changes. I can well imagine that the transistor (or its resistors?) wears out over time and so makes the Trinitron go brighter when cold or even in general. So messing around with G2 voltage (by resistor or WinDAS) may be the wrong place to fix the symptom.
You may be on to something here. I'll check it out these days, unfortunately i haven't had time to mess with it yet. Could you give me a link to the service manual?
Flyback transformers heat up more and die much faster by high horizontal frequencies.
That would be horizontal output transistors. (HOTs) While it is true that the flyback transformer itself gets a bit hotter with higher frequencies, it's the horizontal output transistor (the big one which switches B+ into the flyback) which becomes more and more inefficient as frequency rises.
Is it possible to reprogram the SGI GDM-5411 with WinDAS to display true PAL resolution (50Hz interlaced = 25Hz) or NTSC (60Hz interlaced = 30Hz) without a scandoubler, TV card or other digital picture degraders?
It's not, and this time it really has to do with the flyback. All transformers have a minimum operating frequency depending on core material, size of air gap and number of primary turns. If the frequency is too low, the core saturates. If the core would be allowed to saturate the HOT would blow up.
Comments
The WinDAS cable works great.
My settings, according to WinDAS, even in the dimmed mode, was G2 at 150 and DRIVE_MAX at 255.
Now I want to know, is it normal that my DRIVE_MAX was at 255 on its own? Did someone else made this change? Because I thought the default would be around 210.
I bumped to G2 to 160, kept the DRIVE_MAX at 255. Everything is like new now. Very bright whites and great contrast. Front brightness setting is about 40 and 85 for contrast for perfect rich color.
Should I turn the DRIVE_MAX down a bit? I think I will drop the G2 to 155. Either way this doesn't make any sense that a 10+ bump in G2 made so much difference in the "dim screen" problem. Very weird.
Maybe someone with the F99 chassis can help me find the DRIVE_MAX and DRIVE_MIN values?
From what you're saying the monitor looks in good condition. The higher the original G2 value the better.
Leave your configuration as it is, don't change anything anymore. To give you an ideea, here's my P1130 settings:
G2 = 163
DRIVE_MAX = 220
Brightness = 70
Contrast = 100
Variable Color Mode = 11000K
and everything looks fantastic !
However, now that I've fiddled around with DRIVE_MAX, here's what I've found.
At "default" DRIVE_MAX 210, my 93k GAIN according to the OSD defaults to R: 89, G: 76, B: 95.
At DRIVE_MAX 235, my 93k goes R: 73 G: 63, B: 85
DRIVE_MAX 255, gives 93k R: 73, G: 63, B:78
The contrast seems better as the DRIVE_MAX goes up, but slightly more prone to blooming and slight loss of focus.
What's the best DRIVE_MAX to go with in this case? What's best for longevity? Will going to MAX_DRIVE 235 or 255 be bad for aging?
I didn't even try to write any data to the monitor. Is it fried? What did I do wrong? The TX/RD cables were swapped at first but the +5/GND cables were connected properly from the start.
I had the same thing happen to me except my computer froze when I was in the middle of writing new data into my monitor. What has happened is the monitor didn't get all of the data needed to power up so the monitor refuses to turn on. You will need to get another monitor and hook it up to your laptop and with the power switch on the P991 turned on try to reload the data to your monitor and see if it works the next time.
Success!
I let the monitor sit unplugged for about an hour and reseated the four programming wires on both ends. With this I was finally able to read to 100% and the monitor sprang back to life.
I adjusted G2 down by 8 and then discovered the convergence procedure in WinDAS- I made some rather drastic adjustments to the bottom half of the screen to get rid of a red tinge in the bottom left.
It looks beautiful, esp. for $15! I'll have to set it next to my HP LP2475w (currently in another city) and compare.
Thanks!
Edit: I am a little confused over this design choice though. Why would Sony design a display that renders itself inoperable if you don't finish reading the monitor's settings?
I have never really liked Sony's displays because they break so easily. I am using these Trinitron monitors because I can't find a flatscreen that beats it and also because I have the necessary tools to fix them (brightness, convergence, etc.). I think Sony designed these monitors the way they are right now because they know people like these monitors and will pay good money to get them fixed.
I've battled this beast for a good 4 hours till i got it to work so i'll give a few pointers.
It won't work under 64-bit. You can use an XP installation inside VMWare like i did, or better (since it's free), Windows XP Mode for Windows 7.
Hope this helps. This P1110 was insanely bright with retrace lines visible - i found the G2 value at 156, and i lowered it gradually till i got to 86. Background still isn't pitch black and has a slight greenish tint, but right now i don't even have a spare VGA cable available so i'm only basing my observations on the "no signal" screen. I'll buy a VGA cable tomorrow and see how it looks when connected to the computer.
I also have an IBM P275 at my grandparents' place (same thing basically), and it too has a slight greenish tint until it warms up so i am assuming this to be normal. The P275 i bought 2 years ago IIRC. It has normal brightness, but there are some slight ripples in the picture. Probably dried up capacitors on the signal card, but till now i haven't bothered with it, it looks okay at 1400x1050. Hope i can push this P1110 higher than that, because i run my 19" Nokia at 1440x1080 85Hz and it looks mighty fine. And it is a regular aperture grille.
The greenish tint may be fixed by doing Color Return through the OSD. When I got my P991's, one of them had this slight greenish tint to it. I did a color return and although it did get rid of the green tint the brightness went sky high! (and it was already bright before) If you do the color return there is the chance that the brightness will be way up again. Since you seemed to have a lot of trouble just trying to adjust the screen I don't know if you want to try it. If it doesn't bother you then I would leave it.
Ripples on a CRT monitor. I don't think I have seen that before. My P991's have these skinny lines about 1/2 a cm apart running vertically from top to bottom on the screens. Some lines are more noticeable than others. These lines are most noticeable on the color grey. I have no idea what these lines are or what is even causing them. I don't know if this is what you are referring to or not.
I did that on the IBM last year i think... Didn't change much. Could simply be an aged tube, as apart from the greenish tint the screen is also slightly brighter, and slightly shifted to the right and to the top until it warms up (about 30 minutes).
As for the P1110, i lowered G2 even more, to 76, and now the greenish tint is also gone. With the lights off, the contrast was incredible.
It's like tiny vertical waves running thru the picture. Most noticeable on small text and at window edges. They are also present in the OSD so i'm sure it's a signal card issue and has nothing to do with the cable. I'll have to trace the VGA signal path and replace all capacitors it goes thru, that should fix it.
Could be moire - try a moire adjustment pattern. I like Nokia Monitor Test but it's 16-bit so you'll still need a virtual machine of some sort to run it on a 64-bit OS.
I'm off to buy me a VGA cable now.
I have that program somewhere on my hard drive. I'll have to find it. I used that to fix the convergence and the geometry distortions on my monitors.
I did the color return. It got brighter again. But, the G2 value hasn't changed! Anyway, it just got only a little brighter so i don't mind. The greenish tint stayed the same. I went into Expert color menu and lowered the green bias, that took care of it. I then adjusted the drive levels for gamma 2.2 using these patterns: http://www.photoscientia.co.uk/Gamma.htm
Btw, a more technical explanation: The "bias" controls should have actually be called "cutoff". They adjust black level of each individual gun, while the gains adjust white level. The proper way to set them is lower the contrast to barely visible and input a greyscale pattern. Adjust the bias controls till the gray is neutral. For the gains, return contrast to maximum, calibrate black level (brightness) properly, and adjust using your favorite gamma patterns and desired target gamma... since few of us have access to a color meter so you can properly set the white point, gamma patterns will do. Make the gains as high as possible while keeping your desired gamma.
I also tweaked the convergence a little using the pots on the deflection. Specifically, i adjusted the vertical balance, because the bottom side would always be worse than the top. It looks very good now. I'm still not 100% satisfied but that was just a quick adjustment so it's okay.
And now the negatives... It's nowhere as sharply focused as my Nokia 920C. Someone even confirmed it on here - as refresh rate increases, a Trinitron gets blurry. This doesn't happen with regular tubes. And yes, he is right.
I'll definitely have to take off the antiglare film, as not only is it scratched, it also has this habit of going blue when it is cleaned with regular glass cleaning products, and you have to wipe it with the cloth till it's completely dry... which is annoying to say the least. Also, the monitor is really #$&* reflective. Even tho the anti-glare film is supposed to PREVENT reflections... If i take it off and i'll be greeted with a laptop-style glossy screen, i'll go crazy.
And the deal killer - it's got the exact same bright shadow on the right of dark stuff that my IBM P275 has. No wonder - it's the same platform. It's not that obvious but still enough to make stuff appear that isn't there, making graphics editing impossible. I looked it up and everything seems to point at the video amplifier, either bad capacitors or simply parasitic capacitance caused by dust. I have another monitor, an Eizo F67 19", that exhibits the same issue, and that one isn't even a Trinitron. I took that one apart now and will see if cleaning that area makes any difference.
For a while I have noticed this brighter area which is along the right edge of my P991 screen. I didn't know what to think of it or what had caused it. All I know is it showed up one day. After the memory settings reset, I found out what was causing the right edge to be brighter. It turns out there is a pinkish line that is for some reason being displayed. I can't see it unless I shrink the screen. My guess is when I stretch the screen either the electron beam or the light is being reflected off the edge of the tube and is showing up as just a bright area. I attached a picture of what the line looks like. However my camera is not picking up the bright area when I stretch the screen.
"I too have an issue with bright monitors (though nowhere near as severe as some in this thread), but I have a cable being shipped to me at the moment so I can fix that without too much hassle - I hope.
The question I have is regarded some faint vertical lines (not the two horizontal ones). They appear across the image as:
Code:
| || | | || | | || |
</PRE>
(Fantastic ascii art, I know).
Imagine the two || together are a thicker line. They are faint and run across the whole screen. They do this on two P991 monitors so it's not a specific thing. It's not too noticable except in when playing games, and then it gets annoying.
What causes this? Is it fixable?
The monitors look "OK" at 20 brightness, 90 contrast. I'm hoping playing with the G2 will allow me to get some better definition."
This is what I am seeing, too. However, they don't appear in that pattern. Instead each line is about 1/2 cm apart. I would like to know what causes this. Like I said I am not trying to fix it.
Are the lines ALWAYS in the same pattern? If so, this is consistent with either a bad signal cable, a bad signal source (try another computer), or cold joints inside the monitor. Give it a healthy smack on the back and see if it changes any. If it does, it's time to do some soldering.
As for the pink line, i'm quite sure that one is coming from your video card and not from the monitor. Again, try it on another computer, or try another video card.
The pink line is coming from the video card. I'll decide if I should return it since I just unboxed it 2 weeks ago.
As for the lines, they change with the movement of the grille inside the tube. I think the cause of the issue is the grille. Only way to fix that is replace the tube.
Btw, you did mention about using XP mode on a 64-bit machine. I tried that a month ago and everything was fine except for some reason it isn't detecting my adapter. Are there supposed to be separate drivers installed inside the XP operating system? If so, it is not seeing my adapter in its device manager. Did you have any issues with yours?
The good news: The Dell has gotten better in this regard. I'm currently looking for the bright shadows and i can't see them unless i REALLY try to.
Who knows? Maybe you wiggled something. Have you tried your own method of smacking the back of the monitor to see if it changes any?
Also if you have a good camera could you get a screenshot of what the ripples look like?
As far as slapping it goes, since i was the one who came up with it, it's obvious that i tried it, to no effect. Anyway, i've looked around more and the bright shadows are caused by ringing on the signal, which is due to improper terminations (bad resistors) or bad wiring *somewhere* in the signal path. Unfortunately the VGA signal is very complex, which makes my 100MHz analog oscilloscope basically useless in determining anything helpful. It'll take a while... but i will get to the bottom of this eventually. The ripples on the IBM on the other hand, are most likely caused by bad caps.
I'll also inspect the reason why these monitors start up brighter and slightly offset and then gradually come down to a normal picture. My P1110 takes several hours to 100% stabilize the position... I believe this is directly linked to the G2 issue which causes the abnormally bright screen, but these things don't quite like G2 being measured directly because it is a high impedance signal, so they'll shut down if you attempt measurement, because the internal impedance of the multimeter brings the G2 voltage down. I'll try to devise a solution for that.
Btw, anyone noticed that small monitors always tend to work better than bigger ones? Have here a Sampo Alphascan 718, which is a 17" CRT. Crystal clear at 1280x960 / 72Hz.
This is perfect - i can tinker away at one of them while i use the other one as my main monitor.
Does anyone have working Windas? Seems all the dl links are dead...
I'm working on Sony F520 which has the usual G2 problem. However I'd like to ask if anyone knows if the "dynamic contrast/brightness" feature is something you could turn off from Windas or with some other trick. What I mean is the feature that adjusts brightness dynamically based on what I'm viewing, (web page vs movie for example).
Thanks
As for your "dynamic contrast" problem, that's the ABL (automatic beam limiter) circuit kicking in and reducing the HV supply of the tube to a safe value. When you fix the G2 issue it'll likely go away as well.
Setting CONT_MIN to 1 will ensure the maximum possible range for the contrast control. I like my monitor very dim at night, and with the default value it didn't go low enough.
Setting CON_MAX_BRT_MAX to the same value as CON_MAX_BR_CENT will not decrease the contrast anymore when you crank the brightness up, at the expense of slightly worse focus. Great for watching movies.
BRIGHTNESS_CENT and BRIGHTNESS_MAX are self explaining i believe.
Btw, have 100% confirmed that the bright screen before warmup is NOT a tube problem - i've had the monitor on all day and now after messing with WinDAS the screen was darker for a minute or so from when WinDAS shut it down to when it came back up - ie just like it should be. It probably traces to the same resistor people change to fix the G2 problem without using WinDAS. I'll try and take this beast apart next weekend and replace that resistor to see if the high initial brightness will be cured.
Strange is that during the "colour return" test algorithm the picture temporary steps dark enough to hide the grey and retrace lines, how ever it finishes with adjusting everything too bright again. And in hot summer the problem seems to be less severe. So I concluded that the CRT can not be that faulty and downloaded the service manual. Finally I websearched for a solution and found this huge thread (DELL P1110 monitor too bright) on the Icrontic forum. (Yes, I fully read it to the end.)
I have survived 6 semesters of electrical engineering, so some statements I read on this forum really makes me shudder...
- high voltage won't kill you
Especially the claim that the stored high voltage from a monitor plugged out of mains for hours will kill you is ridiculous, so stop spreading this false urban legend! I do not say it is impossible; people have already died from a baseball hitting their heart in a wrong phase causing fibrillation, or from taser guns those were sold as "non lethal", so never say never. But the chance of dying from the stored "high voltage" is like panic of getting smashed to death by moving the heavy monitor on a desk.
Picture tubes indeed work with 20 to 30kV, but this is not at all like the instantly deadly 20kV high voltage electric locomotives work with. While touching a train power line or 3rd rail indeed has the vigour to burn humans to ash, the current of CRT high voltage is way too low, and because it is DC it is also unlikely to stop a heart. True is that when disconnected from mains, this voltage stays electrostatically stored inside the CRT between the outer and inner conductive layer on the glas like in a capacitor (websearch for "Leyden jar"), but it stays only inside the CRT and the thick cable with the big suction cup that runs to the flyback transformer. Unless you intend to remove or destroy any of these parts (only necessary for replacing CRT or flyback or recycling the whole thing) you will never get in touch with it. While I definitely will not recommend anybody to try that as a courage test, even if you touch it, it would only hurt badly and leave a burnt spot on your finger - as dangerous as sticking fingers into a candle flame. You may know, a flyback is called flyback because the jolt makes you "fly back" some meters. So the worst that may happen is that you jerk so badly that you smash the CRT neck (or touch other dangerous voltages if you are stupid enough to have it plugged into mains). Very hypothetically this might also make the CRT implode and stab you to death by flying shards, but this would be purely a freak accident since modern CRTs are designed not to implode into thousand glass shrapnel even when a TV falls from a truck.
(If you really want to safely discharge a CRT before working, disconnect the monitor from mains, connect a well grounded cable to the metal frame of the CRT (which is the 2nd pole of the stored voltage) and a cable connected to it with a flat metal contact (with well insulated plastic handle) under the suction cup until you hear the crackling of high voltage sparks.)
With flyback transformers the danger is not so much the DC high voltage that comes out of it (safety circuits would surely turn it of, else it can be about 30mA), but the current of the AC voltage that goes into into it (some 100V AC at some 100mA) when the monitor is on. This would be about as dangerous as mains voltage and should not be touched. But any toaster or table lamp may kill with such a voltage while nobody seems to fear to change a light bulb.
The only dangerous voltage that stays stored inside an unplugged monitor (or PC or big amplifier or what ever) is rectified mains voltage in big electrolytic capacitors of the switching power supply. When these caps are large enough, the DC current from these may theoretically kill, but power supplies are designed with safety resistors those discharge the caps within 15 to 30 minutes. (Decades ago I got a jolt on a 400V charged electrolytic cap in a tube radio with defective resistor. I only felt a sting and it left a black spot on my finger.)
High voltages without much current are harmless. Ask YouTube about "violet ray". These electromagnetic healing devices (as painkiller, skin disinfection, burn warts away etc.) contain a small tesla coil that outputs through a glow bulb lightning bolts at radio frequency with even 50kV(!) those you can touch with bare hands. So stop the hysteria that CRTs are so much more dangeorus because they have "highvoltage" in them. (Did you know that ordinary CFL and fluorescent lamps use thousands of volts also?)
- xray protection
The only thing you really should not do with the high voltage is cranking it up higher than intended by design to increase brightness of a worn out CRT, because this exponentially increases the amount of xray (bremsstrahlung), which goes through walls (also to your neighbour or sleeping childs), can cause cancer and crashes the immune system by preventing sleep due to irritation of the melatonin gland. (The light researcher John Ott warned a lot about this.)
With my monitor the HV adjustment pot is RV901 (HV ADJ).
Quote from the service manual: I worry that even the too bright Trinitron picture may output unnecessary xray, because xray is at least proportional with beam current.
- my RS232 adapter
That some people here fried their serial ports and ICs by plugging in the adapter cable wrongways without prior measuring is simply stupid. At least you should always measure with a multimeter that +Vs and GND is correct. (Confusing TX and RX won't harm. TX means "data out" RX means "data in", so they always need to be connected to the opposite.) GND can be easily identified with a resistance meter or continuity tester even when the monitor is plugged out, because GND is connected with the metal shielding. Also doing a firmware or flash rom upgrade using a "breadboard" and sticking a bunch of loose individual wires somehow into a jack is like transporting gelignite bottles in shoe boxes stacked 1m high upon a wheelbarrow; one little shake and everything goes boom and can't boot any more.
So I soldered a 2 transistor circuit for the RS232/TTL converter like this one:
http://www.uchobby.com/index.php//ttl-to-rs232-adaptor-explained.htm
I added an old diskette drive power cable. For the RS232 cable I did a deep look in my bag of dead mice and dug out an old serial mouse cable (with screw mount). Because pin 4,6 and 8 need to be connected, I pried the plug front out by screw driver (which tore the leads from it) to solder wires on its back. The plug was potted with hard hotglue that I had to pry out and pull the cable through the plug cover to re-solder it. Because these monitor apparently tend to make a fun of switching its +5V output off in the least suitable moment (much like old Thinkpad laptops needed a full battery during BIOS upgrades), I used a 100 Ohm resistor followed by a zener diode against GND to turn the +12V from the PC serial port into +5V. I remember that the 12V output is high ohmed, so for the LEDs I used 1 kOhm instead of 220 Ohm to reduce power consumption.
I haven't used the adapter yet. I have to check if I can install WinDAS on my IBM Thinkpad 760XD laptop (Win98SE,100MB RAM,166MHz) to see WinDAS menus even when the monitor crashes by failed uploads. I hope the monitor won't fall onto me or break my spine when I make the hole to access the ECS port. The thing weights about as much as me and was a hell job to move at its place.
- eastereggs/service mode
In the service manual stands: This all does not sound too spectacular. But I found a few strange additional features.
1. The so-called service mode can also be entered by simply switching the VGA port switch to the unused port and holding '->'. The led starts flashing alternately green and orange and I see the following message:
MODEL:GDM-5411
SER NO:2432556
MANUFACTURED:2001-02
(and 4 colour stripes)
2. When I hold the '<-' button, I get a white upright solid white rectangle on a black background, that looks like a crude screen test for something.
3. When I hold the 'menu' button, the whole screen becomes an orange tinted white surface, which slowly grows dimmer. This may be another variant of the "colour return" function.
The up and down buttons seem to do nothing. To avoid trouble, I didn't dare to mess with 'reset' or 'center' button. The service mode is exited by switching back to the VGA port that has a signal. Unfortunately I found nothing sophisticated hidden like advanced convergence or G2 level settings.
- bright warmup fix found?
I read here about the well known resistor fix on the G2 input of the CRT. A look into the schematics reveals that on the A board (CRT neck) the IC406 "cuttoff amp" is connected to something that they call "drift correct". This strange thing consists of the NPN transistor Q406 (2SC2412K) that has its emitter on the shielding GND and its collector on the GND line of IC406. Between emitter and base is R405 (680 Ohm), and between base and collector R404 (2.2k). I guess that the whole thing works as a temperature sensor that changes the GND level of the video cuttoff amp and this way shifts the output voltage of all 3 RGB electron guns when temperature changes. I can well imagine that the transistor (or its resistors?) wears out over time and so makes the Trinitron go brighter when cold or even in general. So messing around with G2 voltage (by resistor or WinDAS) may be the wrong place to fix the symptom.
The IC406 additionally has 3 capacitors C120, C220, C320 (each 0.1uF 250V) on pins 4 (R_BASE), 6 (G_BASE) and 8 (B_BASE) those seem to set reference voltages and may go bad, but I rather suspect the transistor.
- high frame rates & Trinitron blur
I read that at high horizontal frequencies Trinitron CRTs blur more than shadow mask CRTs. I yet always suspected my antique video cards (Riva TNT2 32MB passed through a 3Dfx Voodoo 1) and my homemade VGA switch box (modified "TV Mouse" CVBS converter with RGB spiral cable on Amiga side), but if other people experience this also, it would be plausible that the Trinitron wire mask has to do with it. I imagine that the thin wires may have higher inductance or simply higher resistance than a sheet metal mask and so become the less "visible" to the electron beam (which interacts by a capacitive pulse) the faster it moves.
It may sound funny for some people here, but I anyway run my SGI GDM-5411 only at 64kHz (1280x1024@60Hz), not least to save the flyback. I grew up with 50Hz frame rate (still daily use PAL on my severely modified Amiga 500) and have no problem with screen flicker (unless flu tube lighting pests the room with a beat frequency of additional flicker, so I solely use incan).
Flyback transformers heat up more and die much faster by high horizontal frequencies. (You can even see the energy waste on a socket power meter.) My first Nokia 417TV burnt out after few years by overclocking (it was only rated for 800 lines modes, but on Linux I successfully used 1024 and more). So my new Nokia 417TV is solely used as videophile PAL TV and hasn't degraded much despite it runs roughly 14h per day. Flatscreens jerk way too much and make me seasick and nothing beats a CRT with fast scrolling classic 2D games.
- add PAL/NTSC modes through WinDAS?
Is it possible to reprogram the SGI GDM-5411 with WinDAS to display true PAL resolution (50Hz interlaced = 25Hz) or NTSC (60Hz interlaced = 30Hz) without a scandoubler, TV card or other digital picture degraders?
It would be nice to play also lightgun or Sega 3D Glasses games on it, which is only possible wirh true unaltered PAL output. I still have the old CVBS converter attached (which had worked with a horrible TOEI monitor, NEC MultiSync 3d and Nokia 417TV), but the SGI shows only "out of range" on black background when I deactivate the Amiga scandoubler. Also people with arcade PCBs would certainly like to have true PAL/NTSC modes. Is this only a firmware problem, or is the hardware of late generation CRT monitors generally incapable to process RGB signals with such low frame rates? May also be that they considered that a too fine beam would look ugly on low resolution (like staring through window blinds), but this would be fixable by widening the vertical focus of the Trinitron tube in low resolution modes (but certainly hard to hack due to some kV voltage).
MAY THE SOFTWARE BE WITH YOU! *============================================================================* I CYBERYOGI Christian Oliver(=CO=) Windler I I (teachmaster of LOGOLOGIE - the first cyberage-religion!) I I ! I *=============================ABANDON=THE=BRUTALITY==========================* {http://weltenschule.de/e_index.html}G2 is independent of HV. If HV were too high then the monitor would simply shut down.
You may be on to something here. I'll check it out these days, unfortunately i haven't had time to mess with it yet. Could you give me a link to the service manual?
That would be horizontal output transistors. (HOTs) While it is true that the flyback transformer itself gets a bit hotter with higher frequencies, it's the horizontal output transistor (the big one which switches B+ into the flyback) which becomes more and more inefficient as frequency rises.
It's not, and this time it really has to do with the flyback. All transformers have a minimum operating frequency depending on core material, size of air gap and number of primary turns. If the frequency is too low, the core saturates. If the core would be allowed to saturate the HOT would blow up.