CRT randomly loses power? - Help

This is a discussion on CRT randomly loses power? - Help ; Hello everyone! Over the past several weeks, my monitor has started to blank out at random (and frequent) intervals. Some days, it happens once every hour or so, other days it happens every 30 seconds. The monitor's power LED stays ...

+ Reply to Thread
Page 1 of 2 1 2 LastLast
Results 1 to 20 of 34

Thread: CRT randomly loses power?

  1. CRT randomly loses power?

    Hello everyone!

    Over the past several weeks, my monitor has started to blank out
    at random (and frequent) intervals. Some days, it happens once
    every hour or so, other days it happens every 30 seconds. The
    monitor's power LED stays green, but the CRT itself loses power.
    Note that it doesn't appear to be going into power-save mode, as
    the LED doesn't turn amber... the CRT tube simply loses power.
    Turning the monitor off and then on brings it back to life.

    My (desktop) system:
    - Redhat Linux 9.0, fully updated via Redhat Network
    - Stock kernel 2.4.20-20.9 with glibc 2.3.2
    - Athlon 1800+ running on an ECS Mainboard K7S5A motherboard
    - NVIDIA GeForce 440 MX video card
    - NVIDIA accelerated Linux driver v1.0-4496
    - NEC MultiSync 97F CRT monitor (running at 1280 x 1024)

    Note that this is a dual boot system, and I experience none of these
    problems when running Windows 2000. Another possibly related
    issue is that the right side edge the screen is slightly compressed, as if
    the CRT tube was aimed wrong or a magnet was nearby. No amount
    of adjusting the monitor controls relieves the problem. Again, this
    issue doesn't show up in Windows, was not the case in Linux a month
    ago, and I can't think of any system changes I made that would cause
    this. I'd also add that I had a borrowed LCD monitor a few months
    ago and had no problems with it either under either OS.

    I'm running out of ideas.... can anyone offer some suggestions?

    Many thanks!

    -Brian

  2. Re: CRT randomly loses power?

    In comp.os.linux.setup jonesbr@ecn.purdue.edu wrote:
    > Turning the monitor off and then on brings it back to life.


    It sounds like an hardware problem and not a software one. And it
    looks like is located in the CRT itself. Bring it to a tech or
    just buy a new one.

    Davide

  3. Re: CRT randomly loses power?

    In message , Davide Bianchi
    writes
    >In comp.os.linux.setup jonesbr@ecn.purdue.edu wrote:
    >> Turning the monitor off and then on brings it back to life.

    >
    >It sounds like an hardware problem and not a software one. And it
    >looks like is located in the CRT itself. Bring it to a tech or
    >just buy a new one.
    >

    Agreed. If you can use a soldering iron, or know someone who can, the
    best bet is the pins of the CRT socket. The PCB this socket sits in is
    usually large and heavy, carrying all the video circuitry, and along
    with thermal expansion and contraction, this causes the joints to break.

    Needless to say, be exceptionally careful as the high voltages inside a
    monitor do not go away at the moment of switch-off. Unplug it and
    ideally leave it at least overnight, and still be very careful about
    what you touch.

    Yes, I know CRT monitors are fairly cheap, but that isn't the point. By
    far the best way of recycling anything is to keep on using it. Joint
    failure can occur within a couple of years of manufacture, when the
    monitor really has another five years or so of useful life left.
    --
    Joe

  4. Re: CRT randomly loses power?

    Davide Bianchi wrote in message news:...
    > In comp.os.linux.setup jonesbr@ecn.purdue.edu wrote:
    > > Turning the monitor off and then on brings it back to life.

    >
    > It sounds like an hardware problem and not a software one. And it
    > looks like is located in the CRT itself. Bring it to a tech or
    > just buy a new one.


    I already thought of that, but you'll see in my original post that the described
    behavior does _not_ occur under Windows 2000 (it is a dual boot machine).
    Thus, I'm back to thinking that the problem is software related.


    > Davide


    Thanks for the reply.

    -Brian

  5. Re: CRT randomly loses power?

    On 1 Nov 2003 14:16:23 -0800, jonesbr@ecn.purdue.edu wrote:
    > Davide Bianchi wrote in message
    > news:...
    > >
    > > It sounds like an hardware problem and not a software one. And it
    > > looks like is located in the CRT itself. Bring it to a tech or just
    > > buy a new one.

    >
    > I already thought of that, but you'll see in my original post that
    > the described behavior does _not_ occur under Windows 2000 (it is a
    > dual boot machine). Thus, I'm back to thinking that the problem is
    > software related.


    Not necessarily...

    I have experienced a problem with bad memory that occurred in only Linux
    (KDE crashed on a memory access error) and not Windows. Replaced the
    memory and my Linux problems went away.

    The two OS's utilize h/w differently. It's possible that Linux is using
    different sync / refresh rates (etc) that trigger the h/w error.

    - W. Citoan
    --
    Military secrets are the most fleeting of all.
    -- Spock, "The Enterprise Incident", stardate 5027.4

  6. Re: CRT randomly loses power?

    On Sat, 01 Nov 2003 20:59:20 +0000, Joe wrote:

    > In message , Davide Bianchi
    > writes
    >>In comp.os.linux.setup jonesbr@ecn.purdue.edu wrote:
    >>> Turning the monitor off and then on brings it back to life.

    >>
    >>It sounds like an hardware problem and not a software one. And it
    >>looks like is located in the CRT itself. Bring it to a tech or
    >>just buy a new one.
    >>

    > Agreed. If you can use a soldering iron, or know someone who can, the
    > best bet is the pins of the CRT socket. The PCB this socket sits in is
    > usually large and heavy, carrying all the video circuitry, and along
    > with thermal expansion and contraction, this causes the joints to break.
    >
    > Needless to say, be exceptionally careful as the high voltages inside a
    > monitor do not go away at the moment of switch-off. Unplug it and
    > ideally leave it at least overnight, and still be very careful about
    > what you touch.
    >
    > Yes, I know CRT monitors are fairly cheap, but that isn't the point. By
    > far the best way of recycling anything is to keep on using it. Joint
    > failure can occur within a couple of years of manufacture, when the
    > monitor really has another five years or so of useful life left.


    I'm starting to get one of the colours dropping out every now and again -
    it goes pinkish so I guess it must be the green tube. I'm pretty sure this
    is a fault in or near the plug of the CRT, because it's a new motherboard
    with onboard video, and wiggling the plug and bending the cable seems to
    fix it temporarily.

    I'm thinking that if this goes altogether, I'll try to fix it as follows:

    1. get a suitable 15 pin din plug. (already done this just in case)
    2. cut a short length of the cable sheath off just above the wires.
    3. if the wires are colour coded, go to step 5.
    4. otherwise label them all in some way, putting two labels on each wire
    with a gap between.
    5. Cut all the wires.
    6. Use a circuit tester to work out which wires go to which pins.
    7. Solder up the new plug with the right wires on the right pins. (Not
    forgetting that sneaky gotcha when you remember afterwards that you should
    have slid the plug cover up the cable _before_ you soldered it up.)

    Any comments?

    andy.

    --
    remove 'n-u-l-l' to email me. html mail or attachments will go in the spam
    bin unless notified with [html] or [attachment] in the subject line.

  7. Re: CRT randomly loses power?

    Andy Baxter writes:

    > I'm starting to get one of the colours dropping out every now and again -
    > it goes pinkish so I guess it must be the green tube. I'm pretty sure this
    > is a fault in or near the plug of the CRT, because it's a new motherboard
    > with onboard video, and wiggling the plug and bending the cable seems to
    > fix it temporarily.


    We had a case at my university a while back with a bunch of new nvidia
    cards. After about a year, the picture started getting fuzzy, with
    streaks and shadows, on about half of the cards.

    --
    Måns Rullgård
    mru@kth.se

  8. Re: CRT randomly loses power?

    On Saturday 01 November 2003 14:42 jonesbr@ecn.purdue.edu wrote:

    > Hello everyone!
    >
    > Over the past several weeks, my monitor has started to blank out
    > at random (and frequent) intervals. Some days, it happens once
    > every hour or so, other days it happens every 30 seconds. The
    > monitor's power LED stays green, but the CRT itself loses power.
    > Note that it doesn't appear to be going into power-save mode, as
    > the LED doesn't turn amber... the CRT tube simply loses power.
    > Turning the monitor off and then on brings it back to life.
    >
    > My (desktop) system:
    > - Redhat Linux 9.0, fully updated via Redhat Network
    > - Stock kernel 2.4.20-20.9 with glibc 2.3.2
    > - Athlon 1800+ running on an ECS Mainboard K7S5A motherboard
    > - NVIDIA GeForce 440 MX video card
    > - NVIDIA accelerated Linux driver v1.0-4496
    > - NEC MultiSync 97F CRT monitor (running at 1280 x 1024)
    >
    > Note that this is a dual boot system, and I experience none of these
    > problems when running Windows 2000. Another possibly related
    > issue is that the right side edge the screen is slightly compressed,
    > as if
    > the CRT tube was aimed wrong or a magnet was nearby. No amount
    > of adjusting the monitor controls relieves the problem. Again, this
    > issue doesn't show up in Windows, was not the case in Linux a month
    > ago, and I can't think of any system changes I made that would cause
    > this. I'd also add that I had a borrowed LCD monitor a few months
    > ago and had no problems with it either under either OS.
    >
    > I'm running out of ideas.... can anyone offer some suggestions?
    >

    Run, don't walk, to your XF86Config file, and change this lines left
    hand figure to something the monitor can tolerate, its apparently too
    low. This results in the scan transistor/transformer getting
    dangerously close to the saturation point at the right hand side of the
    screen, which in turn causes it to draw too much currant and get too
    hot.

    Under the "Monitor" section, find this line:
    HorizSync 27-79
    and raise the left hand number to at least 31.

    If that doesn't help, then I'd suspect that the transistor's heat sink
    compound may be dried out and its getting too hot anyway. Remove it if
    you're familiar with hot soldering irons, clean it up and regrease it,
    and re-install it. If hot metal isn't your cup of tea, get somebody
    else, prefereably someone who can put a "CET" after their name like I
    do and have them do it. And check the condition of the electrolytic
    capacitors in that area also.

    > Many thanks!
    >
    > -Brian


    --
    Cheers, Gene
    A mostly retired old coot who is a CET.

  9. Re: CRT randomly loses power?

    > > I already thought of that, but you'll see in my original post that
    > > the described behavior does _not_ occur under Windows 2000 (it is a
    > > dual boot machine). Thus, I'm back to thinking that the problem is
    > > software related.

    >
    > Not necessarily...
    >
    > I have experienced a problem with bad memory that occurred in only Linux
    > (KDE crashed on a memory access error) and not Windows. Replaced the
    > memory and my Linux problems went away.


    This I can easily believe. I guess I wasn't expecting a similar
    problem when
    dealing with my monitor because....


    > The two OS's utilize h/w differently. It's possible that Linux is using
    > different sync / refresh rates (etc) that trigger the h/w error.


    ....I automatically probed the monitor to get the settings. (I forget
    what the
    name of the industry spec that allows this is called at the moment.)
    Then I
    compared the resulting XF86Config settings against the listed specs
    at:

    http://www.necmitsubishi.com/product...9&division=NEC

    The only thing I had to fine tune was the "DisplaySize" option. The
    horizontal and vertical refresh numbers were right on.


    > - W. Citoan


    Thanks for the response... let me know if you have any more thoughts!

    -Brian

  10. Re: CRT randomly loses power?

    Thanks for the response! My comments inline below...

    root wrote in message news:...
    > On Saturday 01 November 2003 14:42 jonesbr@ecn.purdue.edu wrote:
    >
    > > Hello everyone!
    > >
    > > Over the past several weeks, my monitor has started to blank out
    > > at random (and frequent) intervals. Some days, it happens once
    > > every hour or so, other days it happens every 30 seconds. The
    > > monitor's power LED stays green, but the CRT itself loses power.
    > > Note that it doesn't appear to be going into power-save mode, as
    > > the LED doesn't turn amber... the CRT tube simply loses power.
    > > Turning the monitor off and then on brings it back to life.
    > >
    > > My (desktop) system:
    > > - Redhat Linux 9.0, fully updated via Redhat Network
    > > - Stock kernel 2.4.20-20.9 with glibc 2.3.2
    > > - Athlon 1800+ running on an ECS Mainboard K7S5A motherboard
    > > - NVIDIA GeForce 440 MX video card
    > > - NVIDIA accelerated Linux driver v1.0-4496
    > > - NEC MultiSync 97F CRT monitor (running at 1280 x 1024)
    > >
    > > Note that this is a dual boot system, and I experience none of these
    > > problems when running Windows 2000. Another possibly related
    > > issue is that the right side edge the screen is slightly compressed,
    > > as if
    > > the CRT tube was aimed wrong or a magnet was nearby. No amount
    > > of adjusting the monitor controls relieves the problem. Again, this
    > > issue doesn't show up in Windows, was not the case in Linux a month
    > > ago, and I can't think of any system changes I made that would cause
    > > this. I'd also add that I had a borrowed LCD monitor a few months
    > > ago and had no problems with it either under either OS.
    > >
    > > I'm running out of ideas.... can anyone offer some suggestions?
    > >

    > Run, don't walk, to your XF86Config file, and change this lines left
    > hand figure to something the monitor can tolerate, its apparently too
    > low. This results in the scan transistor/transformer getting
    > dangerously close to the saturation point at the right hand side of the
    > screen, which in turn causes it to draw too much currant and get too
    > hot.


    The explanation has a certain logic, but why the right side?


    > Under the "Monitor" section, find this line:
    > HorizSync 27-79
    > and raise the left hand number to at least 31.


    Unfortunately, I already had my minimum HorizSync set to 31, so this
    isn't the final answer for me. For reference, here is the relevant
    portion
    of my XF86Config:

    Section "Device"
    Identifier "Videocard0"
    Driver "nvidia"
    VendorName "Videocard vendor"
    BoardName "NVIDIA GeForce 4 (generic)"
    VideoRam 65536
    Option "NoLogo" "1"
    Option "RenderAccel" "1"
    Option "WindowFlip" "1"
    Option "Overlay" "1"
    Option "CIOverlay" "1"
    Option "UseEdidFreqs" "1"
    Option "UseClipIDs" "1"
    EndSection

    Section "Monitor"
    Identifier "Monitor0"
    VendorName "NEC"
    ModelName "MultiSync 97F"
    DisplaySize 365.8 274.3
    HorizSync 31.0 - 96.0
    VertRefresh 55.0 - 160.0
    Option "DPMS"
    EndSection

    Section "Screen"
    Identifier "Screen0"
    Device "Videocard0"
    Monitor "Monitor0"
    DefaultDepth 24
    SubSection "Display"
    Depth 24
    Modes "1280x1024" "1280x960" "1152x864" "1024x768"
    "800x600" "640x480"
    EndSubSection
    EndSection


    A detailed description of my monitor (with the rate information)
    can be found at:

    http://www.necmitsubishi.com/product...9&division=NEC


    > If that doesn't help, then I'd suspect that the transistor's heat sink
    > compound may be dried out and its getting too hot anyway. Remove it if
    > you're familiar with hot soldering irons, clean it up and regrease it,
    > and re-install it. If hot metal isn't your cup of tea, get somebody
    > else, prefereably someone who can put a "CET" after their name like I
    > do and have them do it. And check the condition of the electrolytic
    > capacitors in that area also.


    I'm always looking for an excuse to tinker, but my skills in this area
    are lacking enough that I don't want to risk permanant damage to
    what is not yet a completely dead monitor.

    Thanks again!

    -Brian

  11. Re: CRT randomly loses power?

    In comp.os.linux.help jonesbr@ecn.purdue.edu wrote:
    > Thanks for the response! My comments inline below...
    > root wrote in message news:...
    > > On Saturday 01 November 2003 14:42 jonesbr@ecn.purdue.edu wrote:
    > > Run, don't walk, to your XF86Config file, and change this lines left
    > > hand figure to something the monitor can tolerate, its apparently too
    > > low. This results in the scan transistor/transformer getting
    > > dangerously close to the saturation point at the right hand side of the
    > > screen, which in turn causes it to draw too much currant and get too
    > > hot.


    > The explanation has a certain logic, but why the right side?


    No logic that I can see. He is saying that going as slow as possible
    (which won't happen since X chooses the fastest config within stated
    limits) comething called a "scan transistor" to get to a "saturation
    point" at the "right hand side".

    I presume he is talking about a driver for the horizontal sweep on the
    beam? If so, I think he is saying that the driver has to pass more
    current to drive the beam further right, and that it can't pass more
    than its maximum (or maybe he is talking about voltage and the
    collector bottoming out above the base .. nah, nobody will drive
    from the collector end, will they? It'll be a darlington follower
    with emitter end following the voltage on the collector of some other
    transistor).

    Why would going slowly increase the voltage or current requirement for a
    given deflection? It wouldn't, on gross physics.

    Maybe he means the beam has to dwell at the right too long? Well,
    adjust your modeline. You can control that precisely. But dwelling too
    long? It'll only make a difference if the fraction of the overall time
    spent there is more! And that is up to you and your modeline.

    Yes, that would make the transistor get hot, but it's the cause, not
    the effect! And getting hot is not good, but will only cause a
    meltdown, and other heat related defects. Nothing repairable.

    > > Under the "Monitor" section, find this line:
    > > HorizSync 27-79
    > > and raise the left hand number to at least 31.


    > Unfortunately, I already had my minimum HorizSync set to 31, so this
    > isn't the final answer for me. For reference, here is the relevant
    > portion


    Forget it. Even if there were something in it, your Modeline would bot
    be exercising the lower limit of the hscan bounds!

    > Option "Overlay" "1"
    > Option "CIOverlay" "1"
    > Option "UseEdidFreqs" "1"
    > Option "UseClipIDs" "1"


    Eh?

    > Section "Monitor"
    > Identifier "Monitor0"
    > VendorName "NEC"
    > ModelName "MultiSync 97F"
    > DisplaySize 365.8 274.3
    > HorizSync 31.0 - 96.0
    > VertRefresh 55.0 - 160.0


    Way too high V limits. Take 'em down to something sensible, like 80.

    > Modes "1280x1024" "1280x960" "1152x864" "1024x768" "800x600" "640x480"



    Peter

  12. Re: CRT randomly loses power?

    ptb@oboe.it.uc3m.es (P.T. Breuer) writes:

    >> Section "Monitor"
    >> Identifier "Monitor0"
    >> VendorName "NEC"
    >> ModelName "MultiSync 97F"
    >> DisplaySize 365.8 274.3
    >> HorizSync 31.0 - 96.0
    >> VertRefresh 55.0 - 160.0

    >
    > Way too high V limits. Take 'em down to something sensible, like 80.


    How is 160 Hz way too high? It's perfectly normal for a monitor to
    support such frequencies. In fact, I have some that do. Only they
    don't support it at high resolutions, since the hsync frequency sets
    the limit there.

    --
    Måns Rullgård
    mru@kth.se

  13. Re: CRT randomly loses power?

    In comp.os.linux.hardware M?ns Rullg?rd wrote:
    > ptb@oboe.it.uc3m.es (P.T. Breuer) writes:


    > >> Section "Monitor"
    > >> Identifier "Monitor0"
    > >> VendorName "NEC"
    > >> ModelName "MultiSync 97F"
    > >> DisplaySize 365.8 274.3
    > >> HorizSync 31.0 - 96.0
    > >> VertRefresh 55.0 - 160.0

    > >
    > > Way too high V limits. Take 'em down to something sensible, like 80.


    > How is 160 Hz way too high?


    Because it's about twice normal. You can't even see flicker over about
    60Hz. By the time you get to 70Hz you've left behind even beat
    frequencies with lighting (50Hz here in europe, 60Hz in the US). VESA
    norms are about 72-75Hz. You can't possibly have any need for any more
    than 80Hz.


    > It's perfectly normal for a monitor to
    > support such frequencies.


    No it isn't. Besides that, a monitor does not "support" a vscan rate
    directly. It's an indirect consequence of the number of lines you
    choose to display on the screen, and the hscan frequency you choose.
    The hscan limit is what counts. The vertical travel rate of the beam is
    relatively slow and is not limiting.

    If you only put 100 dots per line, and the max hscan rate is about
    200KHz, and you only have 10 lines per screen, then you will achieve
    some phenomenal vscan rates! But I just say that to illustrate that the
    vscan figure is not directly related to machine limits.

    What you want to consider is what overall refresh rate on the whole
    picture you are comfortable with. Set the limit at that. Then the
    pixel scan rate will be the slowest compatible with your comfortable
    viewing rate (which in turn also depends on pixel latency, and so on).
    Then the beam will dwell the maximum time on each pixel and be
    "brighter". And you will get sharper images because at lower
    pixel frequencies the inevitable bleedoff of high frequencies in the
    beam will be relatively less.

    > In fact, I have some that do.


    There you are! No, it's not normal. And certainly not normal if human
    beings rather than aliens from outher space are the intended viewers.

    > Only they
    > don't support it at high resolutions, since the hsync frequency sets
    > the limit there.


    It sets it everywhere, being the only mechanical limiting factor.

    Set the vsync at 80, or 85 if you must. 160 is insane,
    counterproductive, and just plain silly.

    Show us the modeline chosen by X for your monitor (do an X -probeonly).


    Peter

  14. Re: CRT randomly loses power?

    On Sunday 02 November 2003 09:40 P.T. Breuer wrote:

    > In comp.os.linux.help jonesbr@ecn.purdue.edu wrote:
    >> Thanks for the response! My comments inline below...
    >> root wrote in message
    >> news:...
    >> > On Saturday 01 November 2003 14:42 jonesbr@ecn.purdue.edu wrote:
    >> > Run, don't walk, to your XF86Config file, and change this lines
    >> > left hand figure to something the monitor can tolerate, its
    >> > apparently too
    >> > low. This results in the scan transistor/transformer getting
    >> > dangerously close to the saturation point at the right hand side of
    >> > the screen, which in turn causes it to draw too much currant and
    >> > get too hot.

    >
    >> The explanation has a certain logic, but why the right side?

    >
    > No logic that I can see. He is saying that going as slow as possible
    > (which won't happen since X chooses the fastest config within stated
    > limits) comething called a "scan transistor" to get to a "saturation
    > point" at the "right hand side".


    Thats what I was also assuming, that X picks the highest rate, so that
    part, which is free for the editing, was the first pass.

    As far as saturation, the limit may not, and probably isn't, in the
    transistor, providing it has adequate drive, which is the reason I said
    to check the electroytic capacitors in the vicinity. Any capacitor
    under 20 uf is subject to deforming and loss of capacitance because its
    also probably a much higher voltage rating than is actually applied, so
    they deform and slowly lose their capacitance while showing no
    externally visible signs of failure.

    But back to the keyword saturation. The current thru the scan
    transistor AND its transformer, once the transistor is turned on at
    about midscreen in the scan from left to right, rises microsecond by
    microsecond according to the inductance, the magnetic inertia if you
    will, of the transformer. The transistor itself is turned on tight and
    represents virtually no loss at this time in the scan. Slower scan
    rates, below those which the monitor is designed to run at, result in
    an excessive amount of time for the currant to rise. The scan
    transformer has a ceramic iron core, called a ferrite core. This core
    material, while an excellent choice at the higher frequencies because
    it has very little eddy currant losses which translate into very low
    levels of heating and high efficiencies, has one other not so desirable
    property. It can only be magnetised to a set maximum amount, often
    haveing a severe negative slope vs temperature as it approaches the
    "curie" point of that particular ferrite compound. Some compounds have
    curie points well below the boiling point of water!

    If the magnetic field within that core hits that limit, then the scan
    transformer, which was controlling the flow of current with its
    inductance, suddenly has virtually zero inductance! So what happens is
    that the currant, which may have been 300 milliamps when the beam is a
    inch from the right end of the scan, can rise very quickly into the
    tens of amperes range and represents a short circuit by the time the
    beam has traveled that last inch. If the transistor can interrupt it
    at the end of the scan, this scene can be repeated, but there are
    storage effects in big power transistors that come into play at very
    high currents, and which may prevent the transistor from doing a good
    turnoff. At that point, the circuits average currant draw is 10x or
    more what it is in normal operation, and parts have but milliseconds to
    live.

    > I presume he is talking about a driver for the horizontal sweep on the
    > beam? If so, I think he is saying that the driver has to pass more
    > current to drive the beam further right, and that it can't pass more
    > than its maximum (or maybe he is talking about voltage and the
    > collector bottoming out above the base .. nah, nobody will drive
    > from the collector end, will they? It'll be a darlington follower
    > with emitter end following the voltage on the collector of some other
    > transistor).


    Generally speaking, darlingtons are too slow for this service.

    > Why would going slowly increase the voltage or current requirement for
    > a given deflection? It wouldn't, on gross physics.


    Then you don't understand the physics of ferrite. All the textbook
    discussions about inductance assume an air core, because thats both
    linear, and predictable. Ferrites let you get the same inductance with
    5% of the copper (or less) in coil windings, thereby reducing the
    overall ohmic losses considerably. But, when they can no longer be
    magnetized any more, their effect on the inductance of the coil simply
    disappears, leading to unlimited current flows.

    > Maybe he means the beam has to dwell at the right too long? Well,
    > adjust your modeline. You can control that precisely. But dwelling too
    > long? It'll only make a difference if the fraction of the overall time
    > spent there is more! And that is up to you and your modeline.
    >
    > Yes, that would make the transistor get hot, but it's the cause, not
    > the effect! And getting hot is not good, but will only cause a
    > meltdown, and other heat related defects. Nothing repairable.


    Its all repairable, if one knows how. Finding the right parts can be a
    cast iron bitch though, and leads to tossing whats borderline junk
    into the bin, where much of it belonged while still sitting on the
    store shelf...

    >> > Under the "Monitor" section, find this line:
    >> > HorizSync 27-79
    >> > and raise the left hand number to at least 31.

    >
    >> Unfortunately, I already had my minimum HorizSync set to 31, so this
    >> isn't the final answer for me. For reference, here is the relevant
    >> portion


    Well, it was something to check anyway. Based on turning over every
    rock

    > Forget it. Even if there were something in it, your Modeline would bot
    > be exercising the lower limit of the hscan bounds!
    >
    >> Option "Overlay" "1"
    >> Option "CIOverlay" "1"
    >> Option "UseEdidFreqs" "1"
    >> Option "UseClipIDs" "1"

    >
    > Eh?
    >
    >> Section "Monitor"
    >> Identifier "Monitor0"
    >> VendorName "NEC"
    >> ModelName "MultiSync 97F"
    >> DisplaySize 365.8 274.3
    >> HorizSync 31.0 - 96.0
    >> VertRefresh 55.0 - 160.0

    >
    > Way too high V limits. Take 'em down to something sensible, like 80.
    >
    >> Modes "1280x1024" "1280x960" "1152x864" "1024x768"
    >> "800x600" "640x480"

    >
    >
    > Peter


    --
    Cheers, Gene
    A mostly retired old coot

  15. Re: CRT randomly loses power?

    "Peter T. Breuer" writes:

    >> > Way too high V limits. Take 'em down to something sensible, like 80.

    >
    >> How is 160 Hz way too high?

    >
    > Because it's about twice normal. You can't even see flicker over
    > about 60Hz. By the time you get to 70Hz you've left behind even
    > beat frequencies with lighting (50Hz here in europe, 60Hz in the
    > US). VESA norms are about 72-75Hz. You can't possibly have any need
    > for any more than 80Hz.


    With my monitor, I can see flicker at 70 Hz with bright colors. At 75
    Hz I can't see it any more.

    >> It's perfectly normal for a monitor to support such frequencies.

    >
    > No it isn't. Besides that, a monitor does not "support" a vscan
    > rate directly. It's an indirect consequence of the number of lines
    > you choose to display on the screen, and the hscan frequency you
    > choose. The hscan limit is what counts. The vertical travel rate
    > of the beam is relatively slow and is not limiting.


    OK, bad choice of words. Still, there are monitors where vrefresh is
    limited to 90 Hz (or whatever), albeit indirectly. Don't forget that
    the maximum pixel rate is also a monitor limitation.

    >> In fact, I have some that do.

    >
    > There you are! No, it's not normal. And certainly not normal if human
    > beings rather than aliens from outher space are the intended viewers.


    Who said I'm not an alien?

    >> Only they don't support it at high resolutions, since the hsync
    >> frequency sets the limit there.

    >
    > It sets it everywhere, being the only mechanical limiting factor.
    >
    > Set the vsync at 80, or 85 if you must. 160 is insane,
    > counterproductive, and just plain silly.


    Perhaps it's silly, but X won't use it either case, being limited by
    hsync range.

    --
    Måns Rullgård
    mru@kth.se

  16. Re: CRT randomly loses power?

    In article ,
    =?iso-8859-1?q?M=E5ns_Rullg=E5rd?= wrote:
    > ptb@oboe.it.uc3m.es (P.T. Breuer) writes:
    >
    > >> Section "Monitor"
    > >> Identifier "Monitor0"
    > >> VendorName "NEC"
    > >> ModelName "MultiSync 97F"
    > >> DisplaySize 365.8 274.3
    > >> HorizSync 31.0 - 96.0
    > >> VertRefresh 55.0 - 160.0

    > >
    > > Way too high V limits. Take 'em down to something sensible, like 80.

    >
    > How is 160 Hz way too high? It's perfectly normal for a monitor to
    > support such frequencies.


    I have 120Hz as an upper bound in mine, but the highest rate is
    320x240@103Hz, and usually I run 1536x1152@61Hz (I don't perceive flicker
    very much).

    > Only they don't support it at high resolutions, since the hsync frequency
    > sets the limit there.


    For higher resolutions, sure, as you said; for my 320x240, the horizontal
    rate is 35KHz.

    --
    -eben ebQenW1@EtaRmpTabYayU.rIr.OcoPm home.tampabay.rr.com/hactar
    "God does not play dice" -- Einstein
    "Not only does God play dice, he sometimes throws
    them where they can't be seen." -- Stephen Hawking

  17. Re: CRT randomly loses power?


    "Peter T. Breuer" wrote in message
    news:4373ob.d6s.ln@163.117.139.117...
    > In comp.os.linux.hardware M?ns Rullg?rd wrote:
    > > ptb@oboe.it.uc3m.es (P.T. Breuer) writes:

    >
    > > >> Section "Monitor"
    > > >> Identifier "Monitor0"
    > > >> VendorName "NEC"
    > > >> ModelName "MultiSync 97F"
    > > >> DisplaySize 365.8 274.3
    > > >> HorizSync 31.0 - 96.0
    > > >> VertRefresh 55.0 - 160.0
    > > >
    > > > Way too high V limits. Take 'em down to something sensible, like 80.

    >
    > > How is 160 Hz way too high?

    >
    > Because it's about twice normal. You can't even see flicker over about
    > 60Hz. By the time you get to 70Hz you've left behind even beat
    > frequencies with lighting (50Hz here in europe, 60Hz in the US). VESA
    > norms are about 72-75Hz. You can't possibly have any need for any more
    > than 80Hz.


    Sigh. New account, new filters, I'm seeing Peter's snarkiness again.

    I've got plenty of users who claim to see the difference between 60 and
    higher, and a few who can tell if it's 75 cycles and higher. The 60-cycle
    frequency for merging successive visual images is a *nominal* frequency.
    Different people have different tolerances for it, and will happily pay
    quite a lot of money and effort to get it higher if they're one of those
    particularly sensitive to it.

    I'd suggest actually taking some courses in neuro-physiology to understand
    why the 60-cycle thing exists, and why televisions and movie films get by
    with so much less.



  18. Re: CRT randomly loses power?

    In comp.os.linux.help Mans Rullgard wrote:
    > "Peter T. Breuer" writes:


    > >> > Way too high V limits. Take 'em down to something sensible, like 80.

    > >
    > >> How is 160 Hz way too high?

    > >
    > > Because it's about twice normal. You can't even see flicker over
    > > about 60Hz. By the time you get to 70Hz you've left behind even
    > > beat frequencies with lighting (50Hz here in europe, 60Hz in the
    > > US). VESA norms are about 72-75Hz. You can't possibly have any need
    > > for any more than 80Hz.


    > With my monitor, I can see flicker at 70 Hz with bright colors. At 75
    > Hz I can't see it any more.


    Fantastic. I congratulate you on your eyesight. Wear dark glasses.
    :-).


    > >> It's perfectly normal for a monitor to support such frequencies.

    > >
    > > No it isn't. Besides that, a monitor does not "support" a vscan
    > > rate directly. It's an indirect consequence of the number of lines
    > > you choose to display on the screen, and the hscan frequency you
    > > choose. The hscan limit is what counts. The vertical travel rate
    > > of the beam is relatively slow and is not limiting.


    > OK, bad choice of words. Still, there are monitors where vrefresh is
    > limited to 90 Hz (or whatever), albeit indirectly. Don't forget that
    > the maximum pixel rate is also a monitor limitation.


    Sure (i.e. "emphatically yes"). But if you are at the max clock, you
    have chosen badly for a modeline. You want to be at the lowest clock
    possible that gives you a refresh rate that you like (as I explained in
    the other post, for reasons of picture brightness and sharpness, not to
    mention stability), and generally speaking "well within monitor limits".

    > >> In fact, I have some that do.

    > >
    > > There you are! No, it's not normal. And certainly not normal if human
    > > beings rather than aliens from outer space are the intended viewers.


    > Who said I'm not an alien?


    Sorry. I didn't mean to imply that you weren't :-/.

    Peter

  19. Re: CRT randomly loses power?

    ptb@oboe.it.uc3m.es (P.T. Breuer) writes:

    >> With my monitor, I can see flicker at 70 Hz with bright colors. At 75
    >> Hz I can't see it any more.

    >
    > Fantastic. I congratulate you on your eyesight. Wear dark glasses.


    Well, the maximum flicker frequency I can see varies with monitors. I
    suppose the decay rate of the phosphor varies. With this particular
    monitor, I can barely notice a flicker at 70 Hz, if there are large
    areas of white or bright colors. Thought I don't see what the dark
    glasses have to do with this.

    --
    Måns Rullgård
    mru@kth.se

  20. Re: CRT randomly loses power?

    ptb@oboe.it.uc3m.es (P.T. Breuer) writes:

    > What is it with you and 60Hz? Even *I* can see flicker at 60Hz - not
    > that it worries me. Over 60Hz I certainly can't distinguish it from
    > general picture instability, and I have used a lot of monitors over a
    > lot of years. At 75Hz I believe myself to be completely happy scanwise,
    > but a bad video card will cause wobble that you will confuse for refresh
    > awareness at whatever frequency you run at, and every video card becomes
    > bad at higher frequencies. So I have learned always to favour a lower
    > pixel clock over a few Hz higher vscan. That's the sum of it.


    There's a reason I use a fairly expensive 3Dlabs card. It's not
    running anywhere near the specified limits of the card.

    > I'm not sure why tvs get by with less. Something with motion sensing,
    > I guess.


    I believe the phosphors used in TVs glow longer than those in computer
    monitors. I'm not sure why they differ. Modern TVs double-scan the
    image, giving a refresh rate of 100 Hz (120 in the US).

    --
    Måns Rullgård
    mru@kth.se

+ Reply to Thread
Page 1 of 2 1 2 LastLast