Maker Pro
Maker Pro

Piddling around with more LEDs

J

JosephKK

Jan 1, 1970
0
Duh! Ya think? The question is *WHY?*

Sorry, i am not all that deep into the neurophysics of vision, i do
electrical engineering primarily.
They are (when I can see the beat). That's the whole point. If the
monitors were driven by separate cards I could understand it.
I would have expected the driver to normally synchronize them rather
than maximize each refresh rate. Maybe something is funny about the
driver configuration. Maybe a rabid gamer set it up?
 
K

krw

Jan 1, 1970
0
Sorry, i am not all that deep into the neurophysics of vision, i do
electrical engineering primarily.

I'm not getting through. It's one card, one system, two monitors, set
up identically. Where is the beat coming from?
I would have expected the driver to normally synchronize them rather
than maximize each refresh rate. Maybe something is funny about the
driver configuration. Maybe a rabid gamer set it up?

Not possible. It's my work system (I went to LCD long ago). There is
no setting other than 60, 75, or 85Hz (at least at 1280x1024).
 
J

Jasen Betts

Jan 1, 1970
0
I'm not getting through. It's one card, one system, two monitors, set
up identically. Where is the beat coming from?

That's really weird.

I once had two monitors connected to different PCs running at almost
the same rate (the difference was about 10hz in the horizintal rate, so
about 0.01hz in the vertical rate) and the beating was clearly visible
as a vertical discontinuity (vertical retrace in the other monitor) that
slowly migrated up the screen of one of the monitors

I solved the problem by rearranging my desktop to have a pc case
between the two monitors.
Not possible. It's my work system (I went to LCD long ago). There is
no setting other than 60, 75, or 85Hz (at least at 1280x1024).

ultimately it comes down to the rate of the pixel clock in the crtc
if they're both phase-locked to the same source (logic suggests that
this would be the case) then it should be possible to phase-lock the
horizontal, and vertical scans also by using the same base pixel
clock, horizontal total and vertical total settings.

doing that requires that the maker of the video driver exposes the
CRTC registers to the OS and that you get in behind the scenes and
tweak them. (In windows this means special software or editing the
registry, in linux it's xvidtune and adding modelines to the /etc/x11/xorg.conf file)

not easly, may or may not be possible.

if it were me I'd (after investigating and rejecting a software based fix)
get a 600mmx600mm piece of .4mm steel sheet (eg pc case side panel)
and wedge that between the monitors and see if it helps, if it only
helps a little I'd try two, mu-metal, or thicker.

seing as it's at work I'd complain to my immediate boss, IT, and or HR,
that's it's causing "stress"
 
G

Greegor

Jan 1, 1970
0
G >Why did he commit to 1000 before seeing a sample?

KRW > No, I don't ask for engineering
KRW > samples when I buy strings of
KRW > Christmas lights from Home Depot.

$150 smackers worth? 1000 lights?
Did you see them in the store?
Usually you can plug them in before purchase.
Even Wal Mart lets you do that.
1000 is a fairly big string of Christmas lights! LOL
 
K

krw

Jan 1, 1970
0
That's really weird.

I once had two monitors connected to different PCs running at almost
the same rate (the difference was about 10hz in the horizintal rate, so
about 0.01hz in the vertical rate) and the beating was clearly visible
as a vertical discontinuity (vertical retrace in the other monitor) that
slowly migrated up the screen of one of the monitors

That's exactly what I'm seeing, except on one system, one card, same
setup.
I solved the problem by rearranging my desktop to have a pc case
between the two monitors.

That would be too far away to make the monitors useful. The *real*
solution is to replace the (at least) decade old monitors with LCD
monitors. I'm workign on it. ;-)
ultimately it comes down to the rate of the pixel clock in the crtc
if they're both phase-locked to the same source (logic suggests that
this would be the case) then it should be possible to phase-lock the
horizontal, and vertical scans also by using the same base pixel
clock, horizontal total and vertical total settings.

doing that requires that the maker of the video driver exposes the
CRTC registers to the OS and that you get in behind the scenes and
tweak them. (In windows this means special software or editing the
registry, in linux it's xvidtune and adding modelines to the /etc/x11/xorg.conf file)

not easly, may or may not be possible.

....or just set them to very different frequencies. ...though if I
look I can still see the beat.
if it were me I'd (after investigating and rejecting a software based fix)
get a 600mmx600mm piece of .4mm steel sheet (eg pc case side panel)
and wedge that between the monitors and see if it helps, if it only
helps a little I'd try two, mu-metal, or thicker.

Steel isn't likely to do anything, though mu-metal might. I don't
have a handy source for that much mu-metal though.
seing as it's at work I'd complain to my immediate boss, IT, and or HR,
that's it's causing "stress"

I have. I shame the boss every time he comes into my cube. ;-) The
whole system is junk but this is a very bad time to spend any money.
 
K

krw

Jan 1, 1970
0
G >Why did he commit to 1000 before seeing a sample?

KRW > No, I don't ask for engineering
KRW > samples when I buy strings of
KRW > Christmas lights from Home Depot.

$150 smackers worth? 1000 lights?

5 reels @ $36 each.
Did you see them in the store?

On the reel.
Usually you can plug them in before purchase.
Even Wal Mart lets you do that.
1000 is a fairly big string of Christmas lights! LOL

Five reels, 200 per.
 
D

Don Klipstein

Jan 1, 1970
0
How about a couple pieces of steel with foam board between them? It
would still be under a half inch thick. I've had decent results with a
single layer of steel in a wood rack with four monitors, but you could
put several layers between monitors. Also, the electromagnetic field is
near the back of the case, so the steel doesn't have to reach the front
of the monitor case to be effective. You just need to keep the yokes
from interacting so it should run from the rear to around 3/4 of the
depth Broadcast monitors are usually in steel cases and no mu-metal.
They are stacked in all sorts of combinations with no problems

<http://www.officemax.com/omax/catal...Poster_Boards&ci_src=14110944&ci_sku=11001275>

for an example, but I have bought 20" * 30" sheets it at 'Dollar Tree'
for a dollar a sheet.

The steel I used was from scrapped equipment cases.

As for mu-metal - many oscilloscopes have sleeves of that around their
CRTs. My father asks me to let him have those whenever I find occaision
to dump an O-scope.

One thing to watch for - bending mu-metal (or similar ones) can change
their permeability at low Gauss, and will have a high rate of changes
being downward. After bending, mu-metal needs to be annealed (heated past
the Curie temperature and cooled at some slow rate afterwards).

Meanwhile, if weight is not a concern and the metal sheet is allowed to
be flat, I suspect 1/4 inch plate steel is not that much less effective
than a usual thickness of mu-metal and probably cheaper. Same for a few
layers of sheet steel - and that can even be bent into shapes.

- Don Klipstein ([email protected])
 
K

krw

Jan 1, 1970
0
How about a couple pieces of steel with foam board between them? It
would still be under a half inch thick. I've had decent results with a
single layer of steel in a wood rack with four monitors, but you could
put several layers between monitors. Also, the electromagnetic field is
near the back of the case, so the steel doesn't have to reach the front
of the monitor case to be effective. You just need to keep the yokes
from interacting so it should run from the rear to around 3/4 of the
depth Broadcast monitors are usually in steel cases and no mu-metal.
They are stacked in all sorts of combinations with no problems

<http://www.officemax.com/omax/catal...Poster_Boards&ci_src=14110944&ci_sku=11001275>

for an example, but I have bought 20" * 30" sheets it at 'Dollar Tree'
for a dollar a sheet.

The steel I used was from scrapped equipment cases.

I might try that. We have some steel laying around, though most of it
is parts for our cases. ;-) Hmm, if I make the lashup ugly enough
I'll get the boss' attention, again. ;-)
 
K

krw

Jan 1, 1970
0
Did you ask them to light them up for you to see?

Nope. The HomeDespot here is pretty lame. Lowes is much better but
they had no LED strings this year. Perhaps I now know why. I just
bought another 10 strings of incandescent lights for a buck-a-string,
though I coudn't find any clear.
 
K

krw

Jan 1, 1970
0
I usually see them at Dollar Tree, but not this year. A few years
ago I bought a pile of sets of 50 white bulbs for 50 cents each.

Yeah, these are strings of 100. The latest ten strings were a buck
each. I bought twenty strings right before Christmas for $1.50/100.
They had red, green, blue, and mixed but no clear. Next year I'll do
blue.
 
K

krw

Jan 1, 1970
0
If you 'really' want his attention, cut up the cover from his
computer. Unless his computer is an 'Etch-A-Sketch'. ;-)

Yeah, but I don't think I want *that* kind of attention. ;-)
 
K

krw

Jan 1, 1970
0
Yes, that is weird. Even i would expect one card to get the vertical
rates the same.
Would this happen to be a linux system (?), you may be able to get to
the dot clock counts if it is.

No, it's my work system, so it's WinBlows. I might try a Linux CD,
if I get a few minutes.
From your own text:


You have set them to different vertical refresh rates is what you
said.

Yes, if they're set the *same* I see the interference bars walking
through the display. I had them set differently so the beat was
too fast to see. I had them set differently so I didn't see the
bar.
It had thought you said that setting them to different
frequencies caused the problem, rather than setting them to the same
frequency.

I infer from this that the two display channels have different
oscillators. Dumb design, but guess that's what I should expect
from Dell.
The resolutions are the same it appears, are the sizes different?

Both 17", though I don't see how the monitor matters.
 
J

JosephKK

Jan 1, 1970
0
No, it's my work system, so it's WinBlows. I might try a Linux CD,
if I get a few minutes.


Yes, if they're set the *same* I see the interference bars walking
through the display. I had them set differently so the beat was
too fast to see. I had them set differently so I didn't see the
bar.


I infer from this that the two display channels have different
oscillators. Dumb design, but guess that's what I should expect
from Dell.


Both 17", though I don't see how the monitor matters.

I wondered about the possibility of different sizes thus different
spatial resolutions (different coverage areas on the retina crashing
the stereo merging or something).
 
K

krw

Jan 1, 1970
0
I wondered about the possibility of different sizes thus different
spatial resolutions (different coverage areas on the retina crashing
the stereo merging or something).
I don't see (NPI) how different sizes would cause the beat to move
through one of the monitors.
 
J

Jasen Betts

Jan 1, 1970
0
I don't see (NPI) how different sizes would cause the beat to move
through one of the monitors.

1024 lines on one 1020 on the other same hozintal rate.
 
K

krw

Jan 1, 1970
0
1024 lines on one 1020 on the other same hozintal rate.

That's not the monitor's decision. I've already said that both are
set to 1280x1024.
 
J

JosephKK

Jan 1, 1970
0
That's not the monitor's decision. I've already said that both are
set to 1280x1024.

At this point, i suspect your only hope is complaining to the video
card vendor. In m$winders display card drivers there pretty much is
no way to get to the clock count data.
 
Top