Grunff wrote:
John wrote:

I am looking to buy a TFT (17") and would like some advice as the 'jargon'
means nothing to me i.e. contrast ratio, cd/m2 etc.etc. What should I look
for in these items? I am only looking at a budget of £350 tops what should
I go for and are there any manufacturers/models to avoid? I was looking at
either a Philips 170S4 or the Liyama 430-B both in black. I would really
like one with a thin outer edge (bezel??), not bothered about integral
speakers.

Obviously I'm biased so bear that in mind, but...

If your budget is £350, *do not* buy a 17" TFT. Or at least
first take a good look at 19" flat CRTs selling at the same
price, or a bit less. You will be gobsmacked by the difference.

A 19" flat CRT screen (like the LG Flatron one for example) is a *lot*
less than £350, I think they're under £200 now.

--
Chris Green )

"Grunff" wrote in message
...
Chris Oates wrote:

and nobody has mentioned 'dead pixels' yet either !

I could live with the odd dead pixel (I have one on my laptop -
it's like a fried to me) if all else was equal. But it's not.

I guess I could possibly live with a dead pixel - but what I wouldn't want
is a pixel which was constantly on. I've seen these any they're infinitely
annoying. A green little dot in the same place on the screen whatever is on
it. Annoyingly retailers apparently have a certain number of pixels which
are allowed to fail yet still be acceptable. If I ever got a TFT display
and had one fixed constantly on, I'd return it. Dead (off) pixel I'd still
try and return it - but may well not kick up a huge fuss if they didn't
allow me to return it.

D

"John Rumm" wrote in message
...
Andy Wade wrote:

whilst graphics were pixel-perfect. I eventually discovered a display
setting in Win-XP (display properties, appearance, effects, "use the
following method to smooth edges of screen fonts") which, when changed
from
"cleartype" to "normal", improved things dramatically. So I'll pass
this on
as a tip: turn ClearType off, it doesn't help.

Clear type uses so called "sub pixel anti aliasing" - the idea being
with a LCD you can control the individual pixels with accuracy, and
hence you can position 1/3rd pixels to give you a smoother edge. It
requires that the monitor has the same ordering of RGB elements in each
pixel that the software is expecting - if you get a monitor with a
different order then you get the colour fringe effect.

Microsoft also have a web-based tool that allows you to configure ClearType
to change how it displays it. Basically you pick the text that is clearest
for you and then Windows will use that setting rather than the default. I'd
really suggest you give it a go and see what you think once its been tuned.

http://www.microsoft.com/typography/...pe/tuner/1.htm

D

On Wed, 17 Dec 2003 11:41:57 -0000, "David Hearn"
wrote:

I guess I could possibly live with a dead pixel - but what I wouldn't want
is a pixel which was constantly on. I've seen these any they're infinitely
annoying. A green little dot in the same place on the screen whatever is on
it. Annoyingly retailers apparently have a certain number of pixels which
are allowed to fail yet still be acceptable. If I ever got a TFT display
and had one fixed constantly on, I'd return it. Dead (off) pixel I'd still
try and return it - but may well not kick up a huge fuss if they didn't
allow me to return it.

The "annoying" retailer near me has a sign up at the checkout pointing out
that TFT screens come in different grades. Only the highest, and most
expensive, grades are free of all pixel defects. I think these are
typically bought for uses like checking X-rays or other medical scans.
Retailers are in the lower-grade market where you may be lucky, may not.

--
Give me patience! RIGHT NOW!

Mail john rather than nospam...

"Juliette" wrote in message
...
In article ,
"nightjar" says...

I started using TFT monitors almost as soon as they became available. I
had
one user who could see flicker on even the highest spec CRT I could buy
at
the time and, at a cost of £2k, the TFT cured the problem. When the
prices
came down to affordable levels*, I converted all my office screens and
my
home machine to flat screen and I wouldn't go back. I like the small
footprint and the fact that I can adjust height, tilt and swivel without
mounting it on a piece of heavy engineering.

We replaced my partner's CRT monitor with a TFT because I could
see the flicker and although he couldn't he was suffering rather
badly with eyestrain and headaches. Like yours, this wasn't a
cheap, nasty CRT - computers are how we make a living so we don't
cut corners on quality.

The Health & Safety Executive publication I checked at the time claimed
that, with CRTs, flicker was considered to be unavoidable for about 5% of
the population.

The TFT did cure the eyestrain and the headaches but the cat was
very annoyed as she used to nap on that monitor.

I would probably have built her a shelf with an electrically heated pad on
it instead.

Colin Bignell

In message ,
writes
geoff wrote:

I use an NEC 18" TFT - it has the same viewable area as a 19" CRT and
cost just over £400. Unfortunately it has a wide enough viewing angle
that my wife can see when I'm looking at what I shouldn't be. It knocks
spots off the 3 19" CTR monitors I have.

... and is over twice the price I should think.

I should think so.
--
geoff

"Chris Oates" none wrote in message
...

Dear Dave
dreadfully sorry but your are going blind
1600x1200 is lurverly (I'm short sighted)

So am I, but......
It's just that bit too far.

I'd go higher but the monitor won't take it.

I've just had a go at the higher resolutions and I'm quite impressed.
Particularly when I changed the icon text size.
The highest my monitor will go is 1600 by 1200. I can go as high as 2048 by
1536, but it falls to pieces and I have to default to something lower :-(

I'll have a better play later in the week, when I have more time,

Dave

On Tue, 16 Dec 2003 22:05:16 -0000, Andy Wade wrote:

Power consumption, they are still remarkably greedy [...]

Not. I measured 42W for a typical screen, 50W flat out white, 2W in
powersave, and 1.5W when "off".

42W is "still remarkably greedy" in my book. This 17" iiyama CRT is
taking around 60W not a great deal of difference.

--
Cheers
Dave. pam is missing e-mail

Dave Liquorice wrote:

42W is "still remarkably greedy" in my book. This 17" iiyama CRT is
taking around 60W not a great deal of difference.

But really, is 40W or thereabouts worth worrying about when your
PC is eating a couple of hundred?

--
Grunff

On Thu, 18 Dec 2003 01:13:49 +0000, Grunff wrote:

42W is "still remarkably greedy" in my book. This 17" iiyama CRT is
taking around 60W not a great deal of difference.

But really, is 40W or thereabouts worth worrying about when your
PC is eating a couple of hundred?

40W for 24/7 is around a unit/day or about =A322/year...

Well the PC is going to be on regardless, now I could possibly
tolerate my percieved short comings of an LCD display if it offered a
decent power saving over a CRT but they don't.

Anyway my PC doesn't take a couple of hundred. It takes around 100
mostly for the CPU, though the video card must take a bit has it runs
at 60C with a fan...

It's not a empty box either, 1GHz Athlon, Ultra 160 SCSI controller,
SCSI CD-RW, SCSI 18G HD, 2 Network Cards, 1 ISDN card, 1 dual port
serial, 1 4xAGP video, 256M of memory (I think). Plus all the normal
on board stuff, floppy, IDE, 2 USB, 2 serial, 1 parallel.

--
Cheers
Dave. pam is missing e-mail

But really, is 40W or thereabouts worth worrying about when
your PC is eating a couple of hundred?

Don't be fooled. The 200W PC power supply rating is just like your consumer
unit being rated at 100A. It doesn't mean it is actually eating it all the
time. Whilst idling, a modern PC uses a tiny fraction of this. Even less if
it turns off the hard disks after a period of inactivity.

Many of the components of a desktop PC are similar to those of a laptop.
That runs for hours off a little battery. OK, the desktop versions are
usually a more power hungry, but not by that much.

Christian.

Christian McArdle wrote:

But really, is 40W or thereabouts worth worrying about when
your PC is eating a couple of hundred?


Don't be fooled. The 200W PC power supply rating is just like your consumer
unit being rated at 100A. It doesn't mean it is actually eating it all the
time. Whilst idling, a modern PC uses a tiny fraction of this. Even less if
it turns off the hard disks after a period of inactivity.

No, really - my PC takes a couple of hundred W. Trust me on this
one. ;-)

--
Grunff

No, really - my PC takes a couple of hundred W. Trust me on
this one. ;-)

What have you got in the thing?

Christian.

Christian McArdle wrote:

What have you got in the thing?

Depends which one you mean, but the hungriest one has a 2.4gig
athlon (about 90W), 2 hard drives, and a GeForce 4. Not
particularly high spec, but easily eats 200W when running.

--
Grunff

Christian McArdle wrote:

Depends which one you mean, but the hungriest one has a 2.4gig
athlon (about 90W), 2 hard drives, and a GeForce 4. Not
particularly high spec, but easily eats 200W when running.


But, we're talking about when idle.

Doesn't spend a great deal of time idle!

--
Grunff

Depends which one you mean, but the hungriest one has a 2.4gig
athlon (about 90W), 2 hard drives, and a GeForce 4. Not
particularly high spec, but easily eats 200W when running.

But, we're talking about when idle.

Christian.

But, we're talking about when idle.

Doesn't spend a great deal of time idle!

Even at 3am? Is it some sort of server? Do you run SETI?

Christian.

Christian McArdle wrote:

Even at 3am? Is it some sort of server? Do you run SETI?

Funnily enough, 3am is backup time, which takes about 2 hours!

But the monitor's off then :-)

No, this one isn't the server - the server is a very modest
600MHz machine which probably does consume a mere 100W or so.

--
Grunff

On Thu, 18 Dec 2003 16:02:43 +0000, Grunff wrote:

Christian McArdle wrote:

What have you got in the thing?

Depends which one you mean, but the hungriest one has a 2.4gig
athlon (about 90W), 2 hard drives, and a GeForce 4. Not
particularly high spec, but easily eats 200W when running.

The Seagate Barracuda 80Gb drive I have in my PC consumes a whole 13W when
active, and your cpu is spec'd at under 70W. Have you upped your bus speed
to 600Mhz or something, or is your graphics card a real hog ? (Still would
expect it to not be using much power in a quiescent mode, though - why have
graphics work no-one can see...)

--
It's not hard to meet expenses, they're everywhere.

Mail john rather than nospam...

John Laird wrote:

The Seagate Barracuda 80Gb drive I have in my PC consumes a whole 13W when
active, and your cpu is spec'd at under 70W. Have you upped your bus speed
to 600Mhz or something, or is your graphics card a real hog ? (Still would
expect it to not be using much power in a quiescent mode, though - why have
graphics work no-one can see...)

The graphics card is ~20-30W IIRC.

But at idle the monitor's powered down anyway, so surely if the
comparison is to determine the significance of the monitor's
consumption, then we need to look at consumotion with the PC
running and outputting graphics?

Say it breaks down like this:
CPU: 70W
Graphics: 30W
Motherboard: 10W
All other cards: 5W
2x hard drives: 30W
Losses in PSU: 30W
(there's a very good reason it has it's own fan!)
2x case fans, cpu fan: 10W

Total: 185W

--
Grunff

The graphics card is ~20-30W IIRC.

But at idle the monitor's powered down anyway, so surely if the
comparison is to determine the significance of the monitor's
consumption, then we need to look at consumotion with the PC
running and outputting graphics?

Say it breaks down like this:
CPU: 70W
Graphics: 30W
Motherboard: 10W
All other cards: 5W
2x hard drives: 30W
Losses in PSU: 30W
(there's a very good reason it has it's own fan!)
2x case fans, cpu fan: 10W

Total: 185W

--
Grunff

This is possibly peak requirements?

The case fans must be pretty hefty to be sucking that much power...
I have an 80mm one in fromt of me, it's only 0.14A (1.68w)

The most reliable way to see would be to actually measure it!

Sparks...

Christian McArdle wrote:

No, really - my PC takes a couple of hundred W. Trust me on
this one. ;-)


What have you got in the thing?



Frying chips?

Seriulsy, POC's don't 'idle' - they are either more or less shut down
with teh clock stopped, or running at teh same curent they always run
at. About 50-200W depending on things like what cards are in them and
how many disc drives are permanently rotating.

CPU typically takes a few amps on its own.


Christian.

Christian McArdle wrote:

Depends which one you mean, but the hungriest one has a 2.4gig
athlon (about 90W), 2 hard drives, and a GeForce 4. Not
particularly high spec, but easily eats 200W when running.


But, we're talking about when idle.


Computers don't idle.



Christian.

Sparks wrote:

The graphics card is ~20-30W IIRC.

But at idle the monitor's powered down anyway, so surely if the
comparison is to determine the significance of the monitor's
consumption, then we need to look at consumotion with the PC
running and outputting graphics?

Say it breaks down like this:
CPU: 70W
Graphics: 30W
Motherboard: 10W
All other cards: 5W
2x hard drives: 30W
Losses in PSU: 30W
(there's a very good reason it has it's own fan!)
2x case fans, cpu fan: 10W

Total: 185W

--
Grunff


This is possibly peak requirements?


There is no difference between a CPU rushing around in a tght loop and
exrecising a full blown compile.

The only thing that opoerartion affects is sometimes disc stepper motors
under high disc access. Otherwise a PC is a constant load device, no
matter what it is doing.

Unless its switched off.



The case fans must be pretty hefty to be sucking that much power...
I have an 80mm one in fromt of me, it's only 0.14A (1.68w)

The most reliable way to see would be to actually measure it!

Sparks...

Computers don't idle.

Yes they do.

Christian. (BEng Electronic and Computer Eng.)

Seriulsy, POC's don't 'idle' - they are either more or less shut down
with teh clock stopped, or running at teh same curent they always run
at.

Not correct. The standard transistor types in a computer consume power when
switching to overcome latent capacitance, but have low quiescent currents
when not doing so. Modern computer equipment is specifically designed to
slow down and avoid switching unnecessarily, meaning that power consumption
drops when nothing much is happening. The power consumption drops markedly
when you have idle CPU cycles.

It isn't like the old TTL logic, where the has to be constant current flow
through a resistor, making the power consumption pretty constant, or related
to state.

Christian.

There is no difference between a CPU rushing around in a tght loop
and exrecising a full blown compile.

There is some difference, as a tight loop won't lead to many state changes.
However, only very old operating systems run the CPU in the tight loop.
Almost all operating systems will put the CPU into idle if there are no
threads demanding execution, giving even greater power consumption drops.

The only thing that opoerartion affects is sometimes disc stepper
motors under high disc access. Otherwise a PC is a constant load
device, no matter what it is doing.

This hasn't been true for about 10 years.

Christian.

Christian McArdle wrote:

There is no difference between a CPU rushing around in a tght loop
and exrecising a full blown compile.


There is some difference, as a tight loop won't lead to many state changes.
However, only very old operating systems run the CPU in the tight loop.
Almost all operating systems will put the CPU into idle if there are no
threads demanding execution, giving even greater power consumption drops.


The only thing that opoerartion affects is sometimes disc stepper
motors under high disc access. Otherwise a PC is a constant load
device, no matter what it is doing.


This hasn't been true for about 10 years.

Well, you mena stepper motors don;t draw current these days?:-)

As far as CPU idling goes, only laptops seem to bother to do anything
when the computer is not actually being tapped upon.

After all, the screen stll need srefreshing, the RAM all needs
refreshing...the sort of 'lets slow down the CPU clock speed and RAM
refresh rate until someone hits a keyboard' kind of thing costs money to
implement...and lots of people myself included simply leave th bloody
things on because parise be to microsnoit, they take too long to boot
anyway...and they poll things like mail servers, and ther are other
background tasks running all the time on a PC.

In short, depsite accepting all your points about switching speeds etc
etc I still maintain that PC's in practice don't Idle.

MA Electrical sciences, Cambridge :-)



Christian.

The Natural Philosopher wrote:

MA Electrical sciences, Cambridge :-)

Don't! You'll summon IMM...

--
Grunff

In short, depsite accepting all your points about switching speeds etc
etc I still maintain that PC's in practice don't Idle.

Do you know about CPU Halt and Grant Idle states? Almost any modern
operating system will enter these states as soon as CPU usage drops below
100%. This isn't some setting in a property page somewhere, it happens
automatically whenever CPU usage isn't 100%.

This isn't an operating system Standby or Suspend state, where it blanks the
screen. It is an instantaneous and effective power reduction mechanism that
is instantaneously reversed as soon as CPU is required.

Christian.

Do you know about CPU Halt and Grant Idle states?

I haven't managed to find any recent figures on AMD/Intel websites, but they
must be on there somewhere.

I've found approximate figures for previous generation processors which
suggest that consumption drops from around 20W to 3W when the processor is
in an idle state. Unfortunately, the source is less than bomb proof.

Christian.

Christian McArdle wrote:
Do you know about CPU Halt and Grant Idle states?

I haven't managed to find any recent figures on AMD/Intel websites,
but they must be on there somewhere.

I've found approximate figures for previous generation processors
which suggest that consumption drops from around 20W to 3W when the
processor is in an idle state. Unfortunately, the source is less than
bomb proof.

There used to be a tool to disable/enable the idle states in some AMD
processors. Temperature sensors built into many motherboards can show that
when the IDLE command was disabled, the temperature of the CPU rose and when
the IDLE command was enabled, the CPU temperature dropped significantly. Of
course, when something was constantly using the CPU (gaming etc) the CPU
temp rose back up and stayed high until the CPU load dropped again.

Just as a pointer, if you're running NT/W2k/XP open up Task Manager, go to
the 'Process' tab and scroll down to the "System Idle Process" process - and
you'll see that generally, its 90% or so idle. As I'm typing this email
I've got loads of IE windows open and OE is checking my email but the Idle
Process is taking about 92-96% of the CPU - and this means that the machine
is only using 4-8% CPU load yet its still displaying the screen and doing
other things.

D

Grunff wrote:

The Natural Philosopher wrote:

MA Electrical sciences, Cambridge :-)


Don't! You'll summon IMM...



Every time.

Christian McArdle wrote:

In short, depsite accepting all your points about switching speeds etc
etc I still maintain that PC's in practice don't Idle.


Do you know about CPU Halt and Grant Idle states? Almost any modern
operating system will enter these states as soon as CPU usage drops below
100%. This isn't some setting in a property page somewhere, it happens
automatically whenever CPU usage isn't 100%.

This isn't an operating system Standby or Suspend state, where it blanks the
screen. It is an instantaneous and effective power reduction mechanism that
is instantaneously reversed as soon as CPU is required.



BUT these states have to be explicitly coded into the BIOS etc. In all
OS's I have designed or been aware of, the CPU at best sits in a tight
loop waiting for some kind of interrupt from somewhere - most frequently
the system timer. Then it probably rushes off to do a few timer related
housekeeping tasks, nips round its scheduler to see if other interrupts
have actually triggered the need to so timeslice code before going back
to doing an endless loop or summat. I agree that a 'halt' instruction
pending interrupt OUGHT to be used somewhere in the core OS..but is it?
Not on my windows 98 I am sure.


Given the size of heatsinks on the CPU, and the fans, its pretty clear
that they are pulling tens of watts no matter WHAT they are doing.
I will grant and exception uin laptop cases. I am sure they do indeed
have sophisticated power management circuitry, but I am pretty sure that
this computer does not.



Christian.

Christian McArdle wrote:
Computers don't idle.

Yes they do.

Especially if they're Linux ones (though I think the latest Windows
OS' have discovered the IDLE instruction as well - or is it HALT?).

--
Chris Green )

The Natural Philosopher wrote:

There is no difference between a CPU rushing around in a tght loop and
exrecising a full blown compile.

But you can HALT the processor when it's idle and it consumes very
little power, Linux kernels do this as do the latest MS ones I
believe.

--
Chris Green )

On 20 Dec 2003 12:33:28 GMT, (Huge) wrote:

The Natural Philosopher writes:

[17 lines snipped]

BUT these states have to be explicitly coded into the BIOS

This has nothing to do with the BIOS. BIOSes are a PC-ism, anyway.

etc. In all
OS's I have designed or been aware of, the CPU at best sits in a tight
loop waiting for some kind of interrupt from somewhere - most frequently
the system timer. Then it probably rushes off to do a few timer related
housekeeping tasks, nips round its scheduler to see if other interrupts
have actually triggered the need to so timeslice code before going back
to doing an endless loop or summat.

The kinds of functionality we are talking about here takes place "below"
the O/S level (not that W/98 is much of an O/S in the first place). (Despite
my earlier facetious remark about O/S idel loops...)

Not one at all AFAICS. A monitor maybe, and not a particularly good
one. A house built on sand? Definitely. OS? Nope. May it rest,
finally, in peace.

..andy

To email, substitute .nospam with .gl

Huge wrote:

The Natural Philosopher writes:

[17 lines snipped]


BUT these states have to be explicitly coded into the BIOS


This has nothing to do with the BIOS. BIOSes are a PC-ism, anyway.


etc. In all
OS's I have designed or been aware of, the CPU at best sits in a tight
loop waiting for some kind of interrupt from somewhere - most frequently
the system timer. Then it probably rushes off to do a few timer related
housekeeping tasks, nips round its scheduler to see if other interrupts
have actually triggered the need to so timeslice code before going back
to doing an endless loop or summat.


The kinds of functionality we are talking about here takes place "below"
the O/S level (not that W/98 is much of an O/S in the first place). (Despite
my earlier facetious remark about O/S idel loops...)



Nothing except internal machine codes takes place below operating system
level.

Its necessary for the OS to call 'halt' or 'idle' or whatever, and wait
for interrupts to restart the CPU.

"The Natural Philosopher" wrote in message
...


MA Electrical sciences, Cambridge :-)



But from Cambridge that's a BA + a wait of a few years after matriculation,
innit?


--
Richard Sampson

email me at
richard at olifant d-ot co do-t uk

RichardS wrote:

"The Natural Philosopher" wrote in message
...


MA Electrical sciences, Cambridge :-)




But from Cambridge that's a BA + a wait of a few years after matriculation,
innit?



Yes, but teh BA course is of MA standard.



--
Richard Sampson

email me at
richard at olifant d-ot co do-t uk