CRT monitor query: Relationship between video input bandwidth and refresh rate/resolution settings possible...

--------------------------------------------------------------------------------------------

Here's a relatively recent link (Jan 2006) that lists CRT models you'll probably still be able to find:

Anyway:

X_pixels * Y_pixels * Vertical_refresh * Nth_harmonic = Bandwidth (asfair)

One person mentioned one should have a monitor that manages 3rd harmonic to get that "crisp" video. While you will get a picture at 1st harmonic ;)

So for a 1600x1200 pixel @ 60 Hz: 1600*1200*60*3 = 345600000 Hz = 345.6 MHz

;)

Minor addition to the above...

There should be a a factor of 1/2 in there somewhere because the max frequency occurs with alternate black and white pixels. Draw the square wave and you can see the base frequency is half the pixel rate.

Whoops.

Being an electronics layperson I'm not sure what, "manages 3rd harmonic", refers to. Would this have something to do with the role of a "3-comb filter"?

Terrific! Thank you.

Ken

sounds right to me... basically if your source emits pixels at greater than twice the bandwidth the display will not respond fast enough to distinguish individual pixels. that's the "ceiling" thing start blurring out before that point, somewhere above half the bandwidth the effect begins to become noticable.

no. the scan and video signals don't intterract inside the monitor (until they both meet at the CRT)

Define realisitic. tubes could be made with bandwiths in the low gigahertz without too much trouble. if they aren't already like that.

ah... TVs do that too...

but at 120Hz wouldn't 60Hz interferance deflect alternate scans in the opposite direction?

the video bandwidth is proportional to the maximum number of pixels times the refresh rate.

you will be able to use the product scanlines * refreshrate to approximate the horizontal rate (add about 20 to scanlines to account for the vertical retrace period)

eg: if aa display can do 1200 lines at 120Hz refresh rate it can do

Bye. Jasen

No. If you want nice sharp edges on your pixels you need to feed tham a square wave not a sine wave. To make a square wave you need a mixture of more than one sine wave. To make a square wave with infinitly steep edges would need to use an infinite number of odd harmonics (multiples of

Yes, but there's a practical consideration as well. If you operate a CRT monitor at or even near its bandwidth limit, image quality will suffer terribly. Quite a bit of headroom is required to avoid this issue.

And to clarify for Ken, the self-confessed "electronics layperson", a periodic wave is composed of a fundamental frequency, which is the frequency we associate with the wave, and various integer multiples of that frequency. These are called harmonics or overtones.

The first harmonic is the base frequency. The second harmonic is twice that frequency, the third three times that, and so on. Other important factors are the amplitudes and phases of these harmonics, which determine the exact wave shape.

Computing all the phases and amplitudes will be left as an exercise for the reader ;-)

The first overtone is the same as the second harmonic, and so on. The fundamental would be the zeroth overtone, but it's not conventional to use that term.

Gino

Some would argue that what you refer to as the first harmonic is the fundamental, with the first harmonic being twice that of the fundamental.

Tom

Thanks for your clarification on this term, "harmonics". It occurs to me that I might do myself a favor by spending some time studying up a bit on all what's involved in CRT monitor technology... all the terms and concepts, etc...

Ken

All I can tell you is that 120 Hz was suggested to me, and then having tried it with a CRT monitor in my house, it seemed to resolve the interference problem pretty much just like the 60 Hz setting does (except minus the bothersome "60 Hz" flicker ). But I only tried it briefly (being that I had to downgrade the resolution setting unacceptably for the CRT monitor I have to refresh at 120 Hz) and thus maybe not thoroughly enough (e.g. during the usual time of day when the powerline interference is at it's worst -e.g. between 9-11 PM). It didn't occur to me that there might be any theoretical reason not to expect this to either work fully or to not work at all. Now that you bring this up, I think I'll test it again; only this time I'll take greater care to be sure to rule out any remote possibility that the PL interference doesn't just happen to be at one of its temporary ebbs at that particular moment.

Thanks,

Ken

than

distinguish

that

Some clarification is in order here, I think. Besides that, it's fun to write about a display technology that I haven't had to worry about professionally for nearly 10 years...:-)

CRTs, and analog video in general, don't know a thing about "pixels." There are no distinct pixels in either the video signal or in the displayed image. Bandwidth, or better yet rise/fall time, is all there is.

You would at first glance think that "bandwidth" IS in fact going to be the relevant specification here, but that turns out not to be the case - nor is it necessary that the specified or measured BW of the video amplifier be high enough to capture the 3rd harmonic of half the pixel rate (since HALF the pixel rate, as mentioned before, is the fastest rate at which you would see a supposed "square wave" representing alternative black/white lines). Due to a number of factors, not the least of which are the spot size of the CRT and the additional "sampling" effect of the color CRT phosphor dot and shadow mask structure, increases in BW beyond a point much higher than about the pixel rate itself (as a rough rule of thumb) will not result in a visible improvement in the image quality. Further, a "bandwidth" spec is often very misleading - depending on how it is measured, "bandwidth" measurements may not take into account slew-rate limiting at the CRT cathode, which is really what you want to worry about when it comes to looking at the quality of alternating vertical lines and similar single-pixel detail.

The bottom line is that in a CRT, you're basically aiming for rise and fall times (at the full video swing) of perhaps 1/3 to

They're not, nor can they readily be made anywhere near that high. With typical cathode drive, you're talking about swinging a signal over the low tens of volts range (20-40V signal swings at the cathode are common) and fairly capacitive load. Gigahertz bandwidth here would be expensive to achieve, but fortunately was never really necessary.

above

than

In terms of the horizontal rate, you're correct - but in the vast majority of CRT monitor designs, the VERTICAL rate will be limited such that you can't ever get to much about 120 Hz, no matter what the horizontal range might imply.

Bob M.

On 4/16/2006, Tom MacIntyre posted this:

They would not be physicists, then.

See my next paragraph about overtones.

Gino (again)

You know, I might have been more help if I hadn't left out some vital information...

An important other thing about this expression of periodic waves as a series of harmonics is that a representation of a given periodic wave as a limited series of harmonics will be an imperfect representation of the original signal. The accuracy of that representation will increase as more harmonics are included. In additon, the harmonics that are present are subjct to phase errors and amplitude errors, all of which change the shape of the signal.

Anyway, the above considerations lead us to ideas of what's good enough, which relates directly to the criterion that we should have at least the third harmonic present in the signal amplifier.

Gino

I've now been able to re-test this. What I've found upon closer inspection is that the 120 Hz setting does indeed improve the PL interference, but not really quite as entirely as the 60 Hz setting does (i.e. there is still some, albeit barely noticeable, "jitter"). What I didn't expect however, (and this turns my "multiples-of-60Hz-hypothesis" on it head) is that when I put the refresh rate even higher than 120 Hz (e.g. to 144 Hz or 150 Hz), the interference problem _remains_ improved as it is at the 120 Hz setting. Obviously the reduction in noticeable interference experienced at 120 Hz has nothing to do with that refresh rate being a whole-number multiple of the 60 Hz, and everything to do with it just being a substantially higher refresh rate. Thus it appears I was wrong in my conclusion that to view high resolution stereo 3D video I needed a CRT monitor capable of displaying those resolutions at a full 180 Hz. This could make my search to buy a suitable CRT monitor for this purpose a tiny bit easier.

Ken

Ken, a thought has been nagging at me for a day or so.

If the problem with your monitor is truly due to the nearby power lines, could it possibly be improved by reorienting the monitor?

My though is that the monitor might be oriented parallel to the oscillating magnetic files, and rotating the monitor so it's perpendicular to those fields may actually help.

BTW, feel free to swap parallel and perpendicular in the above sentence, since I have no idea what exactly causes the physical effects. also consider in-between angles.

In my office layout, there is no way that I could change the orientation of my monitor. If you have the same problem, I offer sympathy :-)

The other possibility is to put up a screen of mu-metal, a magnetic shielding material. You could probably do it for $1000, maybe even less

- it's not cheap stuff. Just wasting bandwidth on this idea :-)

HTH, Gino

On 4/18/2006, Ken Moiarty posted this:

On 4/18/2006, Gene E. Bloch posted this:

Interesting concept, magnetic files. Of course I meant magnetic fields.

Gino

Well it was a worthy try. I just gave your suggestion a try, but to no avail (at least to no noticeable avail). Seems the only CRT related option here is to obtain a monitor capable of, both, at least 1024 x 768 (although my preference would be 1280 x1024) resolution, @ 120 Hz (give or take) refresh rate, simultaneously. So far it appears that monitor models on the market that match (or exceed) this capability are far and few between and therefore not likely to be found in a local vendor selling CRT monitors at good-value prices.

Well the mu-metal option is not all that outrageous really. It's just that (1) it would be strictly utilitarian (i.e not cool to look at) and (2) for about the same investment I can buy a higher performance CRT monitor (through one of a handful of online vendors I've just found) that will provide the high refresh rate that will also do the trick here, whilst enjoy the luxury of the additional features and other (if you will) 'thoughtful touches' that come with a CRT monitor in that price range. For example: the 22" (20" viewable) Iiyama HM204DT, which sports a video input bandwidth of 390 MHz and will handily display 1280 x 1024 at a solid 133 Hz... Its MSRP is $699. And from what I can discern it is (to be brief) 'loaded'! Actually the more I think about it the more I'm inclined to just submit my credit card number and buy it right now. But such impulse buying is a weakness of mine that I must try to hold myself back from. So I'm going to wait until my 'infatuation' with the idea wears off a little so I can possibly weigh the decision more rationally and objectively.

Ken

No problem. I'm not that easily confused.

Ken