CurtPalme.com Home Theater Forum

[spacediver]

As someone who has only owned CRTs, I'm quite passionate about them, and I've been trying to learn as much about them as I can.

There are, however, some questions that I've found it hard to get good answers on. I figured this might be a good place to ask!


All questions are in the context of the trinitron electron gun design with aperture grille.

First question:

In order to produce an accurate image, the beams have to be focused on the correct phosphor stripe (see image below)





I'm having trouble understanding how the beams can be correctly aligned with the appropriate phosphor stripe across a range of horizontal resolutions.

Suppose that in the monitor's "prime mode", the pixel clock is perfectly synced with the stripe frequency, so that each new pixel signal is updated at each new triad of phosphor stripes.

Now what happens if you switch to a lower resolution? The structure of the aperture grille remains the same, yet somehow, the picture remains accurate.

With a lower resolution, I'm assuming that the spot size ("beam width) increases, so it must be stimulating multiple stripes. So how is it possible, for example, to display a purely red image in a lower resolution?

Similarly, when you use the geometry controls to center the picture and shift it up/down left/right, how does it do this while retaining accurate stripe targets? Is it designed so that each click of the geometry control shifts the raster by a pre-calculated amount (e.g. some integer of the width of the stripe triad)?

Second question:

My understanding is that the life of an electron gun is ultimately limited by the amount of coated barium or thorium. The cathode heater element produces heat which releases the electrons, through thermionic emission, and creates a "reserve cloud". Then, depending on G1, G2 voltages, etc. those waiting electrons will be pulled in through the tube and strike the phosphors, causing light to be emitted.

Increasing the brightness of the CRT will increase the rate at which electrons are pulled through the tube, and result in a brighter image.

Will increasing the brightness mean more barium/thorium is used up? It would seem as if the main thing that determines barium usage is how much of the material gets released due to thermionic emission, regardless of grid voltages. Do unused electrons "settle" back onto the coating? Or does using up a lot of the reserve electron cloud mean that more electrons are liable to be released due to thermionic emission (kinda like diffusion is dependent upon relative pressure across a membrane)?


Third question:

When calibrating my trinitrons with WinDAS, during the white point balance procedure, it asks you to set the black level by using a graybar pattern with a pedestal. The idea is to adjust the G2 voltage until the pedestal value blends in with the adjacent bar. I like to set it a touch lower, as I find it gives me deeper blacks. Later on in the procedure, it asks you to meet certain luminance targets in the upper IRE ranges. This procedure has a direct impact on the resulting contrast of the display.


Now I've read that different cutoff voltages can put a burden on the CRT video amplifier and the amplifier driving the cathode (not sure if those are the same thing). See page three of this document: http://www.ti.com/lit/an/snla017/snla017.pdf

Do you think that setting the G2 level too low (while maintaining the same luminance at full white) will wear the amplifier due to making it have to swing between too big of an extreme?

I'm not very fluent in electronics, so I'm having trouble interpreting the document. I'm not even sure what Cutoff voltage means, but it seems to have something to do with black level, so I wonder if changing G2 affects Cutoff voltage. Then again, there is a separate part of the white point balance procedure where it asks you to meet a luminance target for the cutoff voltage, so perhaps I needn't worry about setting G2 too low.


Thanks if you've read this far. If anyone has any insights into any of the questions, I'd be really grateful!

[HD-DAVE]

Pretty sure the answer to question 1. is that there is no relation between the resolution of the incoming signal and whether the electron beam strikes the appropriate phosphor properly or not because the beam is *continuous-scanned* across the face of the shadow mask/crt face and is not turned on/off digital style depending on the incoming resolution...think Old School Oscilloscope here; tv's work the same way, except there are three guns and they each are only allowed to strike a fixed vertical (trinitron), or tri-spot, arrangement of coloured phospor area.

[spacediver]

thanks for the reply dave. Yep I get that the beams are continuously scanned, but the pixel clock determines the frequency that the intensity of those three beams are updated.

Take a resolution of 1920x1200 active pixels, at 85 hz (prime mode for an FW900).

The total number of (active) pixels per frame is 2.304 million, which are scanned in 85 times a second. If we ignore blanking intervals, front and back porch, that means that each pixel is scanned in every 5 nanoseconds.

Let's imagine an image that has vertical lines that alternate from red to green with each successive pixel.

Consider what happens when the each line is (horizontally) scanned.

The red gun has to do to things:

1: it has to turn on and off for each successive pixel (every 5 nanoseconds).
2: when it's on, it has to target the correct phosphor stripe.

The green gun has to do the same thing, but out of phase with the red gun by one pixel.

If the timing is even a tiny bit off, then all sorts of artifacts are going to be introduced, and the image will look like a messy noise of color.

So it's remarkable enough that it can do what it does with one resolution.


Now suppose with a lower resolution, the beam width remains the same size, so that the spot profile doesn't overlap too many stripes. The beam still has to update at the right time in the right place, so it has to "skip" some triads as it scans across horizontally.

I suppose it could work. See the image below:



So I suppose my original assumption that spot size increased as resolution decreased was incorrect.

And I guess when you use geometry controls to move image around horizontally, it does so by shifting the raster in units equivalent to the triad width.

Still, bloody remarkable how everything is synced so well.

[spacediver]

Just thought about something else. How does the resize control work? It doesn't affect scanning frequency, so the number of "spots" per horizontal line remains the same. If it simply changed spot size, then the image would just become blurrier or sharper.

[macgyver655]

I'll give you some short quick answers. If you want to read more on how the beam only strikes it's own color points look up "Purity".

As far as different resolution monitors, the actual on screen resolution does not change. The screen resolution in a color direct view tube is fixed. The incoming signal regardless of what it is, is converted to the screens resolution.

The beam does not change in size.

I did not read your entire last post but I think I answered what you were looking for.

[spacediver]

[spacediver]

I think I've answered that last question: I think the wires that define and support the grille physically block the beam from striking the inappropriate phosphors.