Electronics > Beginners

How did CRT TV sweep down 1 row at a time ?

(1/3) > >>

lordvader88:
For the horizontal sweep I'm thinking the horizontal deflection plate voltage is like a ramp generator, with a smooth continuous linear slope, not stepped. The firing of gun is step-wise tho (on/off depending on image).

But what about the vertical. Wouldn't the voltage on the vertical def. plates look like a big staircase.

So how did early TV's make a big staircase like ramp voltage ?

Or is it also a smooth, continuous, linear slope, which should mean the 1st pixel of a row would be higher than the last pixel, even if we don't notice it ????

sokoloff:
I’ve always assumed it was the latter.

And then, you could rotate the whole setup counter-clockwise slightly to remove some (not all) of the skew.

rs20:
Smooth and continuous, depending on your philosophy/point of view. Even if the small skew that would result would be visible to the human eye, which I doubt; it can also be corrected for by putting the opposite tiny skew into the construction of the coils, or subtracting a tiny version of the horizontal plate voltage from the vertical voltage in order to undo the skew (or to put it another way, generate a suitable approximation to the staircase waveform that you describe).

CRT computer monitors have explicit settings which allows you to mix the deflection voltages in any way you want to achieve arbitrary skews and rotations; removing one pixel's worth of skew is a tiny and trivial adjustment comparatively.

BrianHG:
Vertical deflection is usually smooth.
The yoke is just rotated ever so slightly to correct the horizontal lines.  This is usually easy with monitors which only support a small range of scan rates.  The ramp oscillators were usually almost all analog circuits.  Another fix is to tie a high frequency pass filter, IE series cap+resistor, from the H-yoke to the V-yoke to bend the beam a little.

As for advanced very advanced multiscan monitors with a wide frequency range for horizontal and vertical, with all OSD controls, they have 1 or 2 software DSP IC synths, usually both identical, 1 for vert, 1 for horiz, (DONT ASK WHY it was like this, these were patented IC and all my studio grade 21 inch screens had them except for the earliest which had a massive shielded huge DSP/FPGA brick with 2 dacs in it...) they synthesized the horizontal and vertical separately driven to 2 amps to drive the yoke.

basinstreetdesign:

--- Quote from: lordvader88 on July 28, 2018, 12:44:44 am ---But what about the vertical. Wouldn't the voltage on the vertical def. plates look like a big staircase.

So how did early TV's make a big staircase like ramp voltage ?

Or is it also a smooth, continuous, linear slope, which should mean the 1st pixel of a row would be higher than the last pixel, even if we don't notice it ????

--- End quote ---

You are reading current technology into the past when it is just not there.  Analog TV did not have "pixels".  Each "horizontal" line of video was a continuous waveform which represented darker or lighter areas. After demodulation, a higher voltage in the signal would drive the electron gun to emit more electrons to hit the screen.  A lesser voltage would do the opposite.  If a brighter patch of a pictures line was 10 uSec long or 1 uSec long or .12345 uSec long, that's what it was.  There was no transition from one time increment or "pixel" to the next.  The gun was not "fired" in a series of bursts, it was just on with the smoothly varying brightness voltage that was the video signal.

As you say the horizontal sweep voltage applied to the CRT was smooth ramp at 15,734.3 Hz (in North America after 1953) and moved the electron beam smoothly from left to right.  The ramp ended at the right side of the screen and was reset as fast as could be done to move the beam back to the left side of the screen.  The beam was usually blanked blacker-than-black during the reset so that the retrace would not be seen on the screen.

The vertical travel was similarly controlled by a smooth ramp from top to bottom at 59.94 Hz.  Because neither horizontal nor vertical sweep were "stepped" each horizontal line was just slightly slanted downwards as the beam traveled from left to right.  The slant was not seen for two reasons.  First because the camera was doing exactly the same thing at the same time so ITS scanning was just slightly slanted to match.  Secondly, the slant was so small as to be invisible, anyway.

In an attempt to scan the picture at close to 60 Hz (to reduce flicker) without having to actually use the video bandwidth necessary for 60 Hz scanning of all lines, a compromise was used.  That is, that the video frame was created and transmitted as two "fields" of video that made up the whole frame.  Frames were scanned only 30 fps and made up of two fields.  The lines of one field would be interlaced with the lines of the other.  The horizontal scanning of the CRT was arranged so that that would happen in both the camera and the receiver.  The lines on the screen were then counted down from the top of the screen as 1, 263, 2, 264, 3, 265,... down to 262, 525 at the bottom (at least in North Am.).  This arrangement put some very interesting complications into the synchronizing waveforms that were part of the transmitted signal.

I just scanned the Wikipedia entry on NTSC and see that it could be very confusing if you didn't understand the above basics, but good luck if you want to look it up because it's all there.

Navigation

[0] Message Index

[#] Next page

There was an error while thanking
Thanking...
Go to full version
Powered by SMFPacks Advanced Attachments Uploader Mod