RCA 6T75 Horizontal Linearity

[dobulee]

MadMan had posted some comments in his admiral restoration about working out some horizontal linearity issues with the horizontal driver trimmer, rather than with the horizontal linearity control.

My KCS47T Chassis has always had 'crowding' at the right hand side of the picture (indeed, the 'half' more or less controlled by the Horiz output Tube). Some HO Tubes were worse than others when I tried tube swapping years ago. Given MadMan's comments, I decided to give it another look and play with the horizontal drive trimmer, which this chassis also has.

IT's worth noting that both SAMs and Breitman DO have some comments about adjusting the drive trimmer as far as possible without crowding the picture, but somehow I had always missed it - wasn't on the radar.

I couldn't find my test pattern rig, so I used some repeating patterns in video game graphics as a benchmark. Tell you what - the difference is substantial!

Ive attached before and 'best adjusted' pictures below. Look at the right hand side of the screen.

The linearity isn't perfect, both edges are still crowded compared to center, but it's much less offensive. At this point the vertical linearity is worse, but no amount of fiddling with vert linearity, height, and recentering the picture at the yoke has ever made much difference there.

In a later post, I do need some help -- my horizontal linearity slug is stuck - I'll post some pics, and see if anybody has ideas.

FWIW, this set does have DC Restore at the synch separator tube, so I"ve posted an image with mostly black background for reference. It looks better in pictures than in real life. There is still some image smearing, but I consider that to be the nature of the beast with this set.

Thanks for the help. It was fun to dig into this chassis again.

[dobulee]

Thanks again to everybody in the forum for helping me get interested in the 6T75 again. This KCS47T chassis has no fewer than 3 alternate schematics floating around, and this set uses the 'latest' RCA build with the improved synch separator and 6SN7 as the vertical oscillator.

I have a couple of issues that I'm hoping some folks have come across and can comment on.

Question 1)In the previous post, I showed some horizontal linearity improvement by adjusting the horizontal drive trimmer. I'd like to also try adjusting the horizontal linearity slug -- but the slug is 100% stuck. I've shown 2 pictures of the linearity coil in the pictures. If I hold the coil in one hand and try to turn the adjusting screw, it won't budge. Has anybody encountered this? Would applying some heat to the coil help? some Tuner cleaner to free things up inside the coil? What works in this case.

Also, how available is the width coil as a replacement part. (sorry I don't have the sams in front of me to get the PN -- i'll post it later!) Is it unobtainable? If it isn't out there anywhere, I'd rather live with it AS IS rather than risk breaking the coil.

Question 2) about Horizontal Synch. The TV's horizontal synch is rock solid. It has the ultracommon synchroguide horizontal AFC circuit. I've adjusted the bottom slug for equal peaks as shown in all the guides. However, sams has an additional adjustment which asks pins C and D of the horizontal oscillator transformer to be shorted prior to adjustment of the top frequency slug. When I short pins C and D, the oscillator stops running. Because of this, I've just made the frequency adjustment without shorting the pins. Has anybody encountered this? I dare not leave it on in that condition since it removes the drive from the HOT. It doesn't seem like a problem, since the lock is good, and it holds horizontal lock even at the extreme CCW rotation, as SAMS says it should. Should I just leave well enough alone?

Question 3) on Vertical Linearity. As shown in the previous pics, the screen is stretched at the top 1/4 or 1/5. After some playing around with the Vert height and linearity pots, It actually looks like the linearity is improved drastically with the linearity pot more centered. However, with the linearity pot centered, no amount of adjustment on height gets the screen to fill vertically (maybe 2/3 at best). The only way I can fill the screen is to stretch it out with the linearity pot. So --I've attached the circuit. It's ultra common, and is the same as the admiral TV currently being discussed, save a few tweaks in values. Honestly, there isn't much else to adjust in that circuit, its just the 6SN7, the height pot and and series resistor R150 (270k). Cap C142 has been replaced. Is this a common experience? Other than checking for a weak 6SN7 (or 6K6 sweep output) and drift on R150, where should I poke around?

Any comments appreciated as always. I have ordered an HDMI to composite converter so that I can use my PC to output some better test patterns.

Schematics and images attached.

[Electronic M]

The Lin coil should be freeable. On the stuck one In my first CTC4 what I did was poor some goof off I to the core, and since the wire was fairly stout I connected a D cell across the coil and let it get warm. Once it got warm it broke free...I then totally removed the slug and cleaned it and the form with goof off so it would be less likely to stick again.

[dobulee]

Well, a little bit of heat, and I was able to move the slug slightly. That combined with the horizontal drive trimmer gives me decent linearity.

I got out my DIY pattern Generator and threw the grid pattern up -- plus that gave me a good chance to put the scope on a few interesting points in the set.

The grid pattern looks pretty decent. The vertical linearity is now worse than the horizontal - stretched slightly at the top. That's a problem for another time.

MadMan's discussions on 'ghost / smearing' on his Admiral, as well as the discussion on DC restoration prompted me to look at a few things.

This RCA chassis provides a convient synch negative point to inject composite video right at the grid of the 1st video amp, the captures below are annotated to say if the capture was made with RF injection, or composite video injection.

1st up: Lookint at the Grid pattern with direct composite video injection into the video amp. The linearity is passable. The slight 'bowtie' shape of the horizontal lines didn't used to be there before I had to install an aftermarket deflection yoke :/ Anyhow. Look at the scope, the bottom trace is composite video input, and the top trace is the crt grid. 2 things to note: (1) the video amp obviously has a fairly sloppy underdamped response. The ghosting seen on the screen is clear as day in the CRT grid waveform. Each 'bar' on the composite video is followed by several cycles of underdamped oscillation on the grid. Maybe RCA made this as a compromise to try to squeeze some extra bandwidth out of this 12au7 video amp (I don't have enough TVs to know if low mu triodes were often used as video amps or not...I'd guess not). (2) The bandwidth of the video amp itself is pretty poor (remember the IF stage is out of the mix). Note that the colorburst on the composite video is completely stripped off of the grid waveform. At least this lets me know that I shouldn't expect miracles by tuning up the IF section. The video bandwidth in this chassis is at best a paultry 3MHz.

2nd up. DC Restoration. This is a grid driven set, with DC restoration happening as part of the sync clipper. In fact, a positive going triode grid functions as the DC restoration diode in this configuration. So how good is the DC restoration? Not very. I have taken several captures, an all black screen, an all white screen, and then a center white bar. In all captures the scope settings are the same... the reference and gain are identical.

All Black Composite input to vid amp: The synch tip is at the center scope division. The black level is about one division up (22V). The screen appears gray at best, with retrace lines visible.

All White Composite input to vid amp: The synch tip is now maybe 10V below the previous level, but due fairly high impedance in the diode/cap restoring path, the black level is now only a few volts above the synch tip. The black level has drifted down to ~2V. So, even with DC restoration by the 'black' level at the CRT grid drifts almost 20V under the most extreme cases.

White Bar Composite input. to see DC restoration in action (imperfect though it may be) look at the center white bar. The scope settings are the same, except now were triggered on the vertical, not the horizontal. Note how uniform and flat the 'black' and 'white' sections of the composite input are, while the same portions of the grid drive 'angle' in as the circuit drifts toward the new equilibrium. So, even with DC restoration shown in the schematic, the action leaves a little to be desired.

Finally, I"ll leave you with the response to an NTSC color bar pattern, first going through the RF/IF stage, and then with direct composite injection. With RF injection the color carrier is entirely stripped off the signal when it hits the CRT.

With direct composite injection, the response of the grid is sharper, but the video amp still fails to pass the majority of the color carrier. However, enough 3.57MHz color info remains to give a faint appearance of marching ants on the CRT.

Anyhow. Just a few fun waveforms. I'm not sure how much better this KCS47 chassis can look. Might be polishing a turd, as the expression goes!

Were triode (vs pentode) video amps common? When did they get phased out?

[Spinning Head]

Thanks for the interesting analysis. I have the same set, un-restored. I guess it's a bit of an under performer.