Video input mod and vertical poor interlacing

[Alex KL-1]

Hello folks of VideoKarma,
In my experiences adding direct video input to old TV's, I noted that is very common to vertical intelacing resulting poor, resembling a 240p video instead of a 480i. Is more common on simple transistorized TV sets, and is absent on the TV's having "vertical logic" TDA IC's. Over the air I'm never seen this effect. Sometimes occur with my digital TV converter, and is very common when I use a PC with S-Video output (using the pure luminance signal). For supressing this bad interlacing, I need to critically adjust the vertical freq pot on some TV, on verge of desync or when linearity starts to be affected. I'm wondering, what is going there? Someone know about this phenomenon? And possible interventions to mitigate this (without resorting to things like the LM1881)?

If a thead about this subject exists, I'm happy if someone point to me

[old_tv_nut]

Poor interlace results when horizontal sweep waveforms bleed into the vertical sync section. Because of the high currents and voltages in the H sweep, this means that grounds are critical. Many sets have questionable interlace under normal use without added circuits, but adding a baseband video input may have created a ground loop for horizontal sweep currents/voltages.

It may help to do some trial-and-error experimentation of where the added video input circuit is grounded, but there is no guarantee.

[old_tv_nut]

By the way, a PC with S-video output may actually be running 240P to reduce flicker, so don't judge by that.

[Alex KL-1]

Originally Posted by old_tv_nut

Poor interlace results when horizontal sweep waveforms bleed into the vertical sync section. Because of the high currents and voltages in the H sweep, this means that grounds are critical. Many sets have questionable interlace under normal use without added circuits, but adding a baseband video input may have created a ground loop for horizontal sweep currents/voltages.

It may help to do some trial-and-error experimentation of where the added video input circuit is grounded, but there is no guarantee.

Thanks! This makes sense, considering that some of these things have very simple and critical circuits (to reduce costs etc), not always optimized or routed.

I will play with grounding, routing and with sync separator too, in one particular 10" portable I'm restoring/modding that have horrific vertical interlace. Perhaps it will behaves a little better.

And perhaps will better for me to having only moderate expectations when injecting direct video of course.... hehe

[Alex KL-1]

Originally Posted by old_tv_nut

By the way, a PC with S-video output may actually be running 240P to reduce flicker, so don't judge by that.

Yes, so is good to test on various TV to be sure. I discovered this some time ago, in the worst way... ouch

[old_tv_nut]

In 1953, Motorola touted the quality of interlacing in their sets.
https://www.ebay.com/itm/1953-Print-...cAAOSwke9aIDee

[Alex KL-1]

Very interesting!

[Alex KL-1]

Originally Posted by Alex KL-1

Thanks! This makes sense, considering that some of these things have very simple and critical circuits (to reduce costs etc), not always optimized or routed.

I will play with grounding, routing and with sync separator too, in one particular 10" portable I'm restoring/modding that have horrific vertical interlace. Perhaps it will behaves a little better.

And perhaps will better for me to having only moderate expectations when injecting direct video of course.... hehe

I played with this Philco 10" set, first trying to add some shield for the vertical oscillator, since it is located just below the flyback transformer
This not helped much (but I don't removed it), so I decide to play with circuit.
After changing the grounding of vertical sync filter, from ground to +B (due to reference of oscillator input), and changing the input from sync separator from the original video output tap to a exclusive video buffer, then the vertical sync and interlace becomes very good; so good that the range of adjustment to sync become uncritical and I can easily adjust.
To be true, the major difference was when I changing the Vsync filter, so I presume that the ripple induced by the H switching was intruding into the oscillator, upsetting it.

[Alex KL-1]

By the way, I observed the interlace in my TV's and made some ranking of interlace smoothness-:
Using my most picky device for interlacing (my PC):

#My homemade TV with TDA9373 processor (Philco programmed, 2005): the best so far. Zero interlace troubles. But I made it with shields in critical areas, and is a IC with logic and PLL's.

#Sony TV-700U (1966): Impressive. So simple circuit, but very easy to achieve the interlace point, and very stable. Not needed to change the control anymore. But again, Sony put the flyback inside a shield, maybe for not frying the signal Ge transistors. This confirms the H pulse interfering with V sync, in lesser set's without shielding or good layout.

#Toshiba TS-207 (1987): with TDA2578. Good, but become sensitive ater 30 minutes or more, and not interlaces more. ******

#10" Philco portable (1976): Good after changes. Perhaps better than the Toshiba.

#Telefunken TV-231 (1964): originally horrible; good after vertical mods.

#Telefunken TV-472V (1978): discrete vertical stages. Picky for interlace, but relatively stable when one achieve the point. I don't played with it yet.

******NOTE: in fact, probably my PC become lazy with interlace after some time, but the good TV's yet interlace properly.
The worst TV's didn't interlace with any modern device, not only my PC (like original tube Telefunken or original Philco portable).

[old_tv_nut]

Very nice! You have repeated the engineering that a quality manufacturer might have done. But of course, they had to stop trying to improve at some point and go into production with what they had achieved. It's possible they couldn't find much improvement at that point. The less knowledgable ones might not have tried any of this.

[zeno]

With the resolution thing if you feed it composite to a set without a comb
filter your highs will get thrown away. With a comb filter you jump up
to 330 lines & improved other things. IIRC the most lines on
older sets was 450 lines on S-video etc IIRC.
Combs came out apx 1979 from Maggy. Others followed. Alwas on high end
19" & up.
Great to demo. On the combed set you see individual hairs, tiny
freckles & no more barber poling on checkers & lines. Easy step up
on the sales floor.

[Alex KL-1]

Originally Posted by zeno

With the resolution thing if you feed it composite to a set without a comb
filter your highs will get thrown away. With a comb filter you jump up
to 330 lines & improved other things. IIRC the most lines on
older sets was 450 lines on S-video etc IIRC.
Combs came out apx 1979 from Maggy. Others followed. Alwas on high end
19" & up.
Great to demo. On the combed set you see individual hairs, tiny
freckles & no more barber poling on checkers & lines. Easy step up
on the sales floor.

Interesting about the resolution. Last year I got a VGA to composite video converter for my brother, and I tested it on my homemade TDA9373 TV (it have almost 9MHz of video output bandwidth and >7MHz processor bandwidth). Since it don't have comb filter, is almost impossible to read the text letter fonts from PC workspace. For comparision, with S-Video (from same PC I used in interlace tests), fonts are suprisingly clear and clean, being possible to find videos on Internet I want to watch, see notifications, etc, for example.
And, the same composite video modulator plugged on a LCD TV (with a digital comb filter), is possible to read most of text fonts.

I become S-Video addicted in 2005, when I finished my 29" TDA9373 TV, and I enjoyed to play Playstation 2 in S-Video on it.

In my input TV mods with S-Video, I eliminate the sound trap and/or chroma trap (in case of color TV) in the luma circuits (or leave it behind the video input switch), and resolution become good.

For resolution in general, in modded TV's, most times I perceive better sharpness on B&W TV's, due to absence of phosphour divisions, simpler video circuits without delay lines and many stages etc (less sharp highs cutoff).