Image warping on Ikegami pro monitor despite LM1881

[SamIAm]

The subject line basically says everything.

I recently picked up a very nice Ikegami HTM-2050R2, which was manufactured after 2000 and has fairly modern electronics. I've tried, in RGB, a Mega Drive, a Saturn, a Dreamcast, a Playstation 2, a Super Famicom, an MSX, and a modded PCE, and only the PCE is having trouble. The top 1/4 or so of the picture is badly warped, like a fun-house mirror.

The mod uses composite video for sync, which works fine on my other TVs, but is known to cause problems in general. I tried putting in an LM1881 sync-stripper with this Ikegami, however, and it didn't help. I also noticed that the PS2 apparently uses composite video for sync when in 240p mode, and it doesn't have any trouble.

So, what could be the problem?

Is there something unusual about the PCE's vertical blanking signal? I don't suppose tapping and amping the c-sync from the console directly would help? One thing I'm thinking I can experiment with is the resistor on the LM1881. I'll need to buy a 1 mega-ohm potentiometer to do that right.

Any other ideas, though? 

EDIT:

A few other things.

-I've tried putting resistors in-series with the video signal, as well as a 220uf capacitor. No change.
-The lines all have 75 ohm terminators on them, so grounding them with resistors shouldn't matter.
-If I remove sync completely, the image rolls but is not garbled. At this time, I can see that the warping is gone.

[deubeul]

Not sure if it will help, but it reminds me of what does a PcE on a XRGB3. Setting UP the value of the Low Pass Filter was the solution.

Maybe you could post your issue on the http://shmups.system11.org/ forums, it's full of helpfull tech guys. 

[SamIAm]

That's very interesting. The datasheet for the LM1881 says that some signals will need a lowpass filter to clean them, and it gives a recommendation on how to build one (a simple combination of a resistor and a capacitor).

I posted on that forum. It's been suggested that you are misremembering, and that changing the AFC setting is what fixed your problem. What do you think?

[deubeul]

I think I clearly misremembered, It was the AFC setting .  :oops:

Second time in one week I say bullshit while trying to help people 

I really hope you'll find a solution.

[SamIAm]

No problem. I very much appreciate your effort to help.

This has turned out to be quite interesting. A kind and knowledgeable fellow by the name of viletim took a look at the PCE's composite video signal with his oscilloscope, along with the NES's. It seems that the likely explanation for the warping is that the PCE's vertical sync timing is a little off, and this causes the first few pulses of the horizontal sync to be in the wrong place. In fact, it appears that one is even missing.

I contacted the Japanese guy who did my RGB mods, and in his investigation he found a blog post where the author tapped the digital H and V signals right on the HuC6270, then put them through 74AC123 ICs to change the pulses to something TV-compliant. Apparently, the pulse timing error doesn't yet exist at that stage.

This is reasonable, because it turns out that there is a long delay between the start of an H-sync pulse on the 6260 and the start of the corresponding pulse in composite video. The delay is not based on the dot-clock, either, but on 36 ticks of the main 21.48MHz clock. So, this basically unrelated clock is deciding when to start sync pulses...it's no surprise that this would lead to minor goofs that sensitive professional equipment would get snagged up on.

It's too bad, because I was really hoping that there would be something I could do externally to fix this problem. If I ever go through with opening up the system and modding the mod, it won't be this week.

AFC stands for Automatic Frequency Control. Basically, it's designed to deal with exactly this kind of frequency inconsistency. Consumer TVs typically have it, and Sony BVMs have it, but a lot of pro monitors don't, and maybe PC monitors as well.

I could blow a little money on an Extron box that would probably fix the problem for me, but I'm not quite sure how much money I'm willing to spend just to see the PCE on this monitor.

Here is a link to the thread, if anyone is interesting in reading through it.
http://shmups.system11.org/viewtopic.php?f=6&t=58950

Anyway, thanks again.

[SamIAm]

OK, I think I'm understanding this a little better.



See how there continue to be H-sync pulses during the V-sync period in composite sync? Those are called serration pulses, and it's these that are screwed up in the PCE.

Here are viletim's oscilloscope scans:
http://etim.net.au/temp/forum/v_sync_nes/
http://etim.net.au/temp/forum/v_sync_pce/
(Note: These waveforms are inverted compared to the pic above, as that's how TV sync actually is.)

If I'm reading everything correctly, the situation looks like this:

1. The PCE's V-sync pulse starts 4uS late.
2. H-sync serration pulses during the V-sync period are also delayed relative to the H-sync pulses before the V-sync period.
3. V-sync finishes when it was originally supposed to rather than also being 4uS late.
4. One H-sync pulse that should have happened during the V-sync period winds up being cut out, but the delay in H-sync pulses also disappears.

Whether this is actually being caused by a clock issue, I can't say. Regardless, this isn't something that is easily fixed on the composite video sync signal itself. Rather, it's necessary to get H and V sync directly from the HuC6260, as it appears that these timing errors don't yet exist there.