6128 Plus and sync issue

[emuola]

Hi. It's been years since I last wrote something here in these forums, but I guess everything has been working fine until now


I have a little problem with my 6128 Plus, which (for a perfectionist) is a big problem. I'd say it's sync (caps?) related, but let's see what do you guys/gals say about it.


Yesterday I played with my 6128 Plus for about an hour or so using the magnificent C4CPC. Tried 10 or so games, that I'm not so familiar with yet. It was Rambo First Blood II when the screen started to "jump" up/down one or a few more scanlines, a couple of times per second. So the picture was not rolling, just moving/jumping a bit up - down - up - down. I then thought that it's something to do the Rambo and rebooted the Plus and loaded something else (cannot remember anymore). The menu for the C4CPC did not jump at all, but as the next game got loaded, same jumping stuff again. Rebooted again, same stuff. After a few reboots and trying out Rambo again, the screen stopped jumping and the picture was stable. Unfortunately I did not be clever enough to shoot any video while the problem was "on" :/


I haven't seen this happen with any of my other retro machines, so I'd blame this on the Amstrad at this point.


My setup:


6128 Plus (not recapped)
Retrocomputer Shack PSU.
Internal floppy replaced with HxC floppy
C4CPC (latest fw)
RGB via extron MLS408 to Extron RGB190F to Sony BVM20F1E.


Does this description have any definite problem to be fixed? Recapping needed or some other "basic solution"? 


Thanks already in advance

[emuola]

Did some further testing... It seems the image starts to jump up and down after ~30 in of use. So, there's definitely something wrong with the Amstrad. Tried with MSX2, Atari 800XL etc, all with exactly the same signal path and the Amstrad is the only one with the problems. I also found out that using the Advanced Sync functionality of the Extron, the image get's stable. With all the other comps there's no need to set that on and it actually causes some weird issues with the MSX2 for example. Also, as it's a dip switch, it sucks to change the position of the switch back and fort between computers.


I red that apparently the caps are pretty good on these, but would someone have any clue what this sync problem could be about? It seems there's not so much info on this matter, or then I'm just unable to find it for some reason. Will post video later today.

[Bryce]

Hi emuola,
There are no caps directly involved in the sync signal, it's a straight connection via one resistor from the ASIC. You say that the C4CPC screen doesn't jump. Is that also true if you leave it on that screen for more than 30 minutes?

Bryce.

[emuola]

Hi emuola,
There are no caps directly involved in the sync signal, it's a straight connection via one resistor from the ASIC. You say that the C4CPC screen doesn't jump. Is that also true if you leave it on that screen for more than 30 minutes?

Bryce.

Hi Bryce, been a while


Ok, so no caps involved here. It seems to be a little tricky to reproduce this reliably. Here's the summary so far:


1. C4CPC menu screen -> no problems (yet), no matter how long the Amstrad stays on.Tens of times tested.
2. Games (various, e.g. Rambo II)) -> usually (but not 100%) problem starts after ~30 min of use. Now, for example Action biker, has been sitting there for over an hour, without problems.
3. As comparison to C4CPC, HXC disk selector menu -> one incident with sync problems so far after ~30 mins of use.


I guess this really does not make any sense?

[Bryce]

I'm no software expert, but maybe one of the experts here could look at what CRTC tricks those games use. The games may be changing CRTC settings that your extron doesn't like. If you can't find a solution, solder a proper switch in parallel across the DIP switch that you'd need to switch on for the CPC.


Bryce.

[emuola]

I'm no software expert, but maybe one of the experts here could look at what CRTC tricks those games use. The games may be changing CRTC settings that your extron doesn't like. If you can't find a solution, solder a proper switch in parallel across the DIP switch that you'd need to switch on for the CPC.


Bryce.

OK, I'll try to find more games with problems or some kind of a pattern what works and what does not. Thanks 

[emuola]

Hmm. It seems that the Rambo First Blood pt. II is a good candidate to trigger the problem. Now I managed to get the jumping to happen after 5 mins of use. Once again the Advanced sync setting of Extron fixed it immediately. What is interesting, is that turning the Advanced sync setting back off, does not necessarily trigger the problem back immediately. Weird.

[Bryce]

If you want to try something inside the CPC to resolve the issue, you could replace R87 (just above the third RAM IC) with a lower value. It's currently 220R. Try a 100R resistor instead and see if that improves the syncing.

Bryce.

[GUNHED]

Seriously? For every software problem you suggest to solder something?


Better check for R3 scrolling.

[Bryce]

Seriously? For every software problem you suggest to solder something?


Better check for R3 scrolling.

You mean like I suggested here, BEFORE I suggested any soldering?

I'm no software expert, but maybe one of the experts here could look at what CRTC tricks those games use. The games may be changing CRTC settings that your extron doesn't like. If you can't find a solution, solder a proper switch in parallel across the DIP switch that you'd need to switch on for the CPC.


Bryce.

Bryce.

[emuola]

Seriously? For every software problem you suggest to solder something?


Better check for R3 scrolling.

Just tested Edge Grinder and it's jerky as hell with the Sony CRT with both Advanced sync on or off. :/ I wonder what's the culprit here? At the moment I don't have suitable cables for connecting the Amstrad directly to the Sony BVM. Currently the signal goes through the Extron, but it should not alter the signal with the Advanced sync off.

[Bryce]

The advance sync is most likely just an LM1881 sync splitter cleaning the sync signal. With certain games (try Transputer for example), the programmer used CRTC tricks to smoothen the scrolling and no tweaking or additional hardware will get it to display on a digital screen. These types of games only work on real CRTs.

Bryce.

[emuola]

The advance sync is most likely just an LM1881 sync splitter cleaning the sync signal. With certain games (try Transputer for example), the programmer used CRTC tricks to smoothen the scrolling and no tweaking or additional hardware will get it to display on a digital screen. These types of games only work on real CRTs.

Bryce.

In general, does the LM1881 "ruin" these kind of games with some clever tricks for smooth scrolling? So, does any hw that's fixes the sync somehow, ruin these kind of games? I got kinda confused when you said these games will only run on a real CRT, because that's what I have

[Bryce]

No, the LM1881 cleans the signal, ie: ensure the voltage levels, rise time etc are correct and filters out any unwanted spikes. The bit that messes up the syncing with LCD screens is the fact that when these screens get a "new" signal, they lock onto that sync rate and sample the signal at that rate from then on. Some games change the sync rate mid-game which results in the sampling no longer being in sync with the signal coming from the computer. On analogue TV's this isn't an issue because there's no sampling being done, the actual sync signal is being used to trigger the electron beams scanning/returning, so the picture will adjust to any changes without you seeing any jitter.

Bryce.

[emuola]

No, the LM1881 cleans the signal, ie: ensure the voltage levels, rise time etc are correct and filters out any unwanted spikes. The bit that messes up the syncing with LCD screens is the fact that when these screens get a "new" signal, they lock onto that sync rate and sample the signal at that rate from then on. Some games change the sync rate mid-game which results in the sampling no longer being in sync with the signal coming from the computer. On analogue TV's this isn't an issue because there's no sampling being done, the actual sync signal is being used to trigger the electron beams scanning/returning, so the picture will adjust to any changes without you seeing any jitter.


Ok, thanks for the explanation Bryce. I'm using a analogue CRT, but still there are these weird issues with the R3 scrolling. Is this somehow explainable?


Bryce.

[Bryce]

What's the exact model of TV? There are many later CRT's that weren't purely analogue due to them "improving the picture", 100Hz refresh rates and other crap that nobody needed.


Bryce.

[emuola]

What's the exact model of TV? There are many later CRT's that weren't purely analogue due to them "improving the picture", 100Hz refresh rates and other crap that nobody needed.


Bryce.

That's a Sony BVM20F1E studio monitor   Does not have any of that 100Hz crap for sure   Manufactured in 1996.

[Bryce]

Ok, Sony BVMs are the epitome of "show what came down the cable". There's zero picture manipulation, because of what they were made for.

Bryce.

[emuola]

Some more info, based on tests last night. I bypassed both Extrons, so now the signal path is like this. 6128 Plus -> SyncStrike (has this LM1881 inside) -> Sony. Difference now is that the signal now is fed to Sony via HD15-BNC cable with *separate sync* (one of the 4 bnc connectors) . Before the signal was made RGsB inside the Extron and fed like (only with 3 BNCs) that to Sony.


Guess what? R3 still does not work. Now this is getting wicked. How is that possible? At the moment I don't have a suitable cable for connecting Amstrad directly (without the SyncStrike) to Sony, but as Bryce said, the LM1881 should not cause this. Here's a short video how the Edge  Grinder looks like.



https://drive.google.com/open?id=18lS60MmNgK5ZIA1YMtX-oDJEPeN87sOo



Interesting to say the least. Any more clues based on this?


*edit*


I'll make a cable like this: SCART -> 4 BNCs (R,G,B, CSYNC). Let's see how it goes from there

[GUNHED]

As told before, it's CRTC Register 3 dependent horizontal scrolling. It will not work with a Sony monitor. You need to get a CPC monitor for R3. Anything else should be fine on such a monitor.


I had one too and also developed a patch for R3 scrolling, but eventually I moved and dumped the monitor.


The patch for R3 scrolling will alleviate the effect, but it will not cure it. With that kind of monitors it will happen that way. But they are great for anything else.

[emuola]

As told before, it's CRTC Register 3 dependent horizontal scrolling. It will not work with a Sony monitor. You need to get a CPC monitor for R3. Anything else should be fine on such a monitor.


I had one too and also developed a patch for R3 scrolling, but eventually I moved and dumped the monitor.


The patch for R3 scrolling will alleviate the effect, but it will not cure it. With that kind of monitors it will happen that way. But they are great for anything else.

Ok, now I got it Thank you Bryce and GUNHED for helping me out I actually rearranged the signal path of all my retro computers and now 6128 Plus, MSX2+ and CoreGrafx go via SyncStrike straight to Sony without (any) Extron(s). Based on first tests it seems the original sync problem is gone. Not even Rambo gets crazy. Brilliant!