Sync issues with new 540 build

[nama]

Hi all,
So for those who have been following my progress (and helping me a lot on the way) you will probably know I started with a spare 505 and spare 527 RAM boards. I built a backplane and Klyball data splitter and got the boards working as a serial system with floppy access. I then built a 540 board, and with a bit of modifying I got an early revision of Dave's Unified/Universal Keyboards working. So now I am really close to my end goal, however I'm having some annoying issues with the video sync...
The 540 does display video (so thats good), and I have tried the setup on various CRT monitors with varying success. With one monitor, the best I can get is to de-tune the vertical hold, and I can get the image somewhat stable, but eventually because the vertical hold is right on the edge, the screen will start to roll. If I tune it slightly back, then the image can be in the right place, but jumps and jitters a lot so also not ideal.
I am in NZ which is PAL land, and the 540 board is running a 12mhz Xtal. I have (I hope) built it with all the color circuitry installed, but actually ended up temporarily removing the color crystal for trouble shooting. I have played around a lot with all the adjustment potentiometers too. I have a variety of other US systems (including a C4 and a C2) and I've never really had any issues with getting a clear stable picture out of them on my monitors.

I have attached a video showing what it does on a monitor without all the mucking around with vertical hold. (This monitor actually has no vertical hold adjustment).

I was wondering if anyone could help or if they have any insights into where I could look.
Thanks again.

sync.mov

(3.09 MiB) Downloaded 183 times

[pbirkel]

Newbie here: What is a "Klyball data splitter"?

[nama]

Here you go:
viewtopic.php?f=3&t=220&hilit=d13
Allows you to use drives that are not original MPI.
MPI drives had the data splitter built in.

[pbirkel]

Thank you :-}.

[nama]

And for those interested, here is what the build looks like so far.
I designed and printed a slotted 3d bracket to hold all the board in place so they aren't flopping around. It's also a convenient place to put a reset switch, video out and a small breakout board that makes removing and attaching a floppy cable much easier as access to the data separator is tight.

build.jpg (313.73 KiB) Viewed 3430 times

[bxdanny]

the 540 board is running a 12mhz Xtal.

The color 540B boards built by OSI used a crystal frequency of 12.083915 MHz. That's only about 0.7% higher, but if your board is right on the hairy edge, maybe that would make enough of a difference? I know there are places that sell crystals with custom-ordered frequencies. (The optimal [to the nearest 1 Hz] crystal frequency to produce a horizontal frequency matching the NTSC spec would actually be 12.083916 MHz,) Before ordering a new crystal, maybe try swapping your 12 MHz crystal with the one on the C4P that works well, to see if it makes a difference?

[Thomas]

Looking at video of the monitor you provided, it looks for me like a vertical hold issue of the monitor.
Check if you have V Hold controls at the back of the monitor or inside and try to adjust.
If this does not change anything, you should have a look to the vertical sync level at the video output jack (monitor connected). If the signal level gets to small, vertical hold effects will show up.

[nama]

I have played extensively with vertical hold on multiple monitors (see original post), but with little success.
I will order a 12.096MHz Xtal from mouser and see if that improves things. It's still not perfect, but closer.
Don't want to pull the C4 apart as it always takes a lot of mucking around to get it working again after disassembly (I suspect some dodgy contacts to the backplane).

[nama]

Ok...good news. Had some time to look into the sync issue last night.
I discovered that if I touched the traces around C17, C18 and C19 the constant scrolling would stop. Looking at connected components I focused in on the Aliexpress special 74LS123 at U4A. Replacing that with another branded IC fixed the problem. Now I a have stable (if not a bit noisy) image.

Next I will try another monitor and see if noise persists, and if color works?

[ronin47]

Can I ask what power you're using for this machine?

The reason I ask is that the interference looks really like some I experienced with an Atari 600XL, and the issue turned out to be due to a Chineseum PSU I received with my OSSC.

Simply changing to a different PSU resolved the problem for me.