Strange issue with Sync signal from CPC6128

[tranx]

Hello everyone,

I am having an issue with my CPC6128 with the display. It used to work just fine, but the last time I took it off the shelf it was doing something weird with the display (flickering like if it had the wrong display frequency).

Everything else in the computer works fine, sound works fine, keyboard, I can type BASIC programs and run them...

Confirmed it is not a problem with the CPC-SCART cable, checked all connections and they are good, used a different cable and problem persists.

My monitor has separate RGB+S inputs so I can connect them directly, great for troubleshooting.

You cannot view this attachment.

The monitor detects the signal as 576 lines / 50 Hz. I can remove the color cables (R,G,B) one by one and the problem persists, so the problem is isolated to the Sync signal.

You cannot view this attachment.

Connected the CPC6128 to one of my Extron RGB scalers/converters to see what the Extron detects, and it says the signal is a 15KHz / 50 Hz:

You cannot view this attachment.

The output from the Extron is a blurry signal, so something is affecting the display, like if there was some sort of noise introduced.

Opened the CPC6128 and checked the video connector solder points, which look good (I resoldered and cleaned, picture is before that)

You cannot view this attachment.

The Sync signal goes from the DIN-5 connector to resistor R137, and from there to the Amstrad 40010 chip which I think is the Gate Array.

You cannot view this attachment.

Using a copper wire with a BNC connector, I connect the pin 5 of the Amstrad 40010 (Gate Array) directly to the Sync input of my monitor, and the problem persists (This is not my own picture, but just to show the pin I am using)

You cannot view this attachment. You cannot view this attachment.

Checking this signal with a handheld oscilloscope I get a 4V signal like this (does not look consistent but I think this is because this signal carries HSYNC and VSYNC together?):

You cannot view this attachment.

Looking at the schematic, my understanding is that the Sync signal is generated by the Gate Array based on HSYNC and VSYNC signals coming from the CRTC. So I am thinking I may have a bad CRTC, or a bad GA (?) How can I tell which is the problem?

All connections between the CRTC and GA are good and have good continuity. All the resistors have the correct values.

This is where it gets strange. I connect a different monitor, a Samsung (confirmed to work previously with the CP6128) and on this monitor the Sync looks good but the colors are very dim, almost completely dark:

You cannot view this attachment.

Any ideas what to look for will be appreciated. I can connect my Tektronix oscilloscope but not sure what I am looking for.

Thanks

[Bryce]

It looks like a grounding fault. Are you using two separate switched mode PSU's for display and CPC? Have you a different PSU you could try on the CPC?

Bryce.

[tranx]

It looks like a grounding fault. Are you using two separate switched mode PSU's for display and CPC? Have you a different PSU you could try on the CPC?

Bryce.

Thanks, the monitors I am using are powered by 110V AC and the CPC6128 is powered by an AC/DC adapter. The AC/DC adapter works fine (and has worked fine in the past) and is providing the correct voltage and current, everything in the CPC works just fine except for the sync. I also ran a wire to connect the ground of the BNC connectors from the monitor to the CPC ground, and that didn't make a difference.

Is there a way to tell if the problem comes from the CRTC or the GA?

[Bryce]

The easiest test would be to connect the CPC to a real CRT and see if the problem is still there.

To elaborate on my rather vague "Grounding problem" statement: The sync is a signal that is normally high and drops low to trigger a sync (as nicely shown in your scope screenshot above). If you are using different types of PSU's on the CPC and monitor, there is the possibility that the GND of the two devices are not equal relative to each other. This means that the CPC would be giving (for it's reference) a proper GND pulse, but because of the offset in ground between the devices, the monitor may not always be recognising the pulse as being close enough to ground to be a valid pulse. Try measuring the voltage between the GND of the monitor and the CPC and see if it's really 0V difference.

Another thing that can cause this issue is if the cable between the CPC and monitor is too long or there is resistance somewhere in the cable (frayed wires or a failing solder joint). So always measure the sync signal at the monitor input, not at the CPC end. The CPC is most likely producing the correct signal, but obviously the monitor is not always recognising this for some reason.

Bryce.

[llopis]

The very dark picture looks like a display without the blanking signal turned on. Is that Samsung using a SCART connection? The Amstrad doesn't have easily available power in the video connector, so you usually need to provide 5V (or less) through external means like a jack or a battery.
And if sync is working fine there, there's something about that first monitor that it doesn't like (no idea what it is).
As for the oscilloscope picture, you're too zoomed out at 50ms, so you're getting aliasing. Try zooming in and see if you're getting a regular signal, but I suspect you are.

[tranx]

The very dark picture looks like a display without the blanking signal turned on. Is that Samsung using a SCART connection? The Amstrad doesn't have easily available power in the video connector, so you usually need to provide 5V (or less) through external means like a jack or a battery.
And if sync is working fine there, there's something about that first monitor that it doesn't like (no idea what it is).
As for the oscilloscope picture, you're too zoomed out at 50ms, so you're getting aliasing. Try zooming in and see if you're getting a regular signal, but I suspect you are.

Thanks Noel, as for the 5V, this is one of those cables I purchased from a guy in UK that already includes the connector for the 5V so I am using this just in case, but to be honest my monitors automatically switch to RGB mode and don't need the 5V (I use them with my MSX and have used them with this CPC in the past). 

I will try to get a better picture from a real oscilloscope and will post here. Hopefully that will help diagnose the issue.

Thanks again

[tranx]

The easiest test would be to connect the CPC to a real CRT and see if the problem is still there.

To elaborate on my rather vague "Grounding problem" statement: The sync is a signal that is normally high and drops low to trigger a sync (as nicely shown in your scope screenshot above). If you are using different types of PSU's on the CPC and monitor, there is the possibility that the GND of the two devices are not equal relative to each other. This means that the CPC would be giving (for it's reference) a proper GND pulse, but because of the offset in ground between the devices, the monitor may not always be recognising the pulse as being close enough to ground to be a valid pulse. Try measuring the voltage between the GND of the monitor and the CPC and see if it's really 0V difference.

Another thing that can cause this issue is if the cable between the CPC and monitor is too long or there is resistance somewhere in the cable (frayed wires or a failing solder joint). So always measure the sync signal at the monitor input, not at the CPC end. The CPC is most likely producing the correct signal, but obviously the monitor is not always recognising this for some reason.

Bryce.

Thanks Bryce for your comments. The thing is I used this same cable and monitors in the past, and everything worked fine until last time I tried to use the CPC when it started doing this weird thing with the sync. 

I am almost certain the bad signal is coming from the CPC, I have spent many hours troubleshooting and used my MSX-2 to confirm cables and monitors are working fine. 

[TotO]

Well... When I have first seen the dark picture, that automatically make me remember when I was young and trying to plug a RGB signal on the SCART that only support the Composite / Y/C video. The result is a display with only the Luma and not the colours. But seeing next your nice RGB picture confuded me... (I have not took care that you have hacked the GA pin to display it)

So, yes, the issue is that the video circuit into the mainboard  not drive the colours lines, just the luminosity.

[tranx]

Hi, thanks everyone. Good news!

You folks were absolutely right, it was the 5V for the RGB via SCART. I have never needed that with this monitor, but for whatever reason it was needed with the CPC6128. I thought I had it connected during my troubleshooting, but apparently not. I guess I was so focused on the CRTC and the GA that I wasn't seeing the whole picture , so your suggestions have been very helpful.

You cannot view this attachment.

So that took care of the image for the Samsung LCD. There is some ghosting in the image, so I am now looking into that.

You cannot view this attachment.

As for the other monitor, it's still flickering, no idea why it's having issues with the signal from the CPC since it can detect the type of signal as RGB 576/50Hz Interlaced, this is a professional broadcasting monitor and takes pretty much everything you throw at it. I have tried connecting the grounds as suggested but no luck. My guess is that one of the cables has something wrong (like mixed grounds or missing grounds).

I continue my journey with the CPC6128 and my next step is to add a Gotek drive 

Again, I really appreciate the help provided, you guys are awesome!