CPCWiki forum

General Category => Amstrad CPC hardware => Topic started by: endangermice on 11:47, 03 August 12

Title: Dirty video on 6128 Plus RGB Scart
Post by: endangermice on 11:47, 03 August 12
Hi Guys,


I've just hooked up my Plus with RGB cable to my TV and I'm a bit disappointed with the display. The colours are not as clean or solid as they are on my CPC 6128. If you look at the blue background on the attached screenshot (a little bit hard to see because the camera seems to try filtering it out - but look to the bottom!). With the naked eye I would describe it as broken up by different shades of blue in vertical lines across the screen and is very noticeable - looks like a poor quality composite signal. This does not happen on the CPC and everything is solid.


I made the leads for both machines using exactly the same components apart from the differing DIN and wires for the plus. Everything is grounded correctly. I've made many leads for all sorts of machines in the past and this is the first time I have seen this.


Is this a known issue on the Plus? Is there anything I can do to fix it? It might be the failure of a cap or something on the motherboard but I'd like to know first whether or not this is normal before I start poking around inside!


Cheers,


Damien.

Title: Re: Dirty video on 6128 Plus RGB Scart
Post by: Bryce on 12:36, 03 August 12
Call me blind, but I can't see anything wrong there.

Bryce.
Title: Re: Dirty video on 6128 Plus RGB Scart
Post by: endangermice on 13:03, 03 August 12
I know it doesn't show up properly in the picture, I think the IPhone does some sort of funky filtering. I'll try another one with my D700.
Title: Re: Dirty video on 6128 Plus RGB Scart
Post by: arnoldemu on 13:17, 03 August 12
I know exactly what you mean and I see it on my plus monitor.
I think the gx4000 doesn't suffer from the same problem because it's clock has a different speed and the resolution is then more cleaner.
Title: Re: Dirty video on 6128 Plus RGB Scart
Post by: Bryce on 13:20, 03 August 12
The GX4000 is a completely different matter, because it has a proper SCART driver and circuitry.

Bryce.
(Still disappointed because I thought the "Dirty Video" was going to feature Georgia Salpa) :(
Title: Re: Dirty video on 6128 Plus RGB Scart
Post by: arnoldemu on 13:22, 03 August 12
Quote from: Bryce on 13:20, 03 August 12
The GX4000 is a completely different matter, because it has a proper SCART driver and circuitry.

Bryce.
(Still disappointed because I thought the "Dirty Video" was going to feature Georgia Salpa) :(
I was thinking the gx4000 connected to a colour monitor would not show it.

Title: Re: Dirty video on 6128 Plus RGB Scart
Post by: endangermice on 13:27, 03 August 12
Quote(Still disappointed because I thought the "Dirty Video" was going to feature Georgia Salpa) (http://www.cpcwiki.eu/forum/Smileys/SoLoSMiLeYS1/sad.gif) (http://www.cpcwiki.eu/forum/Smileys/SoLoSMiLeYS1/sad.gif)
Hehe, sorry to dissapoint you! I thought the GX4000 clock speed only affected the composite video not the RGB. The CPC though is fine which I find odd seeing as it's older hardware.

I've taken a better picture using my Nikon, it shows the problem better. Oddly I've had to link to the picture from my webspace since the forum is now rejecting uploads due to a security check failure!?

I'm afraid I'm a bit of a stickler when it comes to video cleanliness, mainly I suspect becuase I work in the industry  ;D .

Anyway this is what I'm talking about....
(http://www.endangermice.co.uk/images/PlusDisplayIssue.jpg) (http://www.endangermice.co.uk/images/PlusDisplayIssue.jpg)

Don't get me wrong, I do like some streaky things, bacon being one of them, but alas video is not!
Title: Re: Dirty video on 6128 Plus RGB Scart
Post by: arnoldemu on 13:36, 03 August 12
Yes I see exactly that on mine. I don't see bacon though.
Title: Re: Dirty video on 6128 Plus RGB Scart
Post by: endangermice on 13:40, 03 August 12
No the bacon is only seen on very special occasions and usually only at the breakfast table!


Cool sounds like it is just how the Plus outputs video down the SCART, a bit of a shame considering the CPC does it much better!


Bryce are you able to confirm that you see the same issue on yours? Being the electronics genius that you are, any suggestions on how it might be improved or is this what comes directly from the ASIC?
Title: Re: Dirty video on 6128 Plus RGB Scart
Post by: Bryce on 13:50, 03 August 12
Strange. The blue on the border is fine, but not in the main window?

I have always had my Plus connected through my S-Video adapter, which definitely doesn't have this effect. I will measure the outputs on mine at the weekend and see where the problem lies. It looks like the Green signal isn't being pulled low enough. So the Blue isn't pure blue, but Blue/Green. But why doesn't it happen on the border too?

Bryce.


Edit: Silly experiment: If you change the text colour to the same blue as the background and do a CLS, is the background still this green or is it now normal blue?
Title: Re: Dirty video on 6128 Plus RGB Scart
Post by: TotO on 13:56, 03 August 12
Quote from: Bryce on 13:50, 03 August 12
Strange. The blue on the border is fine, but not in the main window?
You got the same effect on both, but it's more visible on lighten colours.
Title: Re: Dirty video on 6128 Plus RGB Scart
Post by: endangermice on 14:01, 03 August 12
Toto, you beat me to it :). Yep exactly it is also present on the blue border, but the camera didn't pick it up very well. trying to shoot screen displays is really difficult, but it is there only less pronounced.


Sounds like this is a common problem, I find it very strange and also odd that your SVHS adapter seems to eliminate it. I presume it's taking the RBG video then multiplexing it into a chroma and luminance signal in which case unless it filters too I can't see how the conversion would eliminate the problem which might suggest your plus isn't exhibiting it....
Title: Re: Dirty video on 6128 Plus RGB Scart
Post by: TotO on 14:11, 03 August 12
What RGB cable are you using ?
Is, the video grounds are separated from others ?
What is the length of this cable ?
Do you get a better result by "touching/moving" the CPC video connector ?
Title: Re: Dirty video on 6128 Plus RGB Scart
Post by: Bryce on 14:19, 03 August 12
The S-Video circuit uses an AD724 to do all the RGB mixing trickery, so I'm not sure exactly what it does as far as filtering is concerned.

You could try a few things to find out where it's coming from:

1) Change text and background to pure blue and CLS. Then check pins 2,3,4,5,6,7,8 and 9 of IC115 (AMS48464). They should all be 0. If not, it's the ASIC that's generating the strange colour.
2) If all are 0 (as I suspect), then the problem is leakage between the analogue signals. The RGB outputs from IC115 are pins 15,16 and 17 - B, G and R. With no monitor connected and the CPC turned off, the resistance between 15 and 16 should be about 7.5K, the resistance between 16 and 17 should be about 7.2K and the resistance between 15 and 17 should be about 8.2K - Check if that is what you are getting.

If the values are lower, then you are getting signal leakage from one colour to the other.

Bryce.

Edit: good point TotO, if the cable is too long, you might be getting cross-talk in the cable.
Title: Re: Dirty video on 6128 Plus RGB Scart
Post by: endangermice on 14:21, 03 August 12
I'm using a cable I made myself. It's not individually screened but there is a screen around all 9 wires which is connected to the connector screens on both ends. I'm using exactly the same components that I have used to make cables for my CPC, Archimedes, Amiga and C64 (although C64 is SVHS). The cable is 2m long.


It works just fine on all the other computers and doesn't look like interference since the pattern is completely constant and is not in anyway affected by the sound.


Touching the ground doesn't help which suggests that the grounding is ok. It literally looks like some sort of filtering issue or lack of filtering on the signal. Apart from that it looks really good, sharp and clear just streaky in the solid colours!


I take your point about cable length, but these cables are professionally sourced, we use them in the office all the time and I've used them to make several RGB cables for other computers without issue so I don't think it is the cable but rather something funky at the Plus end....


I'll check out those voltages and let you know how I get on - thanks for the advice guys it's really appreciated :) .
Title: Re: Dirty video on 6128 Plus RGB Scart
Post by: Bryce on 14:23, 03 August 12
Hold on. Is it just the streaks, or is the colour also changed? The picture you've posted, looks like the window is almost touquise and not blue as it should be.

Bryce.
Title: Re: Dirty video on 6128 Plus RGB Scart
Post by: endangermice on 14:30, 03 August 12
No only the steaks. The colour discrepancies will be due to ambient light in the room and me not colour balancing the camera image. It was just a quick and dirty shot to show the streaking. Colour is absolutely fine and consistent with my CPC.


I presume you mean ic numbered 40464 not 48464 and that it's number 15 not 115..?
Title: Re: Dirty video on 6128 Plus RGB Scart
Post by: arnoldemu on 14:35, 03 August 12
Quote from: Bryce on 14:23, 03 August 12
Hold on. Is it just the streaks, or is the colour also changed? The picture you've posted, looks like the window is almost touquise and not blue as it should be.

Bryce.
I get the same display on a 6128 plus connected to an official Plus monitor, with no other cables between.
Title: Re: Dirty video on 6128 Plus RGB Scart
Post by: Bryce on 14:37, 03 August 12
Ah, ok, then that's an easy one - Yes, every Plus does that (except mine, because I've modded that bug out ages ago :) ). The streaks are due to the fact, that the cost cutting idiots at Amstrad in their infinite wisdom decided to delete the RGB output capacitors on the Plus. Add a 220pf capacitor to each colour signal (between colour signal and ground) and the streaks will disappear. Any higher than 220pf and the colours will run into each other and make the picture blurred.

Bryce.

Edit: Here you can see the difference:

Title: Re: Dirty video on 6128 Plus RGB Scart
Post by: endangermice on 14:51, 03 August 12
Typical - why miss these out? It's crazy they cost a few cents each. On well you live and learn. Bryce - thank you for coming to my rescue again! Hopefully this thread is going to help a lot of people who have had similar issues. I suspect arnoldemu, toto and I will be making a trip to our local component suppliers very shortly!
Title: Re: Dirty video on 6128 Plus RGB Scart
Post by: Bryce on 15:04, 03 August 12
Before you go out and buy any parts, let me check exactly what values I used. I remember experimenting a bit before I decided which value to use, but I can't remember what the final choice was.

Bryce.
Title: Re: Dirty video on 6128 Plus RGB Scart
Post by: TotO on 15:05, 03 August 12
I though that all your custom RGB video cables already got capacitors inside. :p
Title: Re: Dirty video on 6128 Plus RGB Scart
Post by: Gryzor on 15:19, 03 August 12
Quote from: endangermice on 14:51, 03 August 12
Typical - why miss these out? It's crazy they cost a few cents each.


A few cents here, a few cents there, plus the fact that probably nobody would notice it then and you get the idea.
Title: Re: Dirty video on 6128 Plus RGB Scart
Post by: endangermice on 16:26, 03 August 12
Cool no probs, no rush :)
Title: Re: Dirty video on 6128 Plus RGB Scart
Post by: MacDeath on 17:02, 03 August 12
Quotethe cost cutting idiots at Amstrad
why plurial...only one... Sir Alan himself.
Title: Re: Dirty video on 6128 Plus RGB Scart
Post by: endangermice on 18:02, 03 August 12
Well considering that previously he was making cheap crappy HIFI systems the fact that the CPC and its successors were excellent machines has to be considered something of an achievement!
Title: Re: Dirty video on 6128 Plus RGB Scart
Post by: Bryce on 19:56, 03 August 12
Quote from: TotO on 15:05, 03 August 12
I though that all your custom RGB video cables already got capacitors inside. :p

My version has one capacitor to hold the RGB switching voltage, but no capacitors for the colours, otherwise it would mess up the colours on a standard CPC.

Bryce.
Title: Re: Dirty video on 6128 Plus RGB Scart
Post by: endangermice on 23:13, 03 August 12
Yeah mine is exactly the same as Bryce's one capacitor to hold the RGB switching voltage but nothing else. If I'm going to solve a problem I'd much rather do it on the motherboard, it's neater and means the thing is outputting something correctly in the first instance.
Title: Re: Dirty video on 6128 Plus RGB Scart
Post by: Badstarr on 23:18, 03 August 12
Curious... My plus doesn't seem to suffer from this problem, this confirmed by me plugging it into several TVs since I bought it a year ago. No evidence of this problem on a GT65 either, although the horizontal picture is shifted slightly to the left. The previous owner had hacked a simple filter onto the power input to solve a slight bit of interference from the PSU (that I believe came from a flat bed scanne the, PSU not the interference  ;) ) perhaps this filter is somehow preventing the dirty video effect on my plus. I do plan on removing the filter and using a different PSU that I recently bought from eBay, however the PSU gets crazy hot so it's either a crap regulator inside or it's not properly heat sinked this makes me a little worried about using it with my plus for now until I crack open the case and see what's what.
Title: Re: Dirty video on 6128 Plus RGB Scart
Post by: endangermice on 23:53, 03 August 12
Interesting, maybe it only affects some machines though that really doesn't make any sense unless there were some late revisions to the motherboard....


I think you're very wise pulling the power supply apart before you do anything with it. A lot of the less scrupulous companies hide a multitude of sins beneath the casing!


I'm running mine on an excellent little switched power supply I bought a few years ago. It's meant to power a shuttle mini atx board but works very nicely with both my CPC and Plus! We used the same supply at work to power a few video playback machines which were run 24/7 for several years without issue which hopefully is a good sign!
Title: Re: Dirty video on 6128 Plus RGB Scart
Post by: TotO on 11:51, 04 August 12
Most quality videogames RGB cable get capacitors on the RGB lines.
Sure, if the system already got them, the colours may look a bit washed.
But w/o, you get worst... (like on the picture)
Title: Re: Dirty video on 6128 Plus RGB Scart
Post by: Badstarr on 11:52, 04 August 12
I have tried the PSU on a recently acquired, sold as faulty 464, which is working ok by the looks of things apart from the big hole in the case which apparently happened while it was on route to me. Kinda annoying since I'm using the case for my über CPC project. Just sat idle with no peripherals attached the PSU was very hot but even with no load the unit gets really hot too but it's output seems rock solid stable. At least it's worked well enough to confirm the 464 main board works fine so I have a spare to add to the pile lol!
Title: Re: Dirty video on 6128 Plus RGB Scart
Post by: arnoldemu on 15:35, 04 August 12
Quote from: Bryce on 14:37, 03 August 12
Ah, ok, then that's an easy one - Yes, every Plus does that (except mine, because I've modded that bug out ages ago :) ). The streaks are due to the fact, that the cost cutting idiots at Amstrad in their infinite wisdom decided to delete the RGB output capacitors on the Plus. Add a 220pf capacitor to each colour signal (between colour signal and ground) and the streaks will disappear. Any higher than 220pf and the colours will run into each other and make the picture blurred.

Bryce.

Edit: Here you can see the difference:
wiki wiki wiki ;)
Title: Re: Dirty video on 6128 Plus RGB Scart
Post by: endangermice on 23:41, 04 August 12
Yeah this should definitely be a wiki article, once Bryce has got back with value of the caps!
Title: Re: Dirty video on 6128 Plus RGB Scart
Post by: TotO on 08:31, 05 August 12
Quote from: endangermice on 23:41, 04 August 12
Yeah this should definitely be a wiki article, once Bryce has got back with value of the caps!
The schematic show 220pF capacitors, as usually on RGB cables or inside videogame systems.
Now, if AMSTRAD remove them it may because the CPC+ Monitor get inputs capacitors instead...
Title: Re: Dirty video on 6128 Plus RGB Scart
Post by: Bryce on 11:01, 05 August 12
I just checked. I used 220pf capacitors, but I've never connected my Plus to a SCART connection, it has an S-Video output. I also power it from a Spectrum +3 power supply and have extra filtering added, so it may be the result of all these factors, that I have no streaks. Adding the 220pf capacitors can't hurt though, so I'd recommend adding them anyway.

@Badstarr: A GT65 definitely won't display this bug, as it doesn't use the RGB signals at all, only the LUM signal is used to produce the Green screen picture.

Bryce.
Title: Re: Dirty video on 6128 Plus RGB Scart
Post by: Badstarr on 15:36, 05 August 12
Quote from: Bryce on 11:01, 05 August 12

@Badstarr: A GT65 definitely won't display this bug, as it doesn't use the RGB signals at all, only the LUM signal is used to produce the Green screen picture.

Bryce.


I thought that might be the case, although I'm a little unclear (probably as I haven't given it much thought) how the GT65 uses the signal from the CPC to create the image, I had an instance where one of the RGB signals didn't reach the GT65 due to a dry joint and I was presented with a poor contrast monochrome image. So am I correct in assuming that the RGB signals are summed together within the GT65? Or are they simply not relevant at all?
Title: Re: Dirty video on 6128 Plus RGB Scart
Post by: arnoldemu on 15:42, 05 August 12
Quote from: Badstarr on 15:36, 05 August 12

I thought that might be the case, although I'm a little unclear (probably as I haven't given it much thought) how the GT65 uses the signal from the CPC to create the image, I had an instance where one of the RGB signals didn't reach the GT65 due to a dry joint and I was presented with a poor contrast monochrome image. So am I correct in assuming that the RGB signals are summed together within the GT65? Or are they simply not relevant at all?
they are summed within the cpc/plus and appear as lum on the monitor connector.
I think the gt65 only uses lum,gnd and sync.
Title: Re: Dirty video on 6128 Plus RGB Scart
Post by: Badstarr on 15:52, 05 August 12
I see, so I wonder why loosing one of the RGB signals resulted in the image I described? I don't fancy cutting the signals again on my GT65 screen to experiment but I wonder if it works like a balanced XLR cable anyone with a bit of studio processing experience will be familiar with this and when they go "one legged" and you get a noisy low gain signal. Just a theory anyways.
Title: Re: Dirty video on 6128 Plus RGB Scart
Post by: arnoldemu on 09:10, 06 August 12
Quote from: Badstarr on 15:52, 05 August 12
I see, so I wonder why loosing one of the RGB signals resulted in the image I described? I don't fancy cutting the signals again on my GT65 screen to experiment but I wonder if it works like a balanced XLR cable anyone with a bit of studio processing experience will be familiar with this and when they go "one legged" and you get a noisy low gain signal. Just a theory anyways.
I don't understand what you are talking about. But do you mean that the gt65 could be sinking/draining the r,g,b signals although it doesn't use them, this produces a load on the signals that modifies their appearance?
Title: Re: Dirty video on 6128 Plus RGB Scart
Post by: Badstarr on 14:32, 06 August 12
This link might explain my thinking ... Balanced Audio Cable (http://www.mediacollege.com/audio/balanced/cable-balanced.html)


I'm not saying this is actually what is happening, it's just an idea really. I'm sure Bryce knows better than me. I'm just wondering if my GT65 (one of the last to be made) has a different method of pulling the signal from the CPC? There are definitely RGB connections present on the connector and they do seem to effect the image on the screen. I used to do a little trick years ago as a quick fix when the original plug dropped off the video cable on the GT65, basically I used to tie the RGB connections together and connect them to the sync pin (as far as I remember) and I would get a decent image, if a little saturated, on the monitor.


Without the RGB signals the picture on screen is washed out. I haven't got access to my older GT65 so I can't check if it also happens on that one.
Title: Re: Dirty video on 6128 Plus RGB Scart
Post by: Bryce on 14:46, 06 August 12
According to the GT65 schematics, RG and B are terminated, ie: they are all connected to ground through a 100R resistor. The reason for this is that they would effectively be an antenna if they weren't terminated, which would interfere with the signals required - LUM and Sync, but the RGB signals definitely aren't used to produce the any part of the picture on the screen. If you had a break in one of the RGB wires, it could still have a negative effect on the picture quality due to the interference it could cause.

Bryce.
Title: Re: Dirty video on 6128 Plus RGB Scart
Post by: Badstarr on 15:33, 06 August 12
I see, thanks Bryce, now it makes sense  ;D
Title: Re: Dirty video on 6128 Plus RGB Scart
Post by: Overkill on 10:39, 04 August 20
First off, thanks for all the information here :)

I'm going to try modding my cable, instead of the machine, that might work, right?
Title: Re: Dirty video on 6128 Plus RGB Scart
Post by: Overkill on 16:26, 08 August 20
Who's had success with 220pf capacitors?  I tried them in the cable and on the mainboard, and I see no picture improvement, still got the vertical bars.
Title: Re: Dirty video on 6128 Plus RGB Scart
Post by: ||C|-|E|| on 18:16, 11 August 20
In the PLUS? I had pretty good results with them installed on the board. I would not say that the bars completely disappeared, but difference was very noticeable  :)
Title: Re: Dirty video on 6128 Plus RGB Scart
Post by: RetroCPC on 03:27, 14 November 20
Quote from: Bryce on 14:46, 06 August 12
According to the GT65 schematics, RG and B are terminated, ie: they are all connected to ground through a 100R resistor. The reason for this is that they would effectively be an antenna if they weren't terminated, which would interfere with the signals required - LUM and Sync, but the RGB signals definitely aren't used to produce the any part of the picture on the screen. If you had a break in one of the RGB wires, it could still have a negative effect on the picture quality due to the interference it could cause.

Bryce.

An old thread, but for anyone experiencing the same issue as Badstarr with his GT-65, the individual RGB signals lines need to be correctly terminated even with the Green Screen monitor to insure the correct "Colour" = Gray scale "palette" (This applies to both Colour & Mono monitors). I simulated the CPC Video circuit to better understand how it functions while designing a new video output circuit for my CPC... Today's High Bandwidth / HiRes monitors tend to expose shortcomings such as noise and poor termination (non constant output impedance) of the CPC original RGB video output.

The CPC's Gate Array produces 3 "Logic" States* per RGB pin:- High, Low & TriState (resulting in the Pin floating / open). The resultant video signal (voltage) in TriState mode is determined by external resistors / termination resistors in the CPC's video output circuit (the termination resistors within the monitor form the final part of this circuit). The Monitors RGB termination resistors form the final part of the voltage divider circuit - thus defining the "Mid Level" video voltage during TriState mode.

The Green screen monitor (GT-65 etc) needs these "Mid level voltages" to correctly display the full colour palette as mono "Gray Scales" - if anyone of the RGB lines is not correctly terminated (say due to a broken monitor cable / connector) then you will shift / loose some of your gray scale palette on a Green screen monitor - this is the primary reason that the RGB lines are required to be correctly terminated by the GT-65 green screen monitor (by the 100R resistors on each of the RGB signal lines, even though they are not used "Directly" by the monitor).

* High / Low / TriState = 3 Voltage "Levels" (with the aid of the external resistor Potential divider circuit on the CPC PCB) - so the Red, Green & Blue video signals from the Gate Array can EACH produce 3 discrete voltage levels hence the 3(R) x 3(G) x 3(B) = 27 colour video palette of the CPC (0 - 26) ...

http://www.cpcwiki.eu/index.php/CPC_Palette (http://www.cpcwiki.eu/index.php/CPC_Palette)
Powered by SMFPacks Menu Editor Mod