Chroma 81 - SCART and Colour interface for the ZX81

[tdg8934]

Hi Tim

I've sent over details of the Chroma interface to you.

Regards

Paul

Thanks Paul. I'll need to research the SCART to composite video converters some more and and place an order in a few days. Some I saw on Amazon were really cheap like under $10 and others were like the ones in the link you sent closer to $100. Seems like a huge price difference. I may have to keep the costs down initially and get only what I need to get started and get the ROM stuff later from you. Let's see... thanks again!

Tim

[tdg8934]

Paul,

It looks like composite video comes out of the SCART pins directly:
http://pinouts.ru/VideoCables/scart_ada ... nout.shtml

So I should be able to get something much cheaper like this right? Am I missing something? Is it more complex then what I am finding?
http://www.ebay.com/itm/RGB-Scart-to-Co ... 0804535154
http://www.amazon.com/Scart-Composite-S ... B006RBMJ46
https://www.youtube.com/watch?v=PabYAQz2uOI
http://skimlinks.pgpartner.com/search_g ... &rd_type=M

It looks like this is doable for a very cheap cost (SCART to composite video).

I also have a 23" widescreen monitor with SVGA and HDMI inputs that I'd rather not take up extra table space for. Those SCART converters are probably the more expensive ones.


Tim

[1024MAK]

@ Tim

Yes, you are missing something!

SCART can carry either RGB video (with the synchronization signal on the pin that would otherwise be used for composite video, it uses a total of 5 pins),
OR S-Video (using one of the pins used for RGB signals and the pin for composite video, it uses a total of 3 pins),
OR composite video (it uses 2 pins).

SCART can only carry one of these signals at any one time, and only if the sending device supports that signal. Of course, the receiving device must also support that signal .

Some devices can switch which type of signal they can send. In Europe, DVD players for example often can produce RGB or S-Video or composite video. Often the selection is done via an on screen setting menu screen. But some devices can only produce ONE of these video signals. For example, the VCRs in Europe often only had a composite video output available on their SCART sockets.

The adaptors that you have linked to are for use when a SCART socket already has a composite video output available and you want to connect it to a Phono/RCA connector/lead. They cannot convert RGB to composite video.

While taking about analogue video, I should point out that there are big differences in quality when colour video pictures are involved. The best quality pictures are RGB video, the next best is S-video, composite video produces poor colour pictures and the worst is RF/VHF/UHF video. So where possible, the preferred option is RGB video.

The interface only outputs a picture in RGB format (not composite video or S-video).

More on SCART here,

Mark

[tdg8934]

Thanks Mark for the detailed info. That really helped.

So to be clear the SCART connector ONLY supplies the RGB signals (and audio) and NOT the composite NOR component video signals (as they are not connected internally on the Chroma81 SCART connector). So this means those cheap converters wont work.

This is my current video capture device: http://www.amazon.com/Kworld-Video-Capt ... B000234SMQ

I will need to find a device that can convert "RGB" (only) SCART to one of the below methods:

1. USB 2 / USB 3 on my laptop (to bypass my Kworld video capture device entirely)
2. Composite video / SVideo (into my Kworld video capture device)
3. DVI or HDMI (into an external monitor) - not preferred
4. SVGA / VGA (into an external monitor) - not preferred

Tim

[1024MAK]

There are converters that convert 50Hz RGB to VGA (I have one).
Also available are converters that convert 50Hz RGB to HDMI. I have one of these as well.
What I don't know is if these will convert a 60Hz picture (living in a 50Hz area, I have never thought to investigate or try).

I don't have any knowledge of video capure devices that convert RGB to USB2/3.

If you want more info on the 50Hz RGB to VGA or the 50Hz RGB to HDMI converters, let me know.

Mark

[tdg8934]

What I have found so far is RGB SCART to HDMI & RGB SCART to Component video only.


$24+$5 for RGB SCART to HDMI (china) - Cheapest one as others were $50++
http://www.ebay.com/itm/720P-1080P-Blac ... 1693106704

$5+$3 for Male to Male SCART cable
http://www.ebay.com/itm/Premium-SCART-C ... 35edf88a7a


$80++ for HDMI to USB video capture device (seems expensive - so I would not buy one and use converter/scaler above with HDMI monitor)
http://www.ebay.com/sch/i.html?_odkw=sc ... e&_sacat=0

[PokeMon]

I have some issues testing a ZX81 chroma interface I got from a friend.

The first issue is the output signal itself (R/G/B on scart adapter). I use these signals to feed into my SONY broadcast monitor with an Scart Adapter having single R,G and B outputs. The monitor did not show up a picture and during testing I found out, that a pictures is shown when I additionally feed the original video output from these test candidates to the sync input and use external sync. Further testing showed up that the picture is only much to dark due to missing a back porch on the R/G/B signals. They just have the sync part and the pixel information but no back porch. If I give maximum brightness the picture can be seen quite good. It is quite funny that the monitor is able to get a black level information from the sync signal - I don't have to change the default brightness setup when connecting it. So if there ever is a modification planned to Chroma it would be nice to add this back porch for R/G/B adapters which do not offer a sync output.

It works normally (give a valid picture) with a standard (original) ZX81 with ULA.

But there are issues when using Chroma with ZX81 clones and I think the timing for grabbing the pixel information from the data bus is a bit strong and does not allow too much tolerance while all clones give perfect video signals to a monitor without Chroma. So I wouldn't say there is really a bug in the video output of these clones but surely a difference to the ULA from Ferranti.

First test candidate was the ZX81NU, a clone developed from a german ZX Team member (holmatic) which offers to use ZX81 or ZX80 directly after startup and some more features like an integrated micro disk to store some programs or data on with a modified LOAD/SAVE routine. This ZX81NU video output via Chrome behaves a bit different depending on HRG or text mode used. The HRG mode is quite readable but with disturbing lines and pixels shown as seen on the picture. As soon as I touch the contacts on the provided expansion connector the video via Chrome is very clear. So this is probably a bit capacitive load with my finger on the data bus which leads me to a timing issue here.

ZX81NU_HRG.JPG (50.28 KiB) Viewed 6423 times

ZX81NU_HRG2.JPG (57.99 KiB) Viewed 6423 times

ZX81NU_HRG3.JPG (37.87 KiB) Viewed 6423 times

[PokeMon]

The last picture shows the text mode after choosing ZX81 and entering one basic line 10 PRINT. So here you can see only two black rectangles which are the inverted chars shown. One is the cursor (should be inverted "K") and one the marker of the active line in the listing (marked with a ">"). If I touch again the data bus on the expansion connector there are characters are displayed but not the correct characters. In the picture the "4" should be a "K" and the basic line is still 10>PRINT. I know the timing is a bit different as the pixel data is not addressed with the I/R registers but with modified address data from the ROM. There are probably small timing differences as well.

The last picture shows the ZX81NU for information.

ZX81NU_TXT.JPG (30.63 KiB) Viewed 6413 times

ZX81NU_TXT2.JPG (25.66 KiB) Viewed 6413 times

ZX81NU_TXT3.JPG (81.83 KiB) Viewed 6413 times

[PokeMon]

The last picture here shows the ZX80CORE, actually a ZX80 based clone with a separate NMI module for slow mode display. This does not bring up a picture on the internal composite video as soon as the Chroma is attached but behave in same way as the ZX81NU in text mode with two inverted block chars without pixel information and rest only space character. Probably the Chroma interface just reads $00 at the data bus due a timing issue. Anyway it is not possible to get pixel data with the "finger trick". The timing here may be a bit more different.

This is the ZX80CORE setup:

ZX80CORE.JPG (64.63 KiB) Viewed 6412 times

Totally different behavior its shown with my latest clone, the ZXmore. Here it shows up only a dozen of pixels in the very first displayed pixel line at the left corner and rest of screen is white/blank. It doesn't hang at all as I can switch to FAST and SLOW mode when entering commands blind. The pixel line disappears in FAST mode and appears again in SLOW mode. I didn't investigate this in very detail now as there seem to be some general timing issues with clones. The Chroma appears to be working good with original hardware only. I am also not sure if there has been any updates to this module, I think the update was for the ROM cartridge only (which I don't use).

Are there any other experiences using the Chroma interface with ZX81 clones ?

[Fruitcake]

The Chroma interface is only intended for use with original ZX81s and I have not tested it with any clones (I don't own any).

The Chroma interface listens to the signals on the expansion bus to replicate the video picture. From your description, it does sound like timing from the clone machines is slightly different and is the cause of the lines and picture corruption you see.

I would hope that modern clones would already include the ability to output a RGB picture and so would not have need for a Chroma interface. Clone machines could also directly include support for Chroma's colour modes (the full specification for these are provided on my website).

I was originally prototyping an alternate version of the interface (Chroma 80) for use with a ZX80 while developing the version for the ZX81 (Chroma 81). I've found that timing of the expansion bus signals on the ZX80 varies slightly to those on the ZX81, and the hardware I was using would not be suitable to allow a single interface for use with both machines. A significant difference in logic implementation was required to allow colour support on the ZX80 (although from a software point of view it appeared identical), so my intention was to create a dedicated interface for the ZX80.

The CPLD that provides the logic for Chroma 81 is fully utilised and so unfortunately there is no scope for adjustments to support clone machines. A design based around a larger capacity device, e.g. a FPGA, would be required to allow finer timing control and therefore able to support a wider range of ZX81 compatible machines.