I wanted to be able to hook up an Amiga 500 to TV with decent picture quality for an upcoming gaming session. Instead of using the lousy composhite signal or the dreadful RF antenna cable I wanted something much better because modern TVs don’t forgive low-quality video signal. In fact, flat panel TVs tend to make old systems look really bad. Low resolution and sharp, pixelled graphics (in contrary to e.g. real life video) transferred using blurry and noisy signal, and then displayed stretched and processed on a HD panel. You know it ain’t gonna be pretty! The best way to prevent the worst eyesores is to use high quality video signal at least.
The most convincing diagram I found for such cable was at Ian Stedman’s site. Instead of blindly building wirings using diagrams found off Internet I like to cross reference things. After having a look at Amiga RGB port’s pinout one of the connections in Stedman’s diagram was starting to look suspicious.
Amiga RGB pin 13 to SCART pin 18
According to NT Rautanen’s pinout information on the Amiga RGB port pin 13 is “Ground Return for External Clock Enable”. Connecting that to blanking signal ground did not seem logical to me. Additionally, NT Rautanen’s page specifically says that do not connect pin 13 with pins 16-20 (video ground) on the Amiga RGB port.
As it turns out Stedman’s diagram is likely to do just that. It is likely that most TVs and monitors with SCART socket have all the ground lines at SCART socket connected. I backed up this assumption with the TVs in our household using a continuity meter. In other words connecting Amiga RGB port pin 13 to any SCART ground connects it with the common ground (including the pins 16-20 in Amiga RGB port).
Other than that, Stedman’s diagram looked good, including all the control signals and some resistors to adjust the voltage levels.
I built my cable according to Stedman’s diagram, except for the above mentioned A13-S18 line. I also used the suggested 220 ohm resistor instead of the 75 ohms in the blanking / RGB mode selection signal (SCART pin 16).
Here is Stedman’s diagram with the modification I did for my cable:
The cable I had and used for this was 8-wire shielded data cable. Six wires was needed for the signals, so two wires and the shielding copper web was left for the ground. I connected all the relevant ground pins making use of the remaining wires and the shielding. The more ground pins connected the better, I think.
Left: The Amiga end of the cable in the making. RGB signal wires and resistors soldered in.
Right: Cable ends ready for test.
I measured the voltage levels on the control and sync signals when the cable was connected to TV, and they were well in the specified range.
- SCART pin 8, status & aspect ratio: 11,5 V, telling the TV to switch to this AV input in 4:3 aspect ratio.
- SCART pin 16, blanking signal & RGB-mode: 1,2 V, telling the TV to use RGB signal.
- SCART pin 20, composite video/sync: average voltage level of 0,6 V
The Amiga RGB port does not provide audio, so the cable needs additional RCA plugs for left and right channels, which can be found at the back of the Amiga. I used some leftover AV cable which I cut and soldered to the SCART plug and ran alongside the video cable.
The cable works fine, and on both TVs I tested it the picture quality is very good (Samsung LCD, Panasonic plasma). The TVs accept the signals without problem, and both TVs switch to AV mode and displays the picture in correct mode and ratio without touching any settings. The picture is very sharp and quite stable. There is only very slight vertical wobbling in the picture, but we are talking about a magnitude of 1/10th of effective (signal) pixel at most. On Panasonic considerably less than on Samsung. It’s possible that using a cable better suited for analog signal would eliminate this slight defect.
Left: Finished cable.
Right: Cable connected to Amiga – excuse me for not having the lid on!
Ok, I was happy, but vintage machines on a modern HD TV just does not look “right”. I dug out my Philips CM 8833 monitor and had a go with that. To my disappointment the picture was missing the red component. I wouldn’t be taking that so I started investigating. I double-checked with completely different system and RGB cable and clearly the fault was not in the cable but in the monitor. I opened the monitor casing hoping that I would find a clear discontinuity in the red signal line, hopefully near the SCART plug. I had to dismantle the monitor internals quite a bit before I got access to the underside of the circuit board where the SCART plug was soldered to. A word of warning: CRT tubes use very high voltages. There’s a risk of serious electric shock if you stick your finger to a wrong spot.
And as it turned out, the fault was obvious. Many of the SCART plug’s connections were broken off from the PCB. I simply heated up the connector legs one by one and applied some fresh resin.
Put everything back together and tried again. Brilliant picture I tell you!
The tube does not scale or process anything, it now displays the video signal in its rawest, purest form. Although low resolution, the picture is sharp! The picture consists of lines of light, not of blocky pixels. The lovely scanlines are there. Just like it should be. It’s perfect!
Now some pictures I took for comparison.
All these photos are directly from the camera without any further processing – so they are quite big. Taking pictures from TV is a bit tricky. I did my best to avoid scan stripes on the CRT. On flat panel TVs there are some moiré on the pictures. Of course you can’t see either of these effects on a plain eye. These pictures are provided only to give an idea of the differences. The best pictures to compare are the Turrican stage photos.
First off some pictures taken from the Philips CM8833 monitor using composite signal. The signal is generated by the external A520 video module.
And then similar pictures from the same monitor, but using the RGB SCART cable.
Now pictures taken from low-end Samsung LCD TV (the model is LE23R86BC if you’re interested of such details) using the RGB cable.
And finally, pictures taken from Panasonic plasma-TV (TX-P42G20E), again using the RGB SCART. The second Turrican stage picture is taken using a zoom mode on the TV, so that the picture fills most of the screen area.