Stedy's RGB to SCART cable

[Stedy]

Hi,

Only two reasons for a white screen I can see, the first has already been covered (Composite video only SCART).

The other possibility is that the RGB mode pin (pin 16) is for some reason missing or at the wrong voltage. With my design and a properly terminated input there should be 2.5V, which I measured during my tests. If the voltage on pin 16 <0.4V, the TV will select the composite video signal on pin 20. The Amiga does not have a composite video signal on the 32 way connector and the TV interprets the signal as a peak white video level as it is approx 1V.

I developed this cable design based upon a reference schematic for one of the analogue video decoder devices commonly used in TVs/monitors, to satisfy all signal levels it could accept.

Ian

[orange]

I would measure the voltage on pin 16 (compared to pin 18) while its connected to TV, if possible.
perhaps pin 18 is not connected properly?

[fgh]

Thanks guys, I'll just order a couple of these from amigakit then..

Another thing - just to make sure after reading this part again:

Originally Posted by Stedy

I undertook a number of measurements on the Amiga video outputs. The first issue I noticed was the voltage on the SYNC signal, it was 4.8V peak to peak.
Modern TVs have video decoder devices [...] They expect a 1V peak to peak signal, connecting the CSYNC signal from the Amiga could harm the device...

This only affects CSYNC/pin10 of the RGB port, and using a normal composite output is safe as always, right?

Thanks again!

[FOL]

Originally Posted by orange

I would measure the voltage on pin 16 (compared to pin 18) while its connected to TV, if possible.
perhaps pin 18 is not connected properly?

Pin 18 is not connected.

[magge]

I built your cable to the specs except i didn't solder SWITCH cable or R/G/B ground since it was a test and i only had 10 scart pins.
Works great though. Both on my CRT 28" and a Viewsonic N2060W-1e LCD TV.
Also i used a 397 Ohm instead of 333 and a 1,18 kOhm instead of 1k resistors. Just because that was what i had at home.
Don't know if it affects quallity.

Interlaced mode is stable, flickers a bit when stuff move around the screen, nice picture quality, i'll test with all the ground cables soldered also when i get the connector.

Thanks!

[Stedy]

@Magge

Glad the cable design worked well.
You can very the resistor values by around 10% and it should work well.

@thread

I tested my new LG LCD TV, got a nice picture on it:


Both normal and interlaced modes were handled without issue. PAL/NTSC was no problem.

I had designed a filter PCB, to filter out a 9 MHz noise component from the DAC output of the A1200. Installed it and tested it, it made absolutely no difference!

Finally, I tried plugging my cable into a composite/CVBS only socket. With the default TTL CSYNC option, I had a black screen. Using the Composite video output of the A1200 as the sync source, I got this naff picture:



That's all for now.

Ian

[orange]

to me, colours on CVBS picture look better, esp yellow (too 'cold' on RGB). did you change some tv settings? perhaps its the camera.

Hi,

I've just bought an LKV362 (http://www.lenkeng.net/Html/Product/...CONVERTER.html) and I think I have the same problem as Ferix

I've made the SCART cable using all resitors and checked it with two LCD TVs I have. It works fine, but all started because my 41" LCD tv introduces too high delay and playing games with the Amiga is impossible.

So I decided to give a try at this SCART->HDMI adapter but I can't get a picture out of it.

Well, the only way I found to see something is to use the Amiga DB23 port but composite-sync from the RCA composite-video connector (I'm using an Amiga 1200).

In this way the LKV362 is able to show me a picture, albeit with some noise and shadows.

Funny is , if I disconnect the DB32 from the amiga and I reconnect it with all other cables connected, the noise-shadows disappears.

My guess is that this device doesn't like Amiga sync signal from DB32 port.

The decoder that gets the RGB or composite-video signals is a TI : http://www.ti.com/product/tvp5160&lp...native_Devices

I'm not the greates expert on video signals, so maybe Stedy/Ferix or someone else on the board can help to shed some light on this?

With above trick seems to work nice, it's cheap and would be the perfect device to finally make that Skidmarks tournment on my 41" LCD that I promised to my friends

Thx - Cheers

-- LuKe

[andy]

Originally Posted by orange

to me, colours on CVBS picture look better, esp yellow (too 'cold' on RGB). did you change some tv settings? perhaps its the camera.

I'd expect a comparison to look similar to that (my old CRT in RGB looked very 'flat' compared to the vibrancy of s-video and composite), but the lighting and response of the camera appears to have made a difference in the images too. Compare the colour of Ian's worktop near the television stand and the green label above '3000', orange. There's a definite yellow tint in the second photograph.

I love how clean that first snapshot looks, Ian. Even if it wasn't the 'best' signal, I rather liked the built-in blending of composite back in the day for gaming - but I'd have loved that sort of quality elsewhere. I don't think I was imagining twenty or so years ago that in Space Year 2011 we'd be seeing an Amiga picture like that on a futuristic flat-screen display! Wow, they'll be making mobile communication devices that let you hold conversations away from home - and with no cables - next!

[Jonathan Drain]

Thanks for posting this schematic. My brother soldered one up for an Amiga that had broken RF and composite. Cost about £7 and worked perfectly. Great picture quality, noticeably sharper than composite.

It worked with CRT, but not a PAL LG LCD TV. We used the resistors exactly as given on the diagram but didn't wire the composite because that Amiga's composite was broken anyway.

[Stedy]

Hi,

With regard to the photos of my TV. The difference in the appearance of the pictures was most likely due to the camera. I am still learning how to use my new camera though the fluorescent lights of the workshop do not help. The RGB video text was much much clearer.

Now on CSI, they would find the 2 pixels, in the reflective plastic and from that reconstruct my face

@LukeJerry/Jonathan Drain

The Amiga's video sync is not compliant with either the PAL or NTSC specifications in non-interlaced mode and is better in interlaced mode, not ideal for games. Some TVs can not handle the non-standard Video sync. If the TV can handle a VCR playing back a macrovision protected video tape, it should work better with the Amiga.

Am I the only one here who still has a VCR?

Bye,

Ian

@Stedy

1st of all thanks a lot for the help and your page/post on SCART cable for Amiga!

I see Amiga video sync is not very compliant, I've tried anyway to set an interlaced video mode with my A1200 connected to this Lenkeng converter but the result is the same (some signal is detected but black screen).

In a previous post you said:

I'm wondering if your TV is using the SCART fast blanking feature with
Alpha blending. This is a mode where it blends the CVBS signal with the RGB.
As you are feeding a TTL CSYNC pulse, it is quite rightly getting confused
as the signal has no front/back porch to detect so black level and peak
white are hard to detect. Just a theory at the moment.

And I see from TI tvp5160 data-sheet:

http://www.ti.com/lit/ds/symlink/tvp5160.pdf

2.6 Fast Switches for SCART and Digital Overlay

The TVP5160 decoder supports the SCART interface used mainly in European audio/video end
equipment to carry mono/stereo audio, composite video, S-Video, and RGB video on the same cable. In
the event that composite video and RGB video are present simultaneously on the video pins assigned to a
SCART interface, the TVP5160 decoder assumes they are pixel synchronous to each other. The timing for
both composite video and RGB video is obtained from the composite source and its derived clock is used
to sample RGB video as well. The fast-switch input pin allows switching between these two input video
sources on a pixel-by-pixel basis. This feature can be used to, for example, overlay RGB graphics for
on-screen display onto decoded CVBS video. The SCART overlay control signals (FSS) are oversampled
at 4× the pixel clock frequency. The phase of this signal is used to mix between the CVBS input and the
analog RGB inputs. This improves the analog overlay picture quality when the external FSS and analog
video signals are generated by an asynchronous source. The TVP5160 decoder has two programmable
delays for component video to compensate for composite comb filter delays and two programmable delays
for digital RGB to compensate AFE and decimation filter delays.

Could this be the explanation of why when I use RCA-Composite-Video for Sync + RBG from the Amiga DB23 connector, I see the image, but with some typical-composite-video shadows/noise added to it?

Seem from the data sheet that this "fast switch" mode can be disabled, I hope to find some time next days to try modding this box and disable it.


-- LuKe

[Stedy]

@LuKeJerry,

The fast-switch mode is controlled by pin 16 of the SCART socket. I feed 5V via a potential divider to set 2.5V @ the TV. This forces the TV into RGB only mode. There is no need to disable the fast-switch mode, my cable does that.

The device could be blending the CVBS and RGB signals, this can be disabled by an I2C register write or two. I'll leave that up to you.

Ian

[ikonsgr]

Hi,

Stedy, i made 2-3 scart cables and tested with my AMIGA 500 but i used only one cable for all the grounds(amiga pin 16->scart pin 17) . Also i didn't use pin 8 of the scart to autoselect av (as almost on any tv, crt or lcd, you can choose "av" from the remote control) . Also instead of connecting 5volt to pin 16 of the scart i took the 12 volts from pin 22 of the amiga connector and through a 1ΚΩ resistor i put it on pin 16 of the scart. Finally i used a 330Ω resistor on the CSYNC as you suggest (amiga pin 10 to scart pin 20).(this was from a design i had, before i saw your site - btw, congratulations for you excellent analysis on the subject!)
When i connected the cable i measured the following voltage levels on the scart pins:
pin 16: 0.83volt
pin 20: 0.67volt (My amiga 500 outputs 4.34volt on pin 10 where scart pin 20 is connected)

All cables i made (even one without the 330Ω resitor) have worked perfect on 3 tv sets i have, 2 crt's and 1 lcd. So my questions are:

1) I've made some measurements with a poly meter on the various ground pins on both amiga 23 pin and scart connectors (on 2 crt tv's and one lcd tv). What i found out is, that ALL ground pins on amiga are internally connected (short circuited) but also ALL ground pins on the scart connectors (at both 3 tv's) are internally connected! So my question is, why use 5 different separate "ground" cables when actually there is only ONE ground on both amiga and scart connectors?

2) If a scart video cable works on a particular tv with an amiga 500, this means that it will also works with every amiga (600,1200,3000,4000)? That is, are there any differences on levels of CSYNC (or other signals too) between various amiga models that will require different resistors or pinout?
3) i gave one of my cables (with the 330Ω) to a guy that has an A4000 and told me that he got picture, but has something like "shadowing". He tried another cable form amigakit and works ok. So, where do you think is the problem with my cable? Is the 0.83volt on scart pin 16, too low to clearly select rgb? Or the 330Ω dropped the signal too much on the A4000? Or what?
4)I noticed on the crt tv's , that when i didnt connect the 2 audio rca to the amiga, i had a slight horizontal "noise band " that keep moving upwards on the screen. This disappeared when i connected the 2 audio rca. What can cause this?

Thanks.

[orange]

1) because of interference/crosstalk.. and I think not all grounds are connected.
2) iirc, they are all the same.
3) perhaps cable too long. or bad ground.
4) also ground

[ikonsgr]

Thanks for your reply orange, but i still cant understand some things...

Originally Posted by orange

1) because of interference/crosstalk.. and I think not all grounds are connected.

I thoroughly checked all the ground pins on the 23 pin amiga connector,i'm 100% sure they are all connected together.AND the same happens with ALL the ground pins of the scart cable, at least to the 3 tv's i checked (2 crts 1 lcd). So, i still can't understand how can be crosstalk or interference, since there is no "isolated" ground anywhere in the circuit!

Originally Posted by orange

3) perhaps cable too long. or bad ground.

Cable was only 1.5meters long(brand new shielded 6 branch low loss), and with my amiga 500 worked perfect on all 3 tv's! How can it be "bad ground" to his Amiga and "good ground" to mine? If we assume it was the guy's tv fault then neither amigakit's cable should work, but it does! On the other hand,if we assume that it was not tv's fault and since my cable work perfect to my amiga, then according to your answer to my second question, it should be working also for A4000. Go figure...

Originally Posted by orange

4) also ground

Can you be more specific about that? You mean that i had a bad soldering or something? Keep in mind that i used separated cable for the audio ground (in fact it was the shielding of the 2 coaxial cables i used for the 2 audio rca connections). As far as the cable concerns, i measured all impedences for every connection (pin to pin) and found out it was below 0.1 Ωm (the limit of my ohm meter)

[orange]

crosstalk and interference are in the cable itself. for same reason IDE, SCSI and floppy cables have bunch of GND pins/wires.

well, I have the same issue 4)
I think its because audio uses Shielding as GND. (and shield should be definitely different from RGB signal GND) so perhaps try connecting 23pin port shield to SCART?
BTW, iirc, low resistance doesn't necessarily mean short circuit. coils have low resistance.

[ikonsgr]

Hmm, that thing with the coils you said, i didn't thought about it, you are right, it is possible....
So you think, that if i use a, lets say 10 branch cable, to connect every ground separate, then i will get a more clean picture than with the current cable?
I suppose that any differences will be more visible with LCD tv's and with workbench environment and not with games, right? Maybe that's the case, because my testing was made mostly with games (where small ghosting/shadowing or other minor defects in picture quality is difficult to detect). However in workbench or any more intensive "video reading" application, where clarity and distinctness of picture is very important the separate ground will make a difference.

[orange]

I think its worth trying out. iirc, you'd only need 3 of them, one GND for each R,G,B. and one more for audio (though that one might be unnecessary if you already connected the shielding).

I remember seeing factory made SCART<->SCART cable (but without connectors). it was designed to be extender, connecting all 21 pins. and it was obvious that each R,G,B signal was shielded with its GND.

one more thing about shielding, they say one shouldn't connect it at both ends (in null-modem cables, for eg.) to avoid (ground) loops that pickup noise or something like that. have you connected it to SCART pin 21?

[ikonsgr]

Well according to stedy's design http://www.ianstedman.co.uk/Amiga/am...iga_scart.html i will need 11 pins but my cable has 10 branches...
Except from R,G,B you will also need separate grounds for pin 20 of the scart (where CSYNC pin 10 of amiga goes) and also for pin 16, the RGB selection. I think the most "essential" ground is pin 17 (the CSYNC/ composite video ground) since the other design i had ,with only one ground connected,was using this pin for all the grounds.
You think i can overlook, lets say, RGB select ground without any compromise in picture quality?Or maybe don't connect pin 8 of scart, since all tv's can select "AV" from the remote control...
I will also try what you say, connect the shielding on both ends to see if there is any difference.