Screen sometimes going momentarily black when using a RGB->SCART cable - An attempt of an explanation

Troubles with your machine? Just want to speak about the latest improvements? This is the place!

Moderators: Mug UK, Zorro 2, spiny, Greenious, Moderator Team

Post Reply
czietz
Hardware Guru
Hardware Guru
Posts: 1246
Joined: Tue May 24, 2016 6:47 pm

Screen sometimes going momentarily black when using a RGB->SCART cable - An attempt of an explanation

Post by czietz »

This is discussed so often that it’s almost a FAQ: When using an RGB→SCART cable to connect their Atari ST to a TV/monitor, people sometimes notice that the screen goes black or “jumps” every few seconds. [1] This has to do with pin 16 of the SCART connector, as I will explain.

This pin 16 is called “blanking (rapid switching)” in the SCART standard [2]. In a typical Atari RGB→SCART cable, it is connected to pin 12 (vertical sync, VSYNC) of the Atari monitor connector. [3] To understand the issue, one has to understand the original intention of the blanking pin: blanking, i.e., masking RGB signals within a video frame. That way, a device such as a subtitle decoder could superimpose transparent text over a composite video frame. Since this therefore is a high-frequency signal, its input impedance – as specified by the standard – is 75 ohms. This will be important later.

However, in practice this signal was usually used for static switching between RGB and composite video, only. For example, a device capable of RGB output, such as a DVD player, would permanently put 1 – 3 V (as per the standard) on pin 16 to keep the TV in RGB mode, while a device only outputting composite video (e.g., a VCR) would keep the pin at 0 V.

If the TV/monitor was fully standard-compliant, it would pose no problem to connect the VSYNC signal of the Atari to the blanking input of the SCART connector. After all, VSYNC is high during the entire time a picture is displayed and, thus, the TV should correctly display the RGB signals. However, it seems that some (modern?) TVs only interpret the pin as the static switching signal, like it was commonly used. If they see the signal going low during the vertical blank, sometimes they switch to composite video mode for an entire frame of video, causing the screen to go black or jump.

This is what the diode and capacitor in the schematic [4] are for: They buffer/smooth the VSYNC signal so that the TV always sees the proper voltage at pin 16 of the SCART connector, stopping the momentary switch to composite video. It’s sometimes suggested to instead use the 12 V present on pin 8 of the Atari monitor connector and a voltage divider to provide the input voltage on pin 16. This is not possible: there is a series resistor of 1200 ohms inside the Atari to protect the 12 V from short circuits. Given the input impedance of 75 ohms (see above), that way one could only get 75 / 1275 ∙ 12 V ≈ 0.7 V on pin 16, less than required by the standard.

The diode and capacitor solution [4] is proven to work – at least with my TV that suffers from the problems described in the introduction. I’ve tested half a dozen STs using that cable without any issues.

[1] This is not to be confused with the issue on some STfms where the screen borders are tearing when the screen is mostly white. See http://www.exxoshost.co.uk/atari/last/S ... /index.htm for that problem.
[2] EN50049: Domestic and similar electronic equipment interconnection requirements: Peritelevision connector.
[3] http://info-coach.fr/atari/hardware/int ... itel_cable
[4] https://www.chzsoft.de/site/site/assets ... tecker.png
tzok
Captain Atari
Captain Atari
Posts: 231
Joined: Fri Jun 30, 2017 7:22 pm
Location: Bielsko-Biala, PL
Contact:

Re: Screen sometimes going momentarily black when using a RGB->SCART cable - An attempt of an explanation

Post by tzok »

In STe the HSync & VSync also have a 33R series resistor. Whats more, the diode introduces further voltage drop out. It should be Shottky diode to ensure minimal voltage drop. 1N4148 is not ideal (won't work with my STe & Sony TV).
siriushardware
Atari Super Hero
Atari Super Hero
Posts: 550
Joined: Thu Aug 21, 2014 7:55 pm
Location: UK

Re: Screen sometimes going momentarily black when using a RGB->SCART cable - An attempt of an explanation

Post by siriushardware »

Good write-up, thanks Czietz.

Tzok has a point about minimising volts drop where Hsync and Vsync have internal resistors, and I wonder if there are some variants of ST (where the Hsync and Vsync may come directly from TTL-level IC outputs) where the Hsync and Vsync outputs may not be able to drive loads as low as 75R.

I have an ST-VGA lead (allowing use of a VGA monitor as an ST Mono monitor substitute) and I have one PC monitor, (Medion brand), which it does not work with because the Medion's Vsync and Hsync inputs are 75R terminated and the ST (an STFM, I think) just can not drive those inputs because the monitor input resistances are too low. I have other VGA monitors where the Vsync and Hsync input impedances measure much higher than 75R and the same computer / lead have no problem driving those.

Incidentally one other approach is to modify the TV / Display internally by cutting the track going from SCART pin 16 and adding a small slide switch next to the SCART socket, arranged so that it either connects SCART pin 16 to the ongoing track (as it originally was) or connects the ongoing track to a strong power supply (via divider resistors) so that you can force RGB input mode just by moving the switch to the appropriate position. It may be worth doing this if you have a quite a few other retro computing devices which can output an RGB signal but don't supply an output voltage from which the force-RGB signal can be derived.
User avatar
GokMasE
Captain Atari
Captain Atari
Posts: 225
Joined: Sun Mar 02, 2003 11:16 pm
Location: Sweden
Contact:

Re: Screen sometimes going momentarily black when using a RGB->SCART cable - An attempt of an explanation

Post by GokMasE »

Thanks for the detailed explanation Christian.

Maybe this should be promotoed to FAQ status officially?
tzok wrote:In STe the HSync & VSync also have a 33R series resistor. Whats more, the diode introduces further voltage drop out. It should be Shottky diode to ensure minimal voltage drop. 1N4148 is not ideal (won't work with my STe & Sony TV).
I am waiting for the diode to try the capacitor/diode solution for the blanking pin. I will test the modified cable on the STe as well.

/J
tzok
Captain Atari
Captain Atari
Posts: 231
Joined: Fri Jun 30, 2017 7:22 pm
Location: Bielsko-Biala, PL
Contact:

Re: Screen sometimes going momentarily black when using a RGB->SCART cable - An attempt of an explanation

Post by tzok »

siriushardware wrote:I wonder if there are some variants of ST (where the Hsync and Vsync may come directly from TTL-level IC outputs) where the Hsync and Vsync outputs may not be able to drive loads as low as 75R.
Actually these serial resistors are present only in STe, in ST/f/fm H/VSync are driven directly.
siriushardware
Atari Super Hero
Atari Super Hero
Posts: 550
Joined: Thu Aug 21, 2014 7:55 pm
Location: UK

Re: Screen sometimes going momentarily black when using a RGB->SCART cable - An attempt of an explanation

Post by siriushardware »

Indeed, but in the STFM at least, Hsync and Vsync appear to come directly out of the GLUE ic, not via any kind of transistor buffer stage. I have no idea what the drive capability of those outputs actually is. 75R is a low impedance to ask any digital IC to drive directly.
siriushardware
Atari Super Hero
Atari Super Hero
Posts: 550
Joined: Thu Aug 21, 2014 7:55 pm
Location: UK

Re: Screen sometimes going momentarily black when using a RGB->SCART cable - An attempt of an explanation

Post by siriushardware »

Just a quick update: On impulse, I've bought an oldish 19-inch LCD TV because, unlike any modern small TV on sale, it had every type of video input known to man including 2 * SCART and 2 * HDMI, and PC-VGA so quite a versatile and useful little display, and with the right video leads it can potentially work as both an ST colour monitor and an ST mono monitor.

However as with many sets of that era it does not have a force-RGB mode option in the menus, so I will have to try something like the above suggestion to get the set to switch to RGB mode.

It happens to be a Sony, a KDL-19L4000. I measured the input resistance from SCART pin 16 down to 0V and certainly on this set at least, pin 16 has an input resistance of 75R so it will need quite a robust voltage source to drive it.

It will be interesting to see if GokMasE can get RGB going on his Sony display by using VSYNC + diode + capacitor.
User avatar
GokMasE
Captain Atari
Captain Atari
Posts: 225
Joined: Sun Mar 02, 2003 11:16 pm
Location: Sweden
Contact:

Re: Screen sometimes going momentarily black when using a RGB->SCART cable - An attempt of an explanation

Post by GokMasE »

Quick update: I did try it on my TV (not a Sony though) and while the previous "bouncing" was completely gone, I instead got slight disturbance/flickering inside the picture every 2-3 seconds.
That is, the picture frame does not move, but there a small cluster of flickering on screen, possibly wandering vertically. Picture itself is crisp and clear apart from the occasional disturbance/flicker.

I might add that I did not use the exact diode from the pic, but a schottky diode: 1N6263 DO-35 60V 15mA.
The capacitor was 10uF/25v though.

I got the same result when testing with MegaST/STf/STe.

One idea might be to resolder the wire and connect GND to more pins on the Scart side, I just have a single one ATM.
User avatar
GokMasE
Captain Atari
Captain Atari
Posts: 225
Joined: Sun Mar 02, 2003 11:16 pm
Location: Sweden
Contact:

Re: Screen sometimes going momentarily black when using a RGB->SCART cable - An attempt of an explanation

Post by GokMasE »

Quite update #2:

Connecting more of the GND pins on the SCART connector did not change anything.
The flickering is seemingly within individual scan lines, so perhaps the switching time of the diode is a factor here?

I will replace the Schottky diode with the suggested 1N4148 diode next.
siriushardware
Atari Super Hero
Atari Super Hero
Posts: 550
Joined: Thu Aug 21, 2014 7:55 pm
Location: UK

Re: Screen sometimes going momentarily black when using a RGB->SCART cable - An attempt of an explanation

Post by siriushardware »

Logic says that the Schottky diode (with a lower volts drop - 0.1V as against 0.6V for a normal silicon diode like the 1N4148) is actually the better choice, but, as I said in your earlier thread, whatever works, works.

Later in the week I will try forcing RGB mode on my 'new' Sony by connecting two charged NiMH batteries (therefore 2.4V) between GND and pin 16, this is probably the most uncomplicated way in which to generate a force-RGB input on SCART pin 16 for test purposes. Also temporary, because the batteries will not stay charged for long driving a 75R load.

If that gives me a rock steady RGB image I will then retrospectively try the vsync + diode + capacitor method to see if that works just the same for me or makes things worse in any way.
User avatar
GokMasE
Captain Atari
Captain Atari
Posts: 225
Joined: Sun Mar 02, 2003 11:16 pm
Location: Sweden
Contact:

Re: Screen sometimes going momentarily black when using a RGB->SCART cable - An attempt of an explanation

Post by GokMasE »

siriushardware wrote:Logic says that the Schottky diode (with a lower volts drop - 0.1V as against 0.6V for a normal silicon diode like the 1N4148) is actually the better choice
Ok, noted.

Then the only thing left to play with is the capacitor. In the versions of SCART RGB => ST cable that was suggested by Ian Stedman in an older thread, he suggested 10-47uF for the capacitor.
Maybe if you have a longer scart cable (mine is a couple of meters at least), it might call for a higher value of the capacitor perhaps? I will see if I can locate a 47uF capacitor and see if that makes a difference.
siriushardware
Atari Super Hero
Atari Super Hero
Posts: 550
Joined: Thu Aug 21, 2014 7:55 pm
Location: UK

Re: Screen sometimes going momentarily black when using a RGB->SCART cable - An attempt of an explanation

Post by siriushardware »

Do you have a meter?

If so try measuring the voltage your setup is supplying to SCART pin 16. Because vsync is high (+5V) nearly all of the time and only low for a very small percentage of the time, the average voltage coming out from vsync (when not significantly loaded) is nearly 5V, and, assuming the drive out from vsync is intended to drive a 75R sync input, it should be not much less than 5V by the time it has gone through the schottky diode (-0.1V) and then been 'averaged' by the capacitor.

But according to the SCART spec we should really be aiming for between 1.5V to 3.0V. Hence my question, what voltage is your present arrangement actually sending to SCART pin 16?

If anyone has an original ST colour monitor such as an SC1224, I would be interested to know what the input resistance is from the vysnc and hsync pins on the monitor plug down to GND - 75R, or some higher value?

Edit: It occurred to me that I could find this out by looking at the schematic for the SC1224. I understand there may be more than one version of the SC1224 and I only looked at the schematic for one, but on that version the input terminating resistances for Hsync and Vsync are 3.3K, which is a much more reasonable load for outputs from a digital IC (ie, Glue) to be expected to drive.

I then looked at the circuit for the SM124 (MT-21 version). On that circuit Vsync has an input resistance of 3.6K and Hsync has an input resistance of 100R + 13K.

This reinforces my suspicion that the Hsync and Vsync outputs from the ST's Glue IC were probably not designed to drive monitor sync inputs with an input resistance as low as 75R. The low value resistors in series with the sync outputs on the STe may have been added to protect the sync outputs in the event that anyone did connect them to monitor sync inputs with very low resistances.
Post Reply

Return to “Hardware”