Woes of Amiga RGB to SCART

[Zetr0]

Hello there my forum chummies,

I have been spending the day, catching up on some projects like helping out our forum friend, Darkwave.

now....

As you know, there has been discussion on the correct or better method of connecting an Amiga to a SCART equipped TV...

infact there are as many methods as there is pins... Anyway a very looooOOooong story short. I have spent the entire day, (as of 10am) building a compatible RGB scart lead for my A600 / A1200 and my 42" plasma tv...

now the tv has 3 RGB scart inputs, all three were tested... so whats the problem i hear you gasp... well after 12 hours, 8 revisions, 4 specifications... not one of them work.... nadda... zip... nothing.....

all i get is a screen switch a report of it going to 50hz PAL and a white screen OR a screen with green / black wavey lines depending on which input is used...

so... heres the 3 main specifications i have used....

type1


type2


type3



I have tried variants of each, mainly with common grounding of non-used lines etc. still the same result...

the Amiga(s) A1200/A600;

works perfectly with the 23pin CBM cable to 9pin DE9 connector of the 1084s
works perfectly with the Composite output to the TV and the 1084s


If anyone has any ideas as to what i have managed to do wrong i would appreciate it... as i am sure would Darkwave LOL

[Merlin]

@ Zetr0

I think you are missing the RGB blanking signal from the cable. What do you have to lose by trying it? I'm pretty sure that's the problem.

Have a look at this page, it could be you need pin 20 or pin 16 connected:-

http://www.idiots.org.uk/vga_rgb_scart/index.html

It could also be pin 16, see below:

Blanking and switching
Two pins provide switching signals.

Pin 8, the switch signal pin, carries a low frequency (less than 50 Hz) signal from the source that indicates the type of video present.

0 V–2 V means no signal, or internal bypass
4.5 V–7 V (nominal 6 V) means a widescreen (16:9) signal
9.5 V–12 V (nominal 12 V) means a normal (4:3) signal

Pin 16, the blanking signal pin, carries a signal from the source that indicates that the signal is either RGB or composite.

0 V–0.4 V means composite.
1 V–3 V (nominal 1 V) means RGB only.
The original specification defined pin 16 as a high frequency (up to 3 MHz) signal that blanked the composite video. The RGB inputs were always active and the signal 'punches holes' in the composite video. This could be used to overlay subtitles from an external Teletext decoder.

0 V–0.4 V means composite with a transparent RGB overlay.
1 V–3 V (nominal 1 V) RGB only.


Good hunting.

[Zetr0]

@Merlicious

Thanks my Old Skool friend, another attempt is now on the go....

[Merlin]

@ Zetr0

I reckon 12v to pin 8 and 1v to pin 16 (5v with an 100R resistor) may solve it, or 12v to pin 8 and composite video to pin 20 could be an alternative.

[TheCorfiot]

@Merlin

I couldn't have been more thorough.. Spot On

Good luck Z man,
Out of Curiosity do you have access to an old analogue TV you can try the cable on as a cross reference.

Merlin's description is exactly how I wire my scarts up too

Let us know my Good friend & the Best of British

TC

[Boot_WB]

@Zetro,

Must be in the air my friend, I've also been catching up on a few projects. I'd been intending to change the oil & oil filter on my car, but the crap weather put paid to that idea.

Repaired my A4K keyboard (thanks to the insert you kindly provided) - all working perfectly now!

Also reinstalled my Deneb, and gave it a nice 1GB flashdisk to play with - gets around 5MB/s transfer speed between the SCSI hard disk on the CSPPC and the flashdisk on the Deneb! Will be updating Poseidon and Luciferin tomorrow if I find the time.

Received a faulty Cyberstorm mk3, so put that through its paces. No boot, no screen flashing, Caps-lock froze after around 6 presses - It's dead Jim!
Gave it a gentle clean, changed the RAM, checked the jumper settings and swapped the CPU for my spare 68060 - ping! Booted up like a charm, no instability or problems!

Some days it all just falls into place, it's just that these days are rare.

[Boot_WB]

Zetro my friend, I drew you a nice RGB-SCART layout recently which should work like a charm: This thread, post #5.

Regards


Rich

[Zetr0]

@Merlcious

thanks for the call of support my friend ... lucky i still have hair left.... i think...

@Boot_WB

Glad those inserts got there and more so glad you got that keyboard sorted , indeed i followed your schematic to the wire... still nothing.... (except green wavey lines on SCART 2)

@TC

my friend, I appreciate your support my daughter has a 21" tv in her room... that has scart.... may have a check on that me thinks

@ALL

you guys are awesome!

heres the latest schematic i tried (its virtually identical to Rich's schema)



still no go I am afraid to say... but something peculiar was noticed....

Firstly SCART 2 now only displays a white screen, when connected (usually only displays green wavey lines)

Secondly when switching from SCART 3 to SCART3S (input source via the tv menu) for a brief moment there seems to be a subtle shadow of what should be workbench.... (two disk icons...) but its a colour shadow that disappears within a blink....

Gonna call it a night at 1am LOL... I shall try another TV and perhaps the VCR as an input as well.. fingers crossed eh?

[Boot_WB]

@Zetro

Have you checked continuity to pin 20?

When I was creating mine from a SCART - SCART cable, I did not realise at first that pin 20 at one end of the cable connects to pin 19 (CVBS out) on the other, and vice versa.

[Zetr0]

oh yes.... i got caught on that one too.... that was around 6pm LOL.....

me and Mr multimeter have been having a wonderful romance today... quadruple checking ones work... if not more so...

I really appreciate you thinking our for me though... but I honestly believe now that the TV is Amiga RGB intolerant! LOL strange as it works on both the XBox RGB AV, the Playstation RGB AV and the DC RGB AV... i believe the VCR to TV is composite... but it has a couple of SCART inputs so in the morning i shall try the latest revised revision

[alexh]

Have you checked to see if the Grounds (return?) on the VGA connector are even wired up?

Most modern TV's sample the inputs and without good grounds I *think* the sample circuits have no reference point.

In one of my Amiga SCART cables there are resistors on all the RGB inputs!?

No point trying a VCR for anything other than pass through. Most (all?) VCR's are composite only.

[Merlin]

@ Alex

"What value are the resistors on the RGB lines, are they 47 or 75 Ohm?", he said, asking a leading question....

@ Zetr0

Try a few different values of Resistors to pin 16, it sounds like you are close, it may just be the voltage to pin 16 it doesn't like. If the voltage isn't right the blanking doesn't work and you get a white screen. The brief flash of image is probably as the voltage ramps up to pin 16.

[Boot_WB]

@Merlin,

Why guess when you can calculate?

Just measure the resistance between pin 16 and ground on the TV. The minimum/maximum value of resistance required to step down 5V to the required voltage for pin 16 (1V-3V iirc) can then be calculated.

Regards


Rich

[Merlin]

Yeah, a fair point well made. I suggested to Zetr0 that he temporarily fit a variable resistor inline to pin 16 to see if varying the voltage has any effect. You state 1 to 3v and I agree, however, we don't know if the TV wants 3v or 1v. I also suggested he check one of his other RGB leads (i.e. the Xbox one) to see what voltage was going to pin 16.

[Boot_WB]

I suggested to Zetr0 that he temporarily fit a variable resistor inline to pin 16 to see if varying the voltage has any effect.

Nice idea. :-)

@Zetro, have you also checked that the remaining (unconnected) pins have not shorted. If you have just cut the excess from the extra wires at the Amiga end, then there may be shorts between them, or to ground.

I'm sure you wouldn't make such a mistake, but this problem appears to defy reason - worth pointing out the obvious just in case.

[darkwave]

Firstly, I just wanted to say a big thank you to Zetr0 for working on these RGB cables for me. You've completely exceeded all sane levels of effort for this matter and I sincerely think you may have a hardware pr0n problem!

It seems as though Plasmas and probably LCDs can be a lot more picky about what they display than our old CRTs. I've had some problems in the past with some of my Megadrives which have completely baffled me.

I have three different megadrives; a Megadrive 2, a MultiMega and a Nomad. Each of these use the same RGB scart to connect to the TV.

The Megadrive 2 works with the CRT.
The MultiMega works only with the CRT.
The Nomad works with both the Plasma and the CRT.

Until I tried it with the Nomad I thought I had a dodgy cable. The only thing that sets the Nomad apart from the other two is that it's a US machine.

The worst thing is that you know it can output the signal. It's in that bunch of wires somewhere.

Anyway, that doesn't really get us anywhere, but I thank you Zetr0, and share your pain!

[Zetr0]

@Darkwave


I appreciate your praise sir, however i was never sane

The good news is guys, I think i have cracked it.... (stay tunned) forgive the pun

[Boot_WB]

Originally Posted by Zetr0

The good news is guys, I think i have cracked it.... (stay tunned) forgive the pun

Excellent news. What was the problem then mate?

[Zetr0]

Okay here is an update

(kinda like live aid this!!! LOL)

right... where to start... hmmmm okay... here we go....

yesterday my Amigo r0jaws popped over for some amiga geekyness and had some retr0licous Xbox emu fun ahh (not going to mention the gatling gun if you dont mention that rocket launcher thingy m8!! )

anyway... popped into the office to get components to build an Amiga RGB to 9 PIN CM8833 RGB input... this is easy, completely passive cable R,G,B,Gnd,Hsync,Vsync job done. anyway moving stuff about on the desk from the umpteenth failed SCART attempt i picked up the packed of resistors... and low and behold.... i had a FACE PALM moment...

on the back of the pack of resitors that i was POSITIVE were infact 110ohm... turn out to be 110k ohm.... yes.... that means less than a 10th of the voltage was getting thorugh.... indeed.... a moment.... and I accept all the mocking now!!!!!

So later this afternoon i decided ok... get back on and sort this puppy out for sure!!

so... removed the 110k and replaced it with an 80ohm as that was handy (yes... i checked the bands!!!) still no go on any scart input... although the SCART3S was flickering with a in-out colour phased ghost like image of workbench.

so tested the output volts

Pin 16 <input 05v> resisted by <80ohms> <output 5.12volts> !??!?
pin 08 <input 12v> no resistance <output = 9.3volts> !?

okay... so removed out the 80ohm and plumbed in a 100ohm (*yes checked the bands*)

pin 16 <input = 05v> resisted by <100ohm> <output = 5.32volts> !?
pin 08 <input = 12v> no resistance <output = 12.32 volts> !?

so.... there we have it.... a little further and still a little beat... I WONT GIVE UP!!!

i popped open the XBox AV to SCART lead and I am now in the middle of doing its schematic... to put on here for further discussion...

Alas i have not had the time to test on another SCART enabled TV/Device yet.. (which I will do before i re-build any cable!)

[Boot_WB]

@Zetro

I assume that you just measured voltage from Pin16 to ground, without the TV attached. If this is the case then the Pin16 voltage readings are not surprising (ignoring the minor variance), as the multimeter will have a very high resistance (eg 100 kOhm) when measuring PD.


(Figure 1)


wrt figure 1 - where:
R1 = the resistor installed in the circuit = 100 Ohm
R2 = the multimeter = 100000 Ohm

Current between input point and ground is given by:

= 5/(100 + 100000) = (almost) 5e-5 A

Voltage across R1 = I x R1 = 100 x 5e-5 = 5e-3 V
Voltage across R2 = I x R2 = 100000 x 5e-5 = (almost) 5V



wrt figure 1 - where:
R1 = No resistor installed in the circuit = 0 Ohm
R2 = the multimeter = 100000 Ohm

= 5/(100000) = 5e-5 A

Voltage across R2 = I x R2 = 100000 x 5e-5 = 5V


-----------------------------------------------


By comparison, the resistance across Pin16 - Ground on the TV is very small (around 100 Ohms).

wrt figure 1 - where:
R1 = the resistor installed in the circuit = 100 Ohm
R2 = TV Pin16 to Ground resistance = 100 Ohm

= 5/(100 + 100) = (almost) 2.5e-2 A

Voltage across R1 = I x R1 = 100 x 2.5e-2 = 2.5 V
Voltage across R2 = I x R2 = 100 x 2.5e-2 = 2.5 V



wrt figure 1 - where:
R1 = No resistor installed in the circuit = 0 Ohm
R2 = TV Pin16 to Ground resistance = 100 Ohm

= 5/(100) = 5e-2 A

Voltage across R2 = I x R2 = 100 x 5e-2 = 5V


---------------------------------------------


Consequently, I'd recommend that you test the resistance between pin 16 and ground on the TV - this will then allow you to calculate the required value of the resistor to install in position R1.

When measuring the voltage on Pin16, this can only be done in a meaningful way with the load (ie the TV) attached, and the multimeter in parallel.

Can't explain your strange variance in readings on pin8 though :-)

Regards





Rich