1084-S RGB trouble with Neo Geo/Megadrive. MSX ok

[carlosemandrade]

1. sorry about my english skills 

When i try to connect directly all wires from neo or megadrive (r,g,b,gnd and csync):

- neo geo always displays a distorted (ondulated on 1/3 top, correct on 2/3 botton) picture. Depending on what is on screen, sync goes (all black with horizontal thin lines)    color & sharpness (when viewed) is very good.

- megadrive (japanese, first model) displays a more stable picture than neo, but has some distortions... very good color & sharpness too

MSX - All good, not a single proglem.

Some users said to use a resistor between sync line (when using neo or mega). What value should i try ? Has anybody experienced same issues ?

[MKL]

There are some neo geo revisions that don't have sync out on the A/V port:

http://gamesx.com/wiki/doku.php?id=av:neogeoav

If this is your problem you can mod it or build a circuit with a sync separator.

[carlosemandrade]

hmmm. it seems not to be the 2 last revisions. Serial of mine is around 100.000

I can see the picture (used KOF 94 for tests). The intro animation displays out of sync (thin lines) or a ondulated picture (1/3 of top screen). in-game picture is a bit less ondulated.

Using composite video, no problems with sync ( but composite.... sux )

[MKL]

100xxx may well be a 3-5 revision since I know for a fact that 116xxx is. If you don't want to open the console (but why not?) you could grab a multimeter and check if you have continuity across pin 3 and 7: if you do then you'll know both pins are composite video and therefore 7 is not sync.

[carlosemandrade]

MKL, you're right about neo aes!

75 ohm resistor on the "sync" line : nada

sync and composite video pins generates exactly the same signal (and both have continuity on multimeter test that you explained).. so i ned a sync separator (LM1881) 

[MKL]

so i ned a sync separator (LM1881) 

The better method would be to restore sync on the A/V port. It's a very easy mod, just one wire soldered on the PCB. If you want I can show you how to do it.

[carlosemandrade]

yeah 

you can show details here or send to my email (carlosemandrade@gmail.com)

thanks 

[carlosemandrade]

SOLVED ! for neo geo 

I built the separator with 1881 and voilá. AES in glorious RGB 

[MKL]

Good stuff!

In case you want to restore sync (so you can use standard cables) all you need to do is desolder the blue thing (ferrite) shown here:

and solder a wire like the yellow one here (ignore the other things):

[carlosemandrade]

hmmm. it's pretty easy ! i'll try later !

[carlosemandrade]

UPDATE. Same cable with 1881 also works with megadrive (amazing picture!) 

[carlosemandrade]

UPDATE 2.  Neo AES: Did the csync modification, no more 1881, works great 

will the same work on megadrive ? (wire csync out directly on cxa like on neo aes)

Thanks MKL   

[MKL]

The Megadrive is like the earlier AES, it already has csync out on the A/V port.

[carlosemandrade]

quite strange here... using csync direct from my old jap. megadrive a/v out causes image instability on some parts of screen, but using csync with 1881 from composite, no problem   

[albino_vulpix]

quite strange here... using csync direct from my old jap. megadrive a/v out causes image instability on some parts of screen, but using csync with 1881 from composite, no problem   

Try sticking a 100uF capacitor in series with the sync line.

[MKL]

quite strange here... using csync direct from my old jap. megadrive a/v out causes image instability on some parts of screen, but using csync with 1881 from composite, no problem   

The Megadrive RGB and Sync outputs lack the RC filtering (470uF or 220uF + 75ohm) recommended for the Sony CXA1145P video encoder:



Only the Composite output has it (220uF + 75ohm). The Neo Geo already has these components for every output (minus for Sync on those revisions that don't have it, like yours):



You can add the RC filter for Sync inside the system (easier than on the cable, especially if it's a DIN-to-DIN).

1) Cut the trace in between the yellow spots:



2) Solder cap and resistor as shown here:



This should fix the problem. What's puzzling though is that your Neo Geo didn't have Sync out and you ran it directly to the A/V port without any filtering and it worked fine. Perhaps the difference is that you tapped it from the input pin (10) of the encoder, whereas if you tap it from the output (pin 11) it needs filtering. Well, try both ways and find out.

[viletim]

MKL,

My Megadrive II has the C-Sync signal from VDP going to both the video encoder and A/V port. The C-Sync output of the video encoder is unused. The signal from the VDP is a high impedance CMOS type and if connected to a TTL input (like a monitor C-Sync input) will be excessively loaded. This results in the video encoder not getting any C-Sync and no video comes out.

I recall it being the same with the original models but I can't really remember and don't have one to check.

In any case, no TTL sync signal needs capacitors and other junk in the signal path. That's guarenteed to make it unreliable!

[MKL]

No, the Megadrive I is not like that. Sync (which is tied high through a pullup resistor) goes from pin 42 of the VDP straight to pin 10 of the encoder and from pin 11 straight to the A/V port:

[viletim]

MKL,

I had the opportunity to check an original PAL Megadrive this evening. It has the sync pin of the AV port connected to pin 10 of the CXA1145 and pin 11 is not connected to anything. Are you absolutely sure yours isn't the same?

The datasheet contains an application circuit with  a resistor and capacitor in series with this pin 11 composite sync output. This suggests that it's not a TTL sync output at all. The Japanese data is meaningless to me but I had a look at the datasheet of a clone IC, the ES71145. If this is anything to go by this pin is basically a composite video output without the video. The output is listed as 0.29Vpp (typical).

[MKL]

MKL,

I had the opportunity to check an original PAL Megadrive this evening. It has the sync pin of the AV port connected to pin 10 of the CXA1145 and pin 11 is not connected to anything. Are you absolutely sure yours isn't the same?

The pics I posted are of an early Japanese model I had a couple of years ago but I don't have any MD to check right now and the trace running under the encoder makes it impossible to see from the pic if it goes to pin 10 or 11 but if you checked that it's connected to pin 10 then I think it'll be the same on any model regardless of region, so the pic is probably to be edited like this:

The datasheet contains an application circuit with  a resistor and capacitor in series with this pin 11 composite sync output. This suggests that it's not a TTL sync output at all. The Japanese data is meaningless to me but I had a look at the datasheet of a clone IC, the ES71145. If this is anything to go by this pin is basically a composite video output without the video. The output is listed as 0.29Vpp (typical).

My suggestion to put the cap and resistor on the Sync output is based on the info in the datasheet (the pic I posted above comes from that) and the fact that the Neo Geo, which has the same encoder, does have these components.

[hellbelly]

Just a quick thank you, thanks to the info in this thread I was able to get my PAL (modified) MD to work on my Samsung LCD after giving up some time ago!

The console has a CXA1145P, when tested with a continuity tester it shown that leg10 was connected to pin 7 of the AV plug, but leg10 is Csync In?  anyway, cut the trace leading upto pin 7 on the AV plug, and soldered at 220uF cap and 75ohm resistor from leg11 to pin7, rewired my RGB lead to use csync instead of comp video and it works brilliantly.

Many thanks
Pete

[marquisor]

FOR MEGA DRIVE 1:

what i did here is i cut the trace of pin 7 and pin 3 (composite video) on the PCB near the RGB connector.


now a very strange issue. it got synced on the 1084S-D1 pin 8 + 9 (H-SYNC + V-SYNC together).
after about one minute the screen fades out! first it gets a instable for some seconds, then it fades out, comes back, instable, fades out, comes back etc. seems like a capacitor issue?

i also tried to connect 75 Ohm + 220 µF to pin 11 of the CXA1145 to PIN 7 RGB OUT, NO sync! neither with H-Sync, V-Sync, both together or composite-sync soldered at 1084S plug.

EDIT1: why i cut the "composite video" trace is, because if trace connected the screen stays black. if cut above issue.. works for 1 minute appr. then fades out, comes back etc. strange...
any s-video chances for MD1?