Atari Jaguar csync TTL?

[FBX]

So I got in a loaner Jaguar with all sorts of RGB cables: sync on composite, csync, and sync on luma. The maker of the sync on luma cable claims the csync coming out of the jaguar is TTL level. Indeed when I do some Googling online, I find this document which states the same thing:

http://old.pinouts.ru/Video/AtariJaguarAV_pinout.shtml

So it sounds like it does output TTL csync, right? Well I opened the csync RGB cable from https://www.retrogamingcables.co.uk/ and determined there's no attenuation of the csync on their PCB board in the SCART head. The trace literally goes from csync pad to the sync pin with no visible components to attenuate the signal. Could they possibly have put the components in the console side of the cable, or did they simply forgot, or does the Jaguar not output TTL sync?

Very curious about this...

[FBX]

Update: just checked the retro gaming cable with my multimeter and there's ZERO resistance from console-side csync pin to SCART pin in the cable. So either they missed this, or the documents about Atari csync being TTL are wrong?

Edit: Here's yet another scanned page from a technical manual:



So yeah, if you have an RGB csync cable for the Jaguar from retro gaming cables, better check that csync!

[maxtherabbit]

it has hsync and csync but no vsync? that's odd

[Bratwurst]

Csync is technically the vsync line. I will try to sit down and measure the output sometime today to verify that it's 5 volts but I'm pretty sure it is and just a misstep on retrogamingcables.co.uk's part. I wasn't impressed when I dissected one of their cables sometime back.

[FBX]

I contacted retro gaming cables, and heard back from a "Robert Fletcher". He claims it does not output TTL csync, but doesn't have his notes on hand to make sure. He said he will double check and get back to me.

Edit: I found an online Atari schematic for the Jaguar that suggests the csync pin is actually "VSL" and the only thing between it and the "Tom" processor is an FB inductor "L13". So apparently vertical sync on the Jaguar also doubles as csync? I guess there's maybe a setting in the Jaguar to set this pin to vertical or csync?

[FBX]

First report came in from a user that measured indeed 5V TTL on pin 5B using his scope. So yeah, we need to get this sorted asap for csync RGB cable users...

Edit: Person wasn't measuring under load, so we still need someone with a scope and proper load testing to get an answer on this.

[FBX]

Found the entire original technical reference document that was written by Atari corporation in 1995:

http://atariage.com/forums/index.php?ap ... _id=369078

Page 8 has the information I showed in the earlier image.

However, another technical manual written in 2001 independently has this to say about pin 32 of the "Tom" chip: 2mA output/TTL input.

So I don't know what to make of that.

[FBX]

Got a reply from someone on the Atariage forums:

Yes, it is a logic level 2 mA output from Tom, without any buffering. So it's fine to drive a high-impedance sync input (a few kΩ or higher) directly, but not for a 75Ω input.
In the latter case, you should use a series 1200Ω resistor to drop the level to the standard 300 mVpp. But I wouldn't recommend it, since without a buffer it's not suited to drive a cable...

So that kind of shocks me (pun not intended) that it would need 1.2k Ohms instead of the typical 470 Ohms. Also he suggests there's no 'buffer', so I'm curious what would be needed to build one in the cable?

The mystery continues!

[Syntax]

Where are they getting 300mv pp??
I thought the standard sat around 600mv pp??

You can make a buffer by copying this design, Ive done it for heaps of my systems now.
http://www.raphnet.net/electronique/dre ... dricas.gif
Supply's clean strong TTL and keeps every safe. Put a 470R on the end if you want around 600mv pp sync.
Available chips are much smaller and a different pinout.
Be sure to ground all unused inputs.

[MKL]

Composite video is 1Vpp but the sync part of the signal is 0.3Vpp so if you're sending a composite sync signal to a composite video input, that's the voltage level you want to have and assuming the source is actually 5V, a 1.2K (1175 ohm) resistor is in fact required in series with it.

[Syntax]

Well you better let Tim Borti Voultar ect know this because all there boards output 600mv pp 75R Csync...

Seriously though, 470R is all I require from a buffer or from an Extron to get 600mv pp.
1.2k seems like overkill and I bet causes issues.

[borti4938]

I’ve designed my boards to output 300mV@75load on CSync. However, everything in between of 1V and 200mV should work fine on virtually all 75ohm sync inputs.

Got a reply from someone on the Atariage forums:

Yes, it is a logic level 2 mA output from Tom, without any buffering. So it's fine to drive a high-impedance sync input (a few kΩ or higher) directly, but not for a 75Ω input.
In the latter case, you should use a series 1200Ω resistor to drop the level to the standard 300 mVpp. But I wouldn't recommend it, since without a buffer it's not suited to drive a cable...

So that kind of shocks me (pun not intended) that it would need 1.2k Ohms instead of the typical 470 Ohms. Also he suggests there's no 'buffer', so I'm curious what would be needed to build one in the cable?

The mystery continues!

You may want to take a look into the TSH122ICT, which is signal channel video buffer with +6dB gain. You just need a handful of components around it: two resistors (e.g. 9.53kohm and 620ohm) for signal attenuation at the input, two decoupling caps on the power rail (10uF and 0.01uF) and output components (22uF, 33uF and 75ohm resistor). The IC is relatively small and does not need that much space. Place the buffer at the console end of the console!

[FBX]

So the basic gist of this all is: Jaguar RGB csync cables that just pass VSL straight to the SCART head pin are doing it completely wrong, correct?

[borti4938]

If "it is a logic level 2 mA output from Tom, without any buffering" is a correct fact, than yes: passing it directly through a cable into a TV (or other 75ohm terminated input) is copletely wrong.

[FBX]

Robert of retro gaming cables reported back with this to say:

My engineer Charlie measured 700mv peak to peak under load however the output is a high impedance which doesn’t match 75 ohm so even though the voltage is with the tolerance the impedance doesn’t match. I was unaware of this when we first designed this cable CSYNC cable.

So this could be solved with a buffer or an lm1881 attached to the sync over composite. We will redesign our CSYNC version.


-FBX

[Johnpv]

What does this mean for those of us with the current Csync cables? Any chance of Robert coming in here and commenting?

[FBX]

What does this mean for those of us with the current Csync cables? Any chance of Robert coming in here and commenting?

It means you've permanently damaged whatever equipment you used them on. Throw out those devices and star over.

On a more serious note, the experts can correct me if I'm wrong, but I believe the impedance mis-match is less 'dangerous' than a voltage one. Not sure though.

[Johnpv]

What does this mean for those of us with the current Csync cables? Any chance of Robert coming in here and commenting?

It means you've permanently damaged whatever equipment you used them on. Throw out those devices and star over.

On a more serious note, the experts can correct me if I'm wrong, but I believe the impedance mis-match is less 'dangerous' than a voltage one. Not sure though.

*throws all my av equipment in the trash*

I was kind of more wondering if RGC would end up replacing or fixing the cables of those of us with the old version, or if we'd have to buy new cables.

[maxtherabbit]

What does this mean for those of us with the current Csync cables? Any chance of Robert coming in here and commenting?

It means you've permanently damaged whatever equipment you used them on. Throw out those devices and star over.

On a more serious note, the experts can correct me if I'm wrong, but I believe the impedance mis-match is less 'dangerous' than a voltage one. Not sure though.

I wouldn't call an impedance mismatch dangerous, but it would mean more signal degradation over longer runs

[retro_gaming_cables]

If anybody wants to return their Jaguar cable for modification please return to the following address and quote your name and email address and order number if you can locate this.

Retro Gaming Cables Limited
151-153 Wincolmlee
Hull
East Yorkshire
HU2 0HD
United Kingdom

Unfortunately we cannot rewire the Jag cables for sync over luma, as the edge connectors are over-moulded, however, we can rewire it for either sync over composite video, or if you want to maintain a CSYNC signal we can rewire as follows.

11B (Composite video) --> LM1881 --> 470Ω --> SCART pin 20.

The customer will need to pay for sending the cable back to us. However we will carry out this work for free and will not charge for sending the cable back to you.

Rob

[FBX]

Ste from HD Retrovision was kind enough to run scope and spice tests on the Jaguar pin 5B, his results suggest all you need is a series 680 Ohm Resistor and series 220uF capacitor. No need for logic buffers as this should be plenty sufficient.

[Johnpv]

I'll probably end up just waiting till the new version of the cables are available and order a new one.

[maxtherabbit]

Ste from HD Retrovision was kind enough to run scope and spice tests on the Jaguar pin 5B, his results suggest all you need is a series 680 Ohm Resistor and series 220uF capacitor. No need for logic buffers as this should be plenty sufficient.

coupling capacitor + attenuation resistor definitely sounds like a better solution than sync stripping the CVBS

[copy]

Just saw this on retro-access.com, if anyone was wondering about their Jag cables:

Additional note: We have received some questions recently about the components used on the c-sync line. We have always wired these cables with a 470 ohm resistor in the Jaguar end and a 220 uf capacitor in the SCART end.

[strygo]

I've had a Retro Access cable since they introduced them and it works great.

[FBX]

I've had a Retro Access cable since they introduced them and it works great.

It's because 470 is MUCH better than nothing, however, 680 is optimal. I had sent Rachel this information, but she didn't seem to trust me on it.

[copy]

Sounds like the Retro Access sync-on-luma cable is the best and safest option currently available from either seller.

[FBX]

Sounds like the Retro Access sync-on-luma cable is the best and safest option currently available from either seller.

We talked some more and she said she's started using 680 Ohms now. Although to be fair, 470 was likely 'safe enough'. It's just that 680 is optimal.

[copy]

We talked some more and she said she's started using 680 Ohms now. Although to be fair, 470 was likely 'safe enough'. It's just that 680 is optimal.

Nice!

[radiokid]

Hi, sorry to resurrect this thread but I am in a sorry position that something has caused my Gscartsw_lite to die, it would appear that my Altera IC is beyond repair (been communicating with superg to try and get to the bottom of the issue).

All my cables are from retrogaming cables, apart from one which is for my NESRGB (just a straight through din to scart).

I have a Jaguar and one of the original Jag scart cables from retrogamingcables. I am trying to work out what killed the Gscartsw. Could the issues discussed in this thread have been a contributory factor?