Amplifying 75 ohm sync to TTL levels effectively?

[MikeIronFist]

I got an Extron VGA switcher (specifically a model labeled SW2 VGA DA2 A) and while it works great with my MiSTer in direct video mode (due to the 3V TTL sync coming from the VGA DAC) I cannot get my Genesis to even output a proper signal from the thing; I'm trying to make an adapter that will take the signal from a proper SCART cable and send it to the switcher with the right levels. Getting it to accept the sync signal is giving me trouble. (I know the Genesis natively outputs TTL Csync but I'm trying to avoid making a custom cable for it if possible, details on that later in the post.)

The manual says the switcher takes 1.7 volts minimum, 5 volts maximum TTL sync. It also has 510 ohms impedance on that input. It also says it can take positive or negative polarity sync. The manual specifically says it can handle 15 Khz RGBs, RGsB,and RsGsBs in addition to standard 31 Khz VGA/RGBHV, and it does do just fine with my MiSTer outputting 15 Khz RGB with Csync, as I mentioned.

The best I managed to do for adapting/amplifying the 75 ohm Csync from my Genesis SCART cable was a quick hack with an OpAmp (Lm324) as a buffer for the Sync signal so that it wasn't getting loaded back down to standard SCART levels or lower once it goes into the switcher's input. Power was drawn from the blanking pin on the SCART cable. However, the final result is shifted way to the left and discolored as any sync signal with significant delay would be expected to do. I've tried one other OpAmp (NE5534P) in hopes the higher slew rate might give me better results but with that one in the same configuration the sync doesn't even lock on, so I don't even get a discolored signal on my PVM 8042q, I get none at all.

Additional info: output devices I have tried include an RGB-modded Trinitron with a SCART input, an RGB-to-YPbPr transcoder into a RetroTink 2X and a PVM 8042q receiving the signals from a female SCART to male BNC adapter. Both my MiSTer and Genesis work fine with all of these when connected directly, but the MiSTer with its TTL-level sync direct from the DAC is the only one that works through the switcher. (yes, the cables coming from the outputs are attenuated to SCART levels using 470 ohm resistors for safe use with my equipment, if you were wondering.)

While I understand I could probably try to make a custom cable that takes the Genesis TTL Sync directly from the port, I would prefer completing my custom SCART-to-DSUB15 adapter for this so that I could potentially use other systems with SCART cables in the future, and maybe even a passive SCART switcher running into the adapter, since this switcher has two outputs and provides exciting options for streaming.

If you have any IC recommendations, better amp options, corrections on my understanding of how to handle sync, literally anything other than "just get another switcher," I'd love to hear it. Or I guess you could recommend me another cheap VGA switcher that handles audio and might be better for the job, I suppose, if you truly feel like I'm on a wild goose chase here. But I'm trying to keep things cheap and go the DIY route where possible and I've already got this switcher, so I'd love to see what could be done with what I have + whatever small components or ICs I might need to add.

[Sirotaca]

I was in a similar situation with a monitor that requires TTL sync, so I used an LM1881 that I had lying around. Works just fine with everything I've thrown at it, including my model 1 Genesis with a 75 ohm c-sync cable from RGC.

[MikeIronFist]

I was in a similar situation with a monitor that requires TTL sync, so I used an LM1881 that I had lying around. Works just fine with everything I've thrown at it, including my model 1 Genesis with a 75 ohm c-sync cable from RGC.

thanks for the recommendation! I'll be sure to give one of those a try. I see it's a popular sync separator, so I bet it might even work with PS1 and N64 sync-on-luma cables if I do this right? I strongly appreciate this.

[Sirotaca]

I see it's a popular sync separator, so I bet it might even work with PS1 and N64 sync-on-luma cables if I do this right?

Yeah, should work with pretty much everything.

[Syntax]

If you upgrade to an Extron VGA Matrix system then most of them reclock and amp any sync signal to 4.5v TTL.

[MikeIronFist]

If you upgrade to an Extron VGA Matrix system then most of them reclock and amp any sync signal to 4.5v TTL.

That sounds fantastic! Thanks for fhe recommendation, will definitely look into one as soon as I have more analog RGB signals to route, it sounds like a great option. Seems like their VGA matrix switchers strike a healthy balance between features, input/output options and space. I'm pleasantly surprised by how flexible they seem to be.

[maxtherabbit]

This may or may not be related, but the Extron VGA based switchers that use auto-switching seem to struggle with 15kHz inputs. I've observed the problem with both RGBHV and RGBS at 15kHz with TTL sync. The problem is the auto-switching logic will occasionally deselect the input momentarily. Turning off auto switching and manually selecting an input resolved it.

[matt]

I have an Extron VGA SW6/AR, it works fine with attenuated csync from most consoles. It does have problem with auto-switching mentioned by Maxtherabbit, which doesn't bother me.

Some Genesis revisions have weird sync outputs, however, and I've run into at least one that didn't work with my Extron switchers no matter what. I just used an LM1881 to pull sync from composite video which worked fine.

[maxtherabbit]

I have an Extron VGA SW6/AR, it works fine with attenuated csync from most consoles. It does have problem with auto-switching mentioned by Maxtherabbit, which doesn't bother me.

Some Genesis revisions have weird sync outputs, however, and I've run into at least one that didn't work with my Extron switchers no matter what. I just used an LM1881 to pull sync from composite video which worked fine.

That's why I buffer csync through a 74 series buffer on my genesis dongle. All genesis revs (NTSC) work with the extron fine after that

[MikeIronFist]

Excellent news! So, I tried Sirotaca's suggestion and finished a simple SCART-to-D-sub15 adapter with an inline LM1881 to handle the sync. Power is drawn from SCART pin 8. The end result is as good as I could hope; the Dsub15 side outputs TTL-level Csync, which the switcher has no issues with. You can check out the pic below for my main application for this setup; one output from the switcher goes to a CRT, the other goes to a transcoder and Retrotink 2X (later it's just going to go straight into a GBS-8200 with gbs-control)


Thanks everyone for the help and recommendations! This has been a great learning experience.

That's why I buffer csync through a 74 series buffer on my genesis dongle. All genesis revs (NTSC) work with the extron fine after that

I'm curious, what part number specifically? The LM1881 is great for my current needs but it wouldn't hurt to know about other useful ICs for these sorts of things


---

[Syntax]

Use pin 16 not 8.
8 can sometimes have 12v.

[Ryoandr]

Use pin 16 not 8.
8 can sometimes have 12v.

Pin 16 should not have more than 3v if wired correctly. You can't power 1881 with it.
If you can it means your cable is out of specs.

[maxtherabbit]

That's why I buffer csync through a 74 series buffer on my genesis dongle. All genesis revs (NTSC) work with the extron fine after that

I'm curious, what part number specifically? The LM1881 is great for my current needs but it wouldn't hurt to know about other useful ICs for these sorts of things


---

SN74LVC2G17DCKR

it's just a buffer, not a sync stripper

[Syntax]

Use pin 16 not 8.
8 can sometimes have 12v.

Pin 16 should not have more than 3v if wired correctly. You can't power 1881 with it.
If you can it means your cable is out of specs.

Pin 16 is usually 5v across a resistor.
This resistor when connected to a sets termination resistor drops the voltage.

If no termination resistor then you get 5v.

But yeah, power the stripper from 8, it's all good till you start throwing 12v sync pulses into your equipment.

[Ryoandr]

pin 16 is usually a resistor away from pin 8
if pin 8 is 12v (and pin 16 is not connected), you get 12v accross a resistor :P

[MikeIronFist]

Pin 16 is usually 5v across a resistor.
This resistor when connected to a sets termination resistor drops the voltage.

If no termination resistor then you get 5v.

But yeah, power the stripper from 8, it's all good till you start throwing 12v sync pulses into your equipment.

Ooh, fair point, I don't have anything that outputs 12v on that pin *right now* but it wouldn't be fun to plug something in some day having forgotten and then end up with a dead switcher, lol.

SN74LVC2G17DCKR
it's just a buffer, not a sync stripper

Thanks! Honestly I originally assumed I would have to use some sort of buffer in the first place, so if I have situations where the LM1881 isn't the best option it doesn't hurt to have simpler alternatives. I understand most people worry about Csync on a per-console basis and don't strip at just one input like I'm doing, haha

[MikeIronFist]

Okay, small problem. Turns out the voltage input gets loaded down when powered from pin 16, as there's resistance inside the IC that causes the SCART cable's resistor to act as a voltage divider. I tried soldering it up to pin 16 to test and the results were the IC basically couldn't stay on.


I took a look at the functional diagram on the data sheet and it shows the resistance right there.


Any recommendations or workarounds for this? I understand the crude solution would be to bridge IC pins 7 and 8 but my first instinct is that this would probably risk damaging the LM1881 if it worked at all. Would some sort of capacitor on IC pin 8/SCART pin 16 help me get around the voltage issues?

[Syntax]

if it all works for now dont worry too much about it, just keep an eye on Pal consoles and wire their cable accordingly.

[MikeIronFist]

if it all works for now dont worry too much about it, just keep an eye on Pal consoles and wire their cable accordingly.

Fair point, honestly just using SCART pin 8 but adding a little bit of label tape with "WARNING: USE ONLY WITH NTSC SEGA GENESIS" as a reminder would probably be the simplest way to go about things.

But I had a sudden bit of inspiration: why not attenuate the LM1881's sync output? That way, even if the voltage on SCART pin 8 were 12v, the final sync voltage could be much lower, and thus safe for the switcher to accept.

So I grabbed my multimeter and hacked together a minimal SCART-to-Dsub15 test adapter with no components except one wire partially-exposed on the sync line so I could measure the voltage of the Genesis SCART Csync before and after connecting to the switcher.

I found that the voltage while not connected was about 4.8v, which was to be expected. After connecting, the voltage dropped to about 1.5v.

I plugged these values into a voltage divider calculator. Vin=4.8 volts, R1=470 ohms, R2=???, Vout=1.5 volts
The value for R2 came out to about 214 ohms; the resistance on the input is probably more like 220 ohms and I ran with this assumption to be cautious, since over-attenuating wouldn't hurt.

I then added a 330 ohm resistor on the LM1881's Csync output, which in this situation brings 5v signals down to 2v and should bring 12v signals down to 4.8v, comfortably within the TTL levels the switcher expects.

I live in North America and I probably won't use a PAL system for anything any time soon but it's nice to know that if I ever want to use Dreamcast or Wii RGB for any reason, this might be the change that makes it safe. If there's anything wrong with this way of going about it, of course, I'd be very happy to be corrected on it.