VGA signal generators vs other alternatives, are there any? (DIY video synths)

Hi! Noob here :slight_smile:

When trying to find information about DIY video synths (not the commercial kits), I’m seeing the CHA/V -type circuit bent VGA signal generator designs everywhere, while alternatives seem to be nonexistent. So I’m just wondering, isn’t there any other ways to convert oscillator signals into video, maybe even straight to yPbPr or av composite? And if not why? In this case audio and video signals are anyway both analog and can use the same kind of rca-cables. In these VGA-conversion-based designs, is the point of the VGA signal generator to send it straight to RGB channels and make hsync and vsync easy, or is it a necessity in all video synthesis?

Thank you!

1 Like

I’m still learning a lot but whatever you do the horizontal/vertical sync pulses have to be there or the monitor, analog or digital, won’t know where to draw whatever. You could insert a signal into RGB but without the sync it’d just drift around on the screen, assuming you even get a screen because without sync the monitor won’t really know what to display.

The CHA/V uses the VGA board to generate sync and provide a way to get those signals in and gives you a lot of options in a relatively inexpensive way. As someone starting out as well, dealing with these sync signals seems to be the biggest headache and it doesn’t help that the standard has evolved over the past decade.

3 Likes

Ah yes, that makes sense! Thanks for the reply

If you really want to get into DIY analog video synthesis at a level beyond what you can do with a CHA/V, I’d go for the LZX Cadet series of modules. They’re not sold anymore (though there may be a few synth DIY places selling some of the last of the PCBs) but their schematics have always been open source. And after they were discontinued, the company also all the gerbers and diptrace files and art files for them, so people could just straight up make the exact same boards for personal use, if they didn’t know how to (or want to bother to) adapt the schematics into their own designs. The schematics and BOMs are here: GitHub - lzxindustries/lzxdocs: LZX Industries Product Documentation (in the “Cadet …” folders and the gerbers/diptrace/art is here: GitHub - lzxindustries/lzxcadet: LZX Cadet Series EuroRack Video Synthesizer DIY Modules

These modules are excellent simple building blocks of video synthesis. An oscillator is one module, a fader is another module, a video input is another module, and so on. With them you get to really understand the signals and how to process them to make neat effects. And there’s meager but enthusiastic and friendly community using them (lots of people often willing to help debug issues you might have on the LZX forums, for example, including Lars at LZX). I didn’t really understand video synthesis until I started with the Cadet modules, and now I love it and am working on several designs myself.

Plus there are some other people/companies who’ve made their own compatible modules to provide even more functionality and are similarly sharing the schematics and selling PCBs (and sometimes kits), such as Syntonie, Visible Signals, and Foxing Hour.

There are some downsides of this approach, when compared to CHA/V (and CHA/V was created in part to address some of these concerns). First, it’s a lot more expensive than CHA/V. The “H” in CHA/V stands for “hacky” because it doesn’t really do things “properly”. There’s no electronic protection involved, there’s no impedance consideration or buffering. There’s no blanking of the incoming signal during blanking periods. Most of this stuff isn’t a big concern for the CHA/V for various reason, but when LZX designed their Cadet modules, they tried to do things as correctly as possible within reason to both make it safer (for the larger eurorack ecosystem) and because it’s a lot more complex than CHA/V and you need to have things like buffered signals in a larger more complex system. So this means there are lot more parts involved and some of those parts (like the op amps) are kind of expensive when compared to things like resistors and capacitors, and even when compared to the op amps used for audio circuits (since video deals with higher frequencies than audio). Another drawback (in some respects) is that you need a eurorack-format case for the modules, since they’re not standalone like a CHA/V. And you also need a power supply, and the power supply needs to be at least a bit better than your average wall wart, since noise in the power can add noise and other issues to your video output. You don’t need to necessarily get a super low noise (aka “clean”) eurorack power supply like some modular video people get (many ones designed for audio will work great or at least OK for video), but there are definitely a few eurorack power supplies that work fine for audio synths but really poorly for video. Another drawback for the Cadet modules is that one specific part on the Cadet 1 (which is one of the modules you need to have for a Cadet system) is effectively no longer available. There are some workarounds some people have figured out that you can read about in the LZX forums if you want. But it might be easier to build a Syntonie VU007B since that’s a module that can replace a Cadet 1 and Cadet 2 in a lot of cases (though I think it only has RCA sync out, which means some Cadets, like Cadet 4 and Cadet 9 won’t work with it, though there are workarounds). I’m not trying to scare you away from this, but in general it’s a lot more expensive and complex than CHA/V, even though it’s still DIY (again, hence why CHA/V was created – the “C” stands for “cheap”).

I’m not sure what you mean by “the commercial kits” but that could encompass what I’ve recommended. Ignoring these things just doesn’t make sense to me. They’re all such incredible resources. You could use just the schematics and make everything 100% DIY if you don’t want to pay the people who are selling the PCBs and you know how to design PCBs yourself if you really want.

4 Likes

As to why there isn’t a large number of DIY (analog) video synth options out there, that mostly boils down to the fact that analog video is old and not that popular. Sure, TVs sometimes still have a composite input on them these days, but some don’t at all because a lot of new video gear doesn’t support that. And sure, TVs still often have a component input on them these days (more so than composite inputs, I think), but they’re also less popular than the much newer HDMI, which is digital not analog.

The lack of popularity of the analog formats (composite and component video) has a very unfortunate impact for people looking to make analog video gear: there are fewer IC’s (integrated circuits aka “chips”) made to do the kinds of electronics needed for analog video. Back in the 70s/80s/90s, there were a lot more kinds of IC’s available for analog video, and some of these made DIY video synths a lot easier to make. In particular, there were sync generator chips available, and sync generation is very important to video generation. Nowadays there are some sync generation chips available, but they’re not very DIY friendly at all, and they usually need some sort of digital brain to drive it, further complicating things. Similarly, there used to be a lot more encoder chips (that take analog inputs) that generate the actual video than there are now. In fact last I checked, there’s only two left that I’d say are remotely DIY friendly (and they’re SOIC SMD chips, which some people can’t handle soldering). In the 70s/80s/90s, there were various easy-to-DIY-with DIP chips for all of these things. Even the LZX Cadet system which I previously mentioned uses a couple SMDs (which were pre-soldered on the boards sold by LZX).

I’m pretty sure that if we had nice DIP sync generator chips and video encoder chips these days, there’d be a lot more DIY video synths out there, given the ease of searching for and sharing circuits these days (with the internet) and the ease of designing PCBs (with open source or free-ish trial software) and the ease of manufacturing PCBs (via all the popular inexpensive Chinese board manufacturers). All of these things were much harder and/or more expensive just 20-30 years ago, when all the nice video ICs were still available.

3 Likes

I realized that despite my long posts, I didn’t directly answer some of your questions. Whoops. I’ll attempt some shorter direct answers…

@jmsiener mentioned the main point: Sync. You can’t really input oscillators directly to YPbPr or composite because those are expecting v-sync and h-sync pulses to be in the signals (though for YPbPr, the sync is only in the Y channel). Without that sync, the tv/monitor/capturecard has no idea when a frame is supposed to start or when a line in a frame is supposed to start.

A more minor issue is also sync related: Without sync, it can be hard to get oscillators to stay at frequencies that look good on displays. Oscillators drift, especially if they’re analog oscillators. Digital oscillators are much more stable, but at video frequencies they tend to still drift when not sync’d. Unsync’d oscillators do indeed have their use in video synthesis for glitchy stuff (when very unsync’d) or for vertically-rolling bars (when slightly unsync’d), but having sync’d oscillators is really nice too.

In a VGA signal, the h-sync and v-sync have their own dedicated pins/wires. That’s why the CHA/V works so well, because it’s easy to inject your oscillator signals onto the R, G, and B signal channels without touching the h-sync and v-sync channels since they’re all separate in a VGA signal. And VGA devices are usually very forgiving of the signals… usually they don’t need the R, G, and B to be blank during the sync/blanking periods. And in the case of an LZX-compatible video modular system, the sync is also separately routed, via some cables inside the case. If you’ve looked into DIY video stuff, you may have come across certain IC’s called “sync separators”. These do just that – they extract the sync out of a signal (composite or the Y channel of YPbPr) and spit out the different sync signals (h-sync, v-sync, and some others) on their own channels, so that whatever device it’s in can easily synchronize to the signals without the picture info getting in the way.

4 Likes

Any interest in a cheap little 8-pin DIP that would produce 640x480 VGA sync signals? Then you could experiment with jamming all kinds of things into the RGB channels and not have to worry about producing Hsync and Vsync. Could be fun. I could spin these out pretty quick.

3 Likes

Thank you Joem for the detailed answers! Good to know that LZX schematics are open source, I’ll take a look at them.

That would be fun! I’m interested

Is there a composite or component signal generator? I’m having trouble finding one… don’t even know where to start…

I don’t know what this is…


Maybe it could work?

This a great idea though! Trouble for me with VGA synth is soldering the back of the signal generator… Right now, I’m kinda imaging big chunky circuits you could plug into to add effects.

Had no idea LZX was open source what an awesome
company!! They have some really cool stuff!!

OK, I currently have a little 8pin DIP that farts out 640x480 75hz VGA signal with relative stability.

Really easy to use: 5v in, VGA HSYNC and VSYNC out

I’m going to continue to bang on the math to try to get it to be 640x480 60hz VGA, but so far it is working.

Would anyone be interested in this chip? Or perhaps a little board that makes a 640x480 H/V Sync and prevents you from frying your VGA display by clamping any input that is more than 0.7v in the RGB channels?

Really simple little board:
5v IN
R IN
G IN
B IN
VGA port OUT

What do ya think? I could probably spin this out pretty quick.

5 Likes

Sounds awesome to me! I’m interested.

Heya, did you eventually end up making these boards? And if so is there a way to acquire some?

2 Likes

Also interested in this chip/breakout board

I think the product has been updated since but you can check it out Here

3 Likes

Yes, you are correct! Thanks for the shoutout. I’ve been busy with holiday stuff and didn’t have a chance to respond yet.

1 Like