Add s-video output to VHS VCR? (Sharp VA40HM)

[bakerie]

I realise this isn't necessarily a repair question, but figured general VCR hardware questions should go here?

I've an old VA40HM on the way that's reportedly working great. (we'll see) I'm interested in the potential of breaking out an S-Video from it. Is this something that's as easy to do as it appears from the schematic?

I can see plenty of points mentioning both PB-Y and PB-C and I think these are playback Y/C, but it goes through so many amp and filter stages, I'm not sure if there is somewhere I could pull a Y & C signal to to grab an S-Video out.

Would I need additional filtering for it to be S-Video compliant, or would this be generated before the final mux to composite anyway?

It looks like around the VHIAN3248NK is where I would need to look, but absolutely no joy in attempting to find a data sheet for that.

Anyone on here with the experience know where I could grab these?

I've attached the specific PDF page I think it relevant, "Vca30Hm-Y_C.pdf".
I've also attached a zip of all the manual in case anyone wants it.

Here's an image of the page: https://i.ibb.co/LhYTCs6x/Vca30-Hm-Y-C-page-0001.jpg

Thanks for any advice!

[radiokom]

There is only one question. Why? There are no "direct mods" anyway but there are lot of VCR with S-video output. So this, in my opinion is something like question "why dogs lick their balls?". Answer is "because they can!". Sorry

[latreche34]

I personally wouldn't bother with such an old basic VHS VCR, If this was another obscure video format with limited supply players I would say it is worth it, VHS was a popular format and technologically progressed over the years in terms of picture and audio quality and features, Just get yourself a S-VHS VCR built in TBC and use the money and effort to bring it up to factory condition.

[dpalomaki]

I agree it is not worth the effort if you are doing it with a view to improving your VHS capture capability. Buying a decent S-VHS machine would be faster and likely assure better results. Do it [B]ONLY[-B] if you want the fun of designing and building circuits and the risk of trashing to existing boards when you tap into them. Note that the Y and C outputs would have to present the correct levels and source impedance to your downstream gear, which likely are not the same as the internals of the machine where you might tap into signals.

[bakerie]

I'm doing this because I want to and it seems feasible.
I have a Panasonic NV-HS830 here beside me.
I also have an oscilloscope and any other tool necessary to do this.
I'm just lost in the circuit design and someone who worked on this hardware back in the day could probably tell me in two minutes where to poke.
Was hoping someone like that might be around here, but maybe that knowledge is lost!

[Bogilein]

Maybe this could help, but it's in german language.

http://www.heimers.ch/de/n1700.html

[bakerie]

That was an interesting read, thanks!

It's not enough to get me over the line but it showed:
1. I know that my thinking is correct and both signals are there just before the composite output.
2. Unfortunately the signal levels might have to be adjusted and that might be a challenge.

[aramkolt]

Your machines has pretty straightforward points to actually tap Y and C. You want the ones that are "PB Y" and "PB C" - PB standing for "Playback"

In this case, the diagram is pretty straightforward as to possible points you could try to tap those on IC201

For Y, you should try Pins 14, 17/18, 21/22.

For C, there aren't as many options, looks like just pin 12


The trick then is getting them to usable levels which if they are close enough to standard levels might be possible with a proc amp, though admittedly, there aren't very many S-Video proc amps available to choose from.

Ideally, I think you'd buffer the signals with an op amp or transistors so that once loaded it doesn't load down the input once attached to the rest of a chain.

That Phillips N1700 circuit is interesting - My guess is it pertains more to the specific levels that machine has at the tap points, but you could probably play around with the resistor values to try to get things to a usable range if using a breadboard for easy component swapping. It appears to be based on a buffering circuit so that current can be amplified without loading down the input signal.

My suggestion would be to use your oscilloscope at those different test points above and post back with what the waveforms look like at each with the voltage scale shown. From there, perhaps someone who knows more about op amps and buffer circuits can comment on how to wire from there.

[bakerie]

Hi!

So I was thinking about this. Could I make a test tape with signals on it using the S-VHS recorder I have and then see if I get the same at those test points?

Or can I just stick in any old tape, connect the oscilliscope and see if the signal is within the standards?

I'm using this as a reference:

Quote:

The S-Video signal consists of separate luma (Y) and chroma (C) components transmitted over dedicated lines, with the Y signal carrying luminance and synchronization information at a nominal voltage of 1.0 V peak-to-peak (Vp-p), including sync and blanking levels. The C signal, which conveys chrominance modulated on a subcarrier, operates at 0.286 Vp-p for NTSC systems. For PAL systems, the C signal voltage is typically 0.3 Vp-p to accommodate the differing color encoding standards.[25] Both the Y and C lines adhere to a characteristic impedance of 75 ohms, consistent with standard analog video transmission requirements, ensuring matched signal propagation and minimal reflections when using coaxial or appropriately shielded cabling.

(I'm in a PAL region for reference)
So if the single C output matches a 0.3 VPTP signal, that's that one sorted.
I can then poke about the Y pins to see if any match.

Next up would be to connect my new S-Video output into something and see if the additional loading causes the signal level to drop. If so, then I need a buffer?

Impedance stuff always confused me. Does this mean that the TV (for example) will present itself as a 75Ohm load? If so, can I just simulate a load on the Y/C signals using a 75Ohm resistor (just for testing)?

[aramkolt]

Yes, making a color bar tape would be useful for locating the signals.

And yes, a TV or capture card should usually present as a 75 ohm load which you can emulate with a resistor on both luma and chroma to ground. The other thing you can do is hook it up to a real TV or other VCR during your testing and use that as the load.

You should be able to verify that the receiving device is a 75 ohm load by measuring resistance from ground to the chroma and luma pins on the S-Video connector, but I believe that is usually the case.

You could also look at a service manual for TVs and VCRs and look if you see 75 ohm resistors to ground right as the S-Video signals enter the machine. I looked at a random HR-S9900U manual I had on my computer and interestingly, the luma has a 75 ohm resistor to ground on luma, but an 82 ohm resistor on chroma through a capacitor to ground.

Unless you are getting signals that aren't pulled down by the ~75 ohm load, you'll likely want to implement variable resistors on whatever ends up controlling output levels so those can be adjusted over using fixed values.

[bakerie]

Any sources on a colourbar tape that will show what I need? I can burn it to a DVD and record it over S-Video, so I have a proper S-Video signal.

[lordsmurf]

Quote:

Originally Posted by bakerie

I'm doing this because I want to and it seems feasible.

Sometimes "can you" and "should you" are vastly different, as is the case here.

- From a usage stance, the entire notion is ridiculous. Buy a good JVC S-VHS VCR, and done.
- But from a purely reverse engineering stance, it's quite interesting.

I will be following this conversation.

[aramkolt]

Quote:

Originally Posted by bakerie

Any sources on a colourbar tape that will show what I need? I can burn it to a DVD and record it over S-Video, so I have a proper S-Video signal.

Color bars from a DVD is fine, Digial Video Essentials has several patterns. I think the NTSC version is on archive.org, but not sure about the PAL version. My guess is you can find the PAL version on ebay for something like $10-$15 shipped and is worth the pickup if you're into that sort of thing.

[bakerie]

So will I just throw together a DVD with maybe 5 minutes of colour bars, 5 minutes of pure white, 5 minutes of pure black and that should cover me.

It's funny, I'm going around the web trying to source something and I keep ending up back in this forum

-- merged --

Ended up grabbing some colour bars and putting them into a video that plays for 1 minute, then 30 seconds of black and then 30 seconds of white, looping the whole thing 5 times. Should do the job.

All the downloads for quick test DVDs I could find where NTSC based.

I authored it in case anyone else needs something quick and simple.

[dpalomaki]

PLUGE type color bars for PAL should include white and black areas and colors at appropriate levels for video system alignment.

Keep in mind that normal VHS VCRS cannot playback a S-VHS recording (although some later production higher end models had that feature built-in.) One would have to force the S-VHS machine to record in VHS mode - using VHS tape (not S-VHS coded tape) should do this for you.

A multi-burst and/or cable sweep signal would be helpful for evaluating band width (frequency response) issues. These include B&W images ranging from 0.5 to 4.5 mHz (notionally 40 to 350 lines resolution). Also a modulated ramp signal, one that provides a uniform ramp for 0 to 100 IRE over the scan width can be helpful

IMO the primary advantage of having separate Y/C outputs is you that can avoid the Y/C separation section of your capture device, often a weak spot. The use if IC is elegant, but for the relatively simple electronics needed to buffer the ourput use of plain discrete components (transistors, resistors and caps) may well be easier. The primary issue .is your skill level as an electronics technician, which I cannot assess.

[aramkolt]

Video99.uk made a sort of video buffer board for that same N1700 based on the design linked above which might make things easier, see here:
https://www.youtube.com/watch?v=_db5KhszagU

He did also put out the Gerber files to make those PCBs.

Interestingly on this build anyway, the chroma uses only passive circuitry and doesn't need to be buffered with the N1700. I don't think there's necessarily anything stopping you from running the same circuit on the chroma side as you do for luma if the chroma signal gets loaded down too much.

Still might not be as ideal as using an op amp though, but you'd probably want to ask someone who knows more about op amps what the best low noise op amp circuit is for S-Video buffering/amplification as well as what the basic circuit design would look like with a variable gain. I always thought it'd be interesting to just use an S-Video distribution amplifier that has variable gain, but I'm guessing those also load the signal before the gain happens.

[bakerie]

A very quick look at this seems like he's just impedance matching the chroma and then using a unity gain amplifier stage as a buffer so he's not loading the weak Y- signal.

I think I'm going to attempt this first without the buffer. The reason that circuit makes sense in those older machines is that they don't even have a composite output, so the signals being generated are just for the RF modulator. As my machine is generating signals destined for composite output, they should be more or less ready to rock before being multiplexed, so the additional circuitry is probably unnecessary.

I've any amount of op-amps and transistors here, but I'm hoping all that won't be necessary. We'll see when I get the scope on it.

[radiokom]

Quote:

Originally Posted by bakerie

The reason that circuit makes sense in those older machines is that they don't even have a composite output, so the signals being generated are just for the RF modulator.

It is not because it is older machine, but because it is BOTL machine. If you want to mess with s-video out, there is, for example, Panasonic NW-W1 multi-format deck (I use it for SECAM). It is decent Hi-Fi deck, bet no s-video out. It could be interesting if someone would make s-video out for that and share his experience. But to do it for bottom of the line deck...

[bakerie]

That's a fair point. I've a Hitachi VT-8000E and it has composite out.

[bakerie]

Quote:

Originally Posted by dpalomaki

PLUGE type color bars for PAL should include white and black areas and colors at appropriate levels for video system alignment....

...A multi-burst and/or cable sweep signal would be helpful for evaluating band width (frequency response) issues. These include B&W images ranging from 0.5 to 4.5 mHz...

I had actually started to generate these, but when I looked at the documentation on how to use them, it started getting too into the weeds for a hacky bit of fun. I'd end up using time to learn something I'd never use again.

Quote:

Originally Posted by dpalomaki

Keep in mind that normal VHS VCRS cannot playback a S-VHS

Of course.

Quote:

Originally Posted by dpalomaki

IMO the primary advantage of having separate Y/C outputs is you that can avoid the Y/C separation section of your capture device, often a weak spot.

From limited experience in the 2000s, the picture quality from S-Video is far superior to composite and wasn't a million miles away from a dedicated RGB signal. We used to convert RGB to S-Video so we could run shotgun coax to get a decent video signal across distances of around 50 meters. Standard composite was so poor, that it wasn't really worth running compared to RF. This was getting signals from Satellite boxes to analog TVs back then.

Composite signals where very washed out and the interference at the edge of digital overlays in particular was horrible looking.

Quote:

Originally Posted by dpalomaki

The primary issue .is your skill level as an electronics technician, which I cannot assess.

I'm at this in one form or another for about 20 years, with a good chunk of that involving designing circuits and PCBs, but all of that is digital low speed stuff and uses modern interactive schematics that are easy to trace. I'm fairly lost when it comes to analog signals and hand drawn schematics!