Hi guys,
In the radio control community FPV flying is what's happening right now. Live video to your ground station from your flying vehicle and you can fly like your playing a video game.

One big issue with this "sport" is LCD screens going blue or to standby when video signal is poor. I've been told this is mainly because the sync pulse deteriorates. Is this as easy as it sounds, just measuring the period and duration of the sync pulses on startup (with good video signal) and then keep pulling the video signal to ground when the video signal is/should be present?

I was also wondering if you know of a circuit like this.

 

I can't figure out what you are thinking of. Are you saying to try to artificially force a sync signal in the ground equipment based on when a sync signal supposedly should occur?

Won't work because the sync signal is there to let the receiver synchronize to the actual signal content. This would be like driving down the road and if you can't actually see the stoplight ahead just deciding whether to stop or not based on when you believe it should be green.

 

I know, I see your point. And you're right about my intentions. But when the sync signal is gone, all we can do is guessing, right? This would be like driving down the road and if you can't actually see the stoplight ahead you can actually have a good estimate which tells you if you should stop or not. Which is better than having no clue?

Of course, the best thing to do is to stop generating pulses when the signal is fine again.

I'm most certainly interested in other ideas that can work.

 

What video standard is your system using? Is it an analog signal, or streaming digital?

And don't forget that you are assuming it is only a sync issue because that is what you have been told. Doesn't make it true.

 

Thank you for your advice. About it being a sync issue or not, you are right and I will be aware of that.

The video signal is composite video and most often I think lcd screens are both NTSC and PAL compatible. If the problem is only sync pulses though, I don't have to think of the black level setup duration for PAL. I've been reading this document from maxim http://pdfserv.maximintegrated.com/en/an/AN734.pdf (TUTORIAL 734 Video Basics).

It seems like I need a timing resolution of at least 0.1us, probably closer to 0.01 would be best.

I found an interesting chip, LM1881 Video Sync Separator. Which could almost eliminate programming needs.

 

Also, this is a nifty chip. Maxim loss of sync alarm.

Although the costs are starting to add up here, they could make an interesting combo. I would prefer the circuit not to cost more than a couple of bucks in large quantities.

 

The analogy is more like driving down the mountain road and a fog rolls in and you cannot see the bend in the road.

 

I know why it's called a sync signal, it keeps the video receiver synchronized with the video transmitter. But am I really that far off? I'm guessing the sync signal will appear at exactly the same time interval with little or less drift, so will it really be a problem producing these sync pulses for a period of time while the real sync is gone?

Of course the best thing to do is some sort of processing based on AT LEAST whether the sync is lost or not (hence the maxim chip).

Does anyone know more details about what missing part of a composite signal makes lcd screen drivers blue or blank the screen?

 

What you need is a syncgen and have it lock in to the received sync.

 

Yep, that's what I was thinking. Any ideas on how this would be best solved?

I'm thinking of using the LM1881 so I can easily extract a sync pulse and detect sync missing. To isolate the input and output video signal is an OPAMP 1:1 gain sufficient? Because if I just fed the sync pulse on the existing input signal the LM1881 would think the sync pulses are back again.

 

...
Does anyone know more details about what missing part of a composite signal makes lcd screen drivers blue or blank the screen?

Loss of a sync signal is very normal and traditional way of testing if video has gone bad. In older TV sets with AV inputs it was common and the sync chip would activate a video blanking upon loss of sync.

These days the sync separation etc is likely to happen in a large complex chip that decodes multiple video formats, so it will very likely go bluescreen on loss of sync BUT might also go bluescreen based on other video failures like signal strength (amplitude) or issues with the colour information etc.

In concept your idea may work, you could use a PLL or micro etc to lock to the hsync frequency and generate it manually.

Also, you don't really know if the fault detect is based on loss of hsync or vsync pulses, or some "smart" combination if it is a new TV.

 

You need to determine the exact HSYNC and VSYNC generated by the camera.
Then you can replicate the exact sync signals with your own circuitry. You can use an mcu with PWM outputs to do this. Then you can use the LM1881 to sync to the incoming signal.

Edit: RB was ahead of me and is saying the same thing.

 

Here's a suggestion circuit. About the sync pulse driving, I'm kinda clueless... Can I have some help on that? The idea is to isolate the input from the output so that when I put the sync pulse on the output, the LM1881 won't know about it.

I've seen measurements done on the sync output of LM1881 when video signal is very, very poor. It starts outputting pulses very fast, so detecting a bad signal is no problem.

 

You need to determine the exact HSYNC and VSYNC generated by the camera.
Then you can replicate the exact sync signals with your own circuitry. You can use an mcu with PWM outputs to do this. Then you can use the LM1881 to sync to the incoming signal.

Edit: RB was ahead of me and is saying the same thing.

But the period of the sync signals on composite video is fixed, isn't it? Or what do you mean by determining the HSYNC and VSYNC, frequency/period or duration/duty cycle? So imagining the sync is missing from the input, as long as I'm in sync with the missing sync (!) I can just generate a pulse width of the duration from composite video specs? I mean, the receiver (in my case an ebay lcd for backup cameras) must already accept all composite video sources, either pal or ntsc.

Bu the way, will the sync pulse on composite video swing below 0v? From some sources I've seen, it seems like it, but the figure on the wikipedia site shows the sync pulse at 0v.

Also, you don't really know if the fault detect is based on loss of hsync or vsync pulses, or some "smart" combination if it is a new TV.

It seems most LCDs will keep showing scrambled video even if only the horizontal synd is maintained. There has been done some experiments on this I can remember from searching the topic a while ago, but of course this must be tested.

Thank you both for informative answers. Does anyone of you have the time to check on the circuit I posted above?

 

Thank you for your advice. About it being a sync issue or not, you are right and I will be aware of that.

The video signal is composite video and most often I think lcd screens are both NTSC and PAL compatible. If the problem is only sync pulses though, I don't have to think of the black level setup duration for PAL. I've been reading this document from maxim http://pdfserv.maximintegrated.com/en/an/AN734.pdf (TUTORIAL 734 Video Basics).

It seems like I need a timing resolution of at least 0.1us, probably closer to 0.01 would be best.

I found an interesting chip, LM1881 Video Sync Separator. Which could almost eliminate programming needs.

Most Broadcast TV standards with the notable exception of fhe old British 405 line system,use what is known as "Negative Modulation".

In this system,the sync pulses correspond to the the highest power level of the radiated signal,& peak white to the lowest power level.
In practice,this means that even the snowiest picture will remain synchronised.
A similar practice is used with frequency modulated video links,where the highest frequency deviation corresponds to the sync tips.

I have no way of knowing what form of modulation you are using,or whether it conforms with these conventions,but if it does,you will find that if the syncs are not usable,the picture information will also be lost.

For many years,TVs & Monitors did not go to a blue screen or standby on loss of syncs.
Instead,the horizontal & vertical oscillators would free run at close to the normal frequency,& display whatever appeared as video---in the case of a TV after station closedown,it would display noise.
Remember the TV in Poltergeist?

The sync circuits in CRT type TVs & Monitors are really phase locked loops,& can tolerate the dropout of a few pulses without losing lock.

Since the late 1970s,loss of sync normally causes the blue screen or standby effect.in all sorts of displays.

Some of the early versions of this let you disable this function,& even later on,you could modify the circuits,but as most modern stuff is done in software,you may have difficulty.

 

There is a multitude of "standard" HSYNC and VSYNC frequencies. On the receiving end you have a certain degree of tolerance and don't have to be exact.

I would look to see what the camera (or source) is using and use that as your standard.
Generate your own sync signals and then lock on to the received signals.

 

If he isn't getting sync pulses now,I goubt that there will be much left of the rest of the video signal to sync to.

I think the OP is approaching this from the wrong end.

If the signal is so low in the noise that he can't see the syncs,maybe he needs to look at the receiver sensitivity.
Perhaps a better antenna?

 

He might still getting sync pulses as the video is not totally missing but is reduced in amplitude or quality?

That leads to an idea; a PLL type setup can probably be used to lock onto the hsync and keep generating hsync if the original is reduced in amplitude. So in other words you make a system that locks to a bad signal sync and constantly reproduces it. Something like a hsync "signal rejuvenator".

 

If he isn't getting sync pulses now,I goubt that there will be much left of the rest of the video signal to sync to.

I think the OP is approaching this from the wrong end.

If the signal is so low in the noise that he can't see the syncs,maybe he needs to look at the receiver sensitivity.
Perhaps a better antenna?

When flying model aircrafts fpv you will at some time experience bad signal reception. Of course you can upgrade your video system, but most people already have done that. Anyway, you don't want to go out flying an airplane knowing that in worst occasions your screen might go blue or in standby mode.

He might still getting sync pulses as the video is not totally missing but is reduced in amplitude or quality?

That leads to an idea; a PLL type setup can probably be used to lock onto the hsync and keep generating hsync if the original is reduced in amplitude. So in other words you make a system that locks to a bad signal sync and constantly reproduces it. Something like a hsync "signal rejuvenator".

Yes, yes. This is exactly what I am thinking. Previous tests done by others show that video receivers only test for hsync, not vsync, before going blue.

I posted a schematic some posts ago, and I need some help in the part about isolating input and output video signals, so you don't start feeding your own PLL or trick your system to think hsync is back up and running.

 

so many people (including myself) are looking for exactly this. Expensive fpv goggles for flying model aircraft cost hundreds of dollars ant this circut would allow them to use cheap lcd screens or make their own fpv headset. If you devellop this you'll make a killing.