How to remove this video interference?
 

So I've got my AG1980 connected to my TBC(ish) ES15+DVK100 and from there to my ATI AIW 7500.

Currently I've just got video going to the purple dongle, and I recorded a short sample of the blue DVK100 picture, which has some pretty obvious rolling lines... that I don't remember the name or the possible cause of. Some audio is getting into the video despite me recording this without any RCA L+R going to the purple dongle, all I have is the audio from the ATI AIW 7500 to the AUX of my Turtle Beach Santa Cruz.

Here's a short HuffYUV video with 500 frames.

https://drive.google.com/file/d/1RNH...ew?usp=sharing

Any ideas what I can do about this? I'm guessing it's electrical interference.

I don't quite understand your issue, can you post a usefull video with objects in it rather than a green screen?

Sorry, I figured it would be easier to see with a solid color. I'll grab another capture tomorrow.

I will mention that I am referring to the diagonal lines scrolling across the video. I think it's EMI, but I'm not sure, and I don't really know what to do about it if so.

-- merged --

Today I took the time to rewire my entire video chain and power it all off of one power strip / surge protector, and that has removed the diagonal lines. I guess it was a ground loop?

There's still some noise, particularly noticeable when I use ExtractU or ExtractV in Avisynth. Not yet sure if that's something that I'll just have to live with or if I should get some professional S-video cables. The ones I have *look* shielded but I'm not sure.

-- merged --

I switched the power strip / surge protector for a Cyberpower UPS that I had. It's... alright? I don't really notice much change. I used an SVHS capture that I got from TGrantPhoto as a test record, I'll try to get a sample uploaded later.

-- merged --

Here's a new sample. Beware, file is 600 or so megabytes.

https://drive.google.com/file/d/1spr...ew?usp=sharing

My main concern now is that audio at the beginning.

Edit: Here's a short audio sample.

What finally got rid of that pulsing was removing the damn CD audio cable from between the AIW7500 and the TBSC's AUX port. But that was my audio source, so this solution isn't exactly ideal.

When I got the cable from ebay I did have to rewire it. I also had it tied into a loop with a twist tie. Perhaps I screwed something up?

I also have a second AIW7500 that I can try.

Edit: Nevermind, when I connected my Monoprice RCA cables to the line-in port with an adapter the noise comes back. So I have no clue.

Edit edit: Oh but the noise is not nearly as bad as it was in the above sample. So it's improved.

-- merged --

Here's a sample of the line input and the aux input. At the end of the line input record I remove the cable entirely to hear what it sounds like with no connection.

Edit: Aaaahhh, it's even quieter from the line input when I switch the VCR to input 1 rather than input 2.

Edit edit: Okay, so it looks like I should keep my AG1980 on input 1 (back) when dubbing, because for some reason when it is set to input 2 (front) it is introducing some noise. No change when I change the inputs on my ES15. Also the AUX input is really noisy for some reason, so while I am going to leave it internally connected I am going to use the RCA outputs on my AG1980 and go straight to the line input on the TBSC via an adapter.

I disconnected all my VCR's front input connections ribbon cables from the main board inside the VCR, They serve nothing but add interference, Always capture from the back connections since they are soldered right on the board.

I was surprised to learn that it would add noise even if I had nothing connected to the front panel input, while switched to it. Maybe sometime I'll try disconnecting them internally, but I'd probably only do that if I thought they were adding noise even when not switched to them.

I found a new AUX cable (with a Y split that I may later snip off) in my storage today, that I didn't even have to rewire. It's got cleaner audio than the AUX cable that I was using, but still not as clean as the line input.

On some systems, AUX cable is just a noise vector. So you have to route RCA audio direct from VCR to capture card via quality RCA>mini cable. I like AUX, cleaner wiring. But some AIW models, some VCRs, some systems, it's a no-go. I have a system like that myself.

But interesting about disconnecting ribbons. I never thought of that. I need to open some decks.

@latreche, which exact models have you done this on?

Probably yes. Running equipment from different power outlets and then connecting them to each other is a recipe for ground loops. As is leaving connections to the VCR inputs. For tape capture it's safer to use only the video out and the audio line outs.

I also noticed that the clip you uploaded seems to be in mono. That's unusual.

JVC's, I never bothered to do it on Betamax, Too much stuff in the way and I'm sure Sony has coaxials running not like JVC running a strip of ribbon cables from the front left all the way to the back center with no shielding what so ever.

I dont understand this. Normally in play mode, arent all video and audio input signals electronically disconnected or muted from appearing at the outputs?


I dont understand this either. Are there VCR's with video and audio output sockets at the front?

It's not about the video signal itself, the ground side is shared. Anything extends is a target for interference and acts as an antenna, I even though about removing the TV tuner box, just wasn't sure how the system control would react afterwards and may give an error code.

There's nothing inherently 'wrong' with the shared ground/earth (or 0V proper) - it's how these devices work ultimately, check out any schematic and you'll find the same unless I've misunderstood your point and you're saying it's not the problem that they're shared - I'll grant you that if I've misunderstood.

We always 'terminate off' any unused terminals on our machines (as would generally be done in broadcast) as a matched termination will curtail a good deal of interference. I know enough about electrical noise to know I don't know enough about it. There are libraries stuffed with volumes detailing methodologies for minimising electrical noise and still lots to discover.

Mismatched impedances are a nightmare, and like you say tend to become very effective aerials (antennas) with cable runs etc. I'm (to my genuine shame) a ham so this is a deeply complex topic I'm not really qualified to give advice on beyond general advice. Sometimes knowing 'a bit' highlights how little you know under scrutiny, but these are interesting experiments and something I might look at myself over my long awaited break over Chritsmas.

I guess later on I'm going to try bypassing parts of my video chain to see if the noise that appears is caused by something specific. It kinda looks like hum bars to me, but I'm not sure.

-- merged --

I'm thinking it might be the DVK-100. In isolation it seems to be the noisiest device, with rolling bars. Anybody have any advice on how to ground this thing? It has the back nut thingy for doing so, but I don't... know how, electrically speaking.

Odd. Never seen that with a DVK. It has to be local power issue.

If you are the UK, I would suggest connecting directly the terminal to mains earth - however I am not qualified to give advice for any other domestic wiring system. As you used the term ground rather than earth I guess you're not in the commonwealth nations?

Does the device use an external PSU or does it just connect directly to the mains?

I'm from the USA. I kinda swap between the terms. :)

I'll have to find a means of connecting it to ground. The DVK has a tiny SMPS I think, that plugs into a Cyberpower UPS.

Here's a sample of the DVK, the ES15, and the AG190, all with a signal generator with SMPTE bars going in and a direct record out. I think I'm hitting placebo territory, because I am not sure if I can easily see the difference. I think the AG1980 might be cleaner though.

Go simpler. Your power grid may hate that power supply. And that's easy to fix.

Well I'm already pretty simple (i.e. stupid), so I have to ask, what do you mean by that?

I realized that I had the Cyberpower UPS plugged into a rack and I hadn't considered our facility's UPS, so I moved it to a plug that I know is not on the facility UPS. Still analyzing a test capture now.

The AC adapter that came with unit.

It came with a standard 12V 3A, but 2A works fine, even up to 5A. The amps don't matter as much, but I'd floor it at 2A. Standard 12V center-positive, 2.1/2.5 tip. If you want to buy a new one for under $10, APD (good brand), I can point you in the right direction. Or perhaps some other devices near you already uses this rating, such as an external 3.5" powered hard drive.

Also realize that UPS can leak if other devices on the UPS leak badly.

Aaaahhh. Yes please, I'd like to give that a try.

Finally had a chance to try this after a long holiday break. Unfortunately the replacement power supply for the DVK100 didn't really seem to help.

I've got my entire system unplugged due to the power going up and down... winter is hard on the KUB system apparently.

My next couple of ideas is to replace the UPS with a new one, and then try replacing the PSU in the PC itself. I've been trying to get into electronics and power supply theory so I can understand how to even diagnose that, but I'll probably just swap it with a separate PSU.

Would anyone happen to have any recommendations for specific UPS and PSU models?

Edit: Oh also I've ordered a cheap isolation transformer, doesn't seem to have a ground pin though. As I understand it a lot of isolation transformers have the ground bounded from the input to the output, so I don't really understand if a regular IT would prevent ground loops.

I wouldn't fart about with UPSs until you've exhausted the basics.

As the noise is evenly spaced and continuously rolling it's most likely mains referred, mathematically anyway as best I can see it on a tiny screen using your sample at the top.

Did you try the supplementary earth bonding to see if that eliminated it? Also are there any coaxial connections in your system?

Also - to add, an isolation transformer with a common earth connection across the windings won't help with the issue you describe. It may help, but not if you're trying to 'lift the earth'.

So I added the ground wire with some spare wire I had. I'm pretty much a novice, so my idea was to solder that to a ground pin which I then plugged into the power strip (which is connected to the battery side of the UPS). ...Then the resulting image had really bad hum bars, I don't remember if they were there before that. I then plugged a ground pin isolator between the power strip and the UPS, here's the resulting image.

I have some coax *around* the PC, but not connected directly to it.

Yeah, the isolation transformer I purchased has no ground pin.

So I dunno, I still see some kind of noise in the chroma, but the luma looks squeaky clean. Is this the best that I can expect?

If doing that is exacerbating it, we can probably say it's mains referred.

There are some things that might be tried, however if you're a novice (and I'm not familiar with US domestic wiring, it's totally alien to the Britbong 'sparky writing this) I'm not going to suggest anything further regarding touching the power side. Await an approriately qualified US response!

Dump the UPS, honestly, they emit more broadband HF 'crap' than just about any other non-RF device, don't add things that do high-power switching into the signal chain at the moment! There are appropriate power-conditioning units, but a cooking model UPS isn't one of them when it comes to video in my opinion.

I've not really followed the thread, but that looks like an S-Video connection, have you tried a different cable? Could just be a basic shielding issue?

Another monkey wrench is the TVA. I'm betting you're on that grid. I had issues with DataVideo gear, now some 15 years ago. Reports are the this was resolved years ago, and this is the first time it's been possible in ages.

UPS are necessary in USA, our power is too prone to fail and flicker, and you'll lose a pricey item if plugged into a wall outlet directly. Some Europeans swear by using outlet-only power, but I'm dubious. There's really nothing special there, or anywhere. Power spike/fail can easily mean a $2K TBC = instant paperweight. I'm not risking my gear.

I have to try something like this soon, for a problem bass speaker.
https://amzn.to/33m2D9z
Not sure that'll fix it, but here's hoping. If not, Amazon return.

UPS are very electrically noisy, whether that has a bearing on your work is for you to determine and satisfy yourself with, or at least most basic 'interactive' units are, I can't speak for your unit in your application.

Our 'quality and continuity of supply' regs (harmonised across the EU & Britbongland) are pretty draconian and tightly adhered to. In my years of working around sensitive electronic appliances, the total I've seen damaged by 'power spikes' is nil. Like most of Europe we work on a national grid (helpfully in the UK, called the 'National Grid') which probably isn't as practical in the vastness of the US, many of the European grids are linked too which improves continuity.

I can't speak for all areas, but we had a power-cut around two weeks ago in one of my offices for around 30s and that caused quite a stir, power cuts do occur but they're very rare, probably a couple a year and for no more than a few seconds. We don't get a tremendous amount of lightning either, which is obviously climate and terrain dependent. It makes the national news if we get a half-decent thunderstorm; not to mention the nation grinds to a halt on anything deeper than 3mm of snow.

Also, in Britbong land often sockets will have voltages approaching 270V-RMS (@32A, in most cases, that's why we wire our homes differently and fuse our sockets) which is 330V PP which may mean the equipment is less sensitive to transient high voltage by design, possibly, I can convince myself either way on that one.

We need electric kettles due to the millions of cups of tea consumed daily here in Brexistan - we need high socket-power in every room and continuous supply. Speaking of which, my mug is empty..... :smack:

Anyway, @Plubbingworth, do you have a multimeter?

UPS noisy? No. Most actually purify the power, even the switchover types.

My coffee maker is on the PFC in the kitchen. And my mug needs refill as well. :D

Ah, I'll leave that there, I fear this may get dragged somewhere else. If you're satisfied it's not having an effect, carry on.

I got nothing there unless you're talking about Power-Factor Correction, in which case I'm excited?! :laugh: Oh, and I'm a coffee drinker actually, I'm not a full national stereotype. Tea tastes vile to me, but don't tell anybody or ma'am will want my passport back.

My cables don't appear to be especially shielded, I'm just using some cheapish ones rather than any from like BJC or something. I'm willing to go that route, but I'd prefer to try the less expensive options first? That being said, I did just drop the cash on the product that LS just linked to....

My concern with the UPS is that it's kinda old and I don't think the battery is very good.

I have a good multimeter, yes.

Two things you can do:

Firstly, check for resistance value between the outer 'ring' of both ends of the S-Video cable.

Then, (if you feel confident) see if you have an AC voltage between the casing of the machine and earth (ground), however, you get earth in the US, I think your sockets may have it exposed.

That 'thing' looks like a cheap cap & choke in a plug, doesn't mean it doesn't work, but I'd wager there's about 50p worth of components in it. As said above, send it back if it doesn't prove anything.

Oh well speaking of my UPS, I'm suddenly unable to power anything off of the battery side. So that's encouraging!

I'm glad you had me check my S-video cables, because the one that I had going between my DVK and my purple dongle actually had an open connection on the outer ring. A second cable I have on hand tests at roughly .35 of an Ohm. Unfortunately simply swapping the cables didn't seem to fix anything.

Uh well I have low Ohms between the chassis and the ground pin of the PSU if that's what you mean? Or do you mean to test for voltage between the metal of the chassis and the ground pin of the socket while the PC is on?

And I... also have low Ohms between the ground pin and the open screw holes of the rack that this is racked into, since I have this PC built into a rack case. That seems not ideal.

I feel really foolish for not considering continuity on that path. I'm going to have to take the entire PC out and test all of this isolated from the rack.

Yes, it's very common with some lowly cables for them to be unshielded. Shielding is a whole other complex topic, but fundamentally it's quite important for this and 0.350Ω is effectively a 'dead short' for the purposes of this conversation; you can't measure resistances in that range with conventional technique anyway, just a heads-up if you're new to this. Either way, it's fine.

Like I mentioned above, it's AC voltage you're looking for between any earth point on the device and your mains earth - if you have an AC voltage of any appreciable amount (you'll have a few mV due to technique and instrument) that means your equipment is 'above' earth. Anything more than a Volt or two probably needs investigating.

It's one of a range of things you could examine, but it's the only one I'd suggest. Obviously only do this if you understand what you're measuring and how to do it safely.

Otherwise, we're looking at power analysis and politely, if you're asking these questions, I'd suggest it's probably not for you yet.

Well done on getting this far if you're new to this sort of thing, a decent multimeter is still one of the most useful tools in any tool kit if you play with anything connected to electronics in my opinion.

So today I disconnected the PC from the rack, and set it on a stool. I powered the PC, the monitor, and the DVK100 from one power strip. I tried it connected to ground and I also tried it without ground. I've tried a different power supply in the PC.

The noise is still present regardless of these changes. This is getting immensely frustrating.

I've also tested for AC voltage on the ground connections... there's nothing there that I can see.

I should try and find a different test pattern generator than directly from the DVK100. But if neither the isolation transformer nor the hum remove device works, I'm tempted to look into replacing caps on the motherboard.

The trouble with replacing capacitors is that the 'scatter gun' approach may just shift the problem elsewhere -or not resolve it after considerable time and expense.

This isn't mains referred in the sense that it's caused by 'unclean power' (how I hate that term) by how you describe it but by your fundamental mains frequency encroaching on to your picture signal, doubly so if you've added a device in to reject common-mode and harmonic 'crap' and still getting the same results. We've lifted the earth to no benefit too, and proven there's no potential between device and earth.

I would now say this is a device issue, but without seeing the device and having test equipment handy it'll be difficult to determine. S-Video is fairly noise immune if you're using a reasonable quality cable due to various design features (gross common mode noise will be rejected at the input on the device, and it's a differential pair) so it's unlikely to be coupling in - not impossible, but unlikely.

A catastrophic impedance mismatch is unlikely with quality commercial products too.

I can't really offer much more without having the equipment in front of me and on a bench, but there are a few EEs here who may be able to chip in.

I can't quite make out from the thread, have you just tried another video signal source?

I skimmed over this thread and I don't see the OP is using a process of elimination method, That's the only way to figure out the problem or he will be posting here for the next ten pages or so.

First identify the faulty component by removing other components, like:
- VCR playing good tape to TV
- Stable video source to VCR to TV
- Stable video source to capture card
- Stable video source to TBC to TV ....etc. you got the idea.

Once the component is identified, then breakdown the area where it happens, like:

- Does it happen on composite?
- Does it happen on S-Video ? If so try turning chroma to zero, is the problem still there
- Does it matter if other components are too close?

This is just an example of the basic troubleshooting, Keeping all the component together and try to identify the problem by just buying and adding more stuff is not going to get you anywhere.

Good catch.
I just assumed this had been done already. Perhaps not.
Start over, step 1.