(Solved) iScan Pro problem (losing video signal)

[Billkwando]

I recently picked up an iScan Pro to upscale my Neo Geo to 480p, so I could run the SLG-HD (it's on the way) but I'm having a couple problems. The main one is that I keep losing the signal, or so it seems, and I only get a brown or gray screen. Eventually, the image comes back on its own. Restarting the console doesn't do anything (and the brown screen remains, even as the iScan tries to autoscan all 3 inputs for a signal. If I pull the power on the iScan it will usually pick it right back up.

Here's a video:

http://www.youtube.com/watch?v=f6bU00fg5Pk

The signal drops out around 02:25 and comes back around 05:25.

The hookup is Neo Geo (or Genesis, it does the same thing) via scart > CSY 2100 clone > iScan Pro > Panasonic Viera plasma (TC-P50U50)

Funnily enough, Waku Waku 7 hasn't dropped out yet, but it gets crazy squiggly lines when there's bright stuff on the screen (but not always): http://www.youtube.com/watch?v=1MmN-uyyXIA

Any ideas? Thanks!

[Fudoh]

Likely the same problem as discussed here: http://shmups.system11.org/viewtopic.php?f=6&t=46202

Did you use the transcoder directly on your TV successfully before ? Can you try the iScan via VGA to some other display, maybe a PC monitor ? I'm pretty sure that it's not the iScan's fault, but either the transcoder's or the combination of both.

[Billkwando]

Likely the same problem as discussed here: http://shmups.system11.org/viewtopic.php?f=6&t=46202

Did you use the transcoder directly on your TV successfully before ? Can you try the iScan via VGA to some other display, maybe a PC monitor ? I'm pretty sure that it's not the iScan's fault, but either the transcoder's or the combination of both.

I had a slight bendy issue at the top of the screen, but otherwise it was fine. Here's what it looked like, pre-iScan:

http://www.youtube.com/watch?v=AUd1HfMRPEw

Losing the signal like in the Ninja Masters video has never happened before. I have a thread at n-g.com:
http://www.neo-geo.com/forums/showthrea ... ost3568295

and there's a theory for what's happening in WW7, but it doesn't explain losing the signal on the Genesis as well, as that worked flawlessly with just the YUV converter. Weird huh?

[Fudoh]

same old problem with non-standard video or sync levels, even if it presents differently.

I have no experience with those CSY2100 clones. If you have a multimeter you can check what actually happens on it's output.

[Billkwando]

same old problem with non-standard video or sync levels, even if it presents differently.

I have no experience with those CSY2100 clones. If you have a multimeter you can check what actually happens on it's output.

I do have one.

Can you go into more detail about what I should be looking for? I also have a truckload of 75 ohm resistors, should the need for them arise, lol.

[Fudoh]

There is no straight forward solution, start with adding resistors to the RGB lines. You might need to lower the sync level as well and if it doesn't help you can add capacitors to flaten out the voltage peaks.

[Billkwando]

There is no straight forward solution, start with adding resistors to the RGB lines. You might need to lower the sync level as well and if it doesn't help you can add capacitors to flaten out the voltage peaks.

Ok, but what specifically should I be measuring with the multimeter? I'm trying to take your advice here. Should I be measuring ohms or what? Are there measurable standard levels we're trying to conform to here?

I don't mean to be bitchy (I'm failing here in that regard) but for all I know, this device could be defective. As I sit here, typing this post, Viewpoint (Neo Geo) which had been running fine for the past half hour, just dropped out to a gray screen. Not a flicker like the other post you sent me to (which was left unresolved with the OP still requesting help, I might add), it's just gone. 5 minutes later, comes back. Very frustrating.

The weirdest thing is that it's happening on multiple consoles. If it were just the Neo, which is known to be problematic, I'd be less surprised. Right now I'm looking at adding multiple resistors to multiple cables, some of which most likely already have circuitry I'm not entirely familiar with already added to them. It's a daunting process.

My apologies for the need for extra hand holding, but at least if I can resolve this issue we would know the solution to advise future users with similar problems. I appreciate any help you can offer.

Do you have any theories as to why the iScan would lose the input, and start scanning across the 3 inputs for a signal?

[Fudoh]

The problem is certainly (or very likely) not the iScan. Just grab any other component source to verify that it works as it's supposed to. Have a PS2 on hand ?

The AES is very easy to work on, so start with this one so work out the problems. What you're measuring is the voltage peak level on the RGB and the sync lines. With a white flash on screen (or a white transition) the voltage level should peak at 0.8V max. (0.7V is the standard consumer peak level). It's normal for the sync line to peak a little higher. Arcade systems can go up to 3+V and some AES and MD/GEN revisions certainly hit the 2V level without the proper attenuation.

The iScan loses the signal because the the inputs are "out of bound" (voltage levels too high). The difference in behaviour (pure grey vs. additional flicker) is probably caused by what follows the voltage peaks.

Once you understand why the NG drops out and how to fix it, it's very easy to apply the same fix to other systems.

The only question mark in your particular setup is that transcoder box. Before you start soldering you should make a table to note down which input voltage levels (on the console's output) correspond to which output levels from the transcoder.


PS: resistors lower the signal level in general, while capacitors cap the signals and just flaten out the peaks.

PPS: and verify that both your systems (AES & MD) use the proper sync line for RGB sync instead of composite video.

[Billkwando]

The problem is certainly (or very likely) not the iScan. Just grab any other component source to verify that it works as it's supposed to. Have a PS2 on hand ?

I do indeed! I assume I need to use a PS1 game to get it to output 240p?

The AES is very easy to work on, so start with this one to work out the problems. What you're measuring is the voltage peak level on the RGB and the sync lines. With a white flash on screen (or a white transition) the voltage level should peak at 0.8V max. (0.7V is the standard consumer peak level). It's normal for the sync line to peak a little higher. Arcade systems can go up to 3+V and some AES and MD/GEN revisions certainly hit the 2V level without the proper attenuation.

The iScan loses the signal because the the inputs are "out of bound" (voltage levels too high). The difference in behaviour (pure grey vs. additional flicker) is probably caused by what follows the voltage peaks.

Awesome. I'll give that a try. I was wondering, since it's the component signal that feeds the iScan, is modifying a component cable a potential option for correcting the problem, or does it have to be on the RGB side? Is it possible that the voltages sent from the YUV converter will be unaffected by changes made to the RGB side?

Also, for testing it with the bright flash, is there a way you'd suggest that I can watch it and test with a multimeter at the same time? Can I unhook the scart plug and stick the probes in there, while listening on headphones for the parts that have bright flashes....or does it need to be connected?

Once you understand why the NG drops out and how to fix it, it's very easy to apply the same fix to other systems.

I was wondering, assuming this is a universal fix, if maybe I could put the resistors or caps into my scart breakout box/stereo pass-through adapter? I haven't tried it with my Saturn or SNES yet, but I'm assuming it's going to do the same thing, lol.

The only question mark in your particular setup is that transcoder box. Before you start soldering you should make a table to note down which input voltage levels (on the console's output) correspond to which output levels from the transcoder.

I've just tried measuring the output from the component cable that goes to the iScan. Maybe it's just my crappy Radio Shack multimeter, but it seems to be saying that the green line is putting out around .012 volts??

Also, the box has some knobbies inside that adjust the color and various other things that nobody knows what they do. I have a thread about that here: http://www.neo-geo.com/forums/showthrea ... bleshootin

I'm wondering what they do and how they might factor in. I'm gonna try to find out more info.

PS: resistors lower the signal level in general, while capacitors cap the signals and just flatten out the peaks.

Thanks for putting it into layman's terms for me. That helps a lot.

PPS: and verify that both your systems (AES & MD) use the proper sync line for RGB sync instead of composite video.

Both my MD and Neo cables are from here:

http://www.retrogamingcables.co.uk/snk- ... -lead.html

Do you happen to know how he wires his cables? I assume some folks here use them.

[Fudoh]

Shouldn't matter if you use a PS1 or PS2 game, 240p or 480i. Doesn't affect the signal per se...

You cannot modifiy the component end, since it uses color difference signals (meaning the three colors are calculated from two signals and those calculations are not linear). If you alter those you change the color composition.

I'd just try to find a game which "drops outs" during the intro. This way you don't have to actively do anything. You can just watch the multimeter for 30 seconds or so. Levels change slightly when connected, but you can measure it unconnected.

The type of fix is universal, but the values of the required resistors vary. It's also unlikely that you require a mod on all systems. Saturn's levels are pretty standard (again assuming you're using the sync output of the system, not the composite video signal), PS1/2 with RGB should run flawless already.

but it seems to be saying that the green line is putting out around .012 volts??

that doesn't say anything. You can just use it to compare it to a measurement AFTER modding the RGB output.

If the cable doesn't explicitely say "raw sync" it doesn't use it. But it's an easy mod, since both systems already output raw sync out of the box. NG depending on the board revision and MD requires a capacitor in line with the sync signal. Just for example: on the XRGB-3 both MD and NG caused dropouts like this. Using the proper sync signals and decreasing the RGB levels made both systems work flawlessly on the XRGB.

[Billkwando]

Shouldn't matter if you use a PS1 or PS2 game, 240p or 480i. Doesn't affect the signal per se...

You cannot modifiy the component end, since it uses color difference signals (meaning the three colors are calculated from two signals and those calculations are not linear). If you alter those you change the color composition.

I'd just try to find a game which "drops outs" during the intro. This way you don't have to actively do anything. You can just watch the multimeter for 30 seconds or so. Levels change slightly when connected, but you can measure it unconnected.

The type of fix is universal, but the values of the required resistors vary. It's also unlikely that you require a mod on all systems. Saturn's levels are pretty standard (again assuming you're using the sync output of the system, not the composite video signal), PS1/2 with RGB should run flawless already.

but it seems to be saying that the green line is putting out around .012 volts??

that doesn't say anything. You can just use it to compare it to a measurement AFTER modding the RGB output.

If the cable doesn't explicitely say "raw sync" it doesn't use it. But it's an easy mod, since both systems already output raw sync out of the box. NG depending on the board revision and MD requires a capacitor in line with the sync signal. Just for example: on the XRGB-3 both MD and NG caused dropouts like this. Using the proper sync signals and decreasing the RGB levels made both systems work flawlessly on the XRGB.

Did you watch the videos? I'm experiencing 2 different types of dropouts. The flicker/squiggly lines one that happens during Waku Waku 7 and the Neo startup, and the MAJOR one, on Neo and MD, which happens at random (normally after the system's already been running a long time) where the picture drops out completely and stays gone for minutes. There's no way to anticipate when that's going to happen.

The lesser dropout, according to the n-g.com thread I linked above, is supposed to be a result of the caps and resistors in my revision Neo. It's a 1st gen/ 1st revision according to that same thread.

[Fudoh]

see it this way: if you got another component source that works fine (which I assume you do) it eventually always comes down to "recondition" the signal to a degree that the line doubler will handle it. With the systems you have on hand, that's extremely easy, especially since both offer a sync output.

I didn't have an AES for a very long time, but I always got a MD in my test lineup and that into the iScan Pro is usually not a problem. I use another color transcoder though.

is supposed to be a result of the caps and resistors in my revision Neo

that's very well possible, but you can definitely counter that by modding the cable. It's an extremely common problem. That's why SNK added the flash off function to many of it's later fighting games.

[Billkwando]

OK, this issue has officially gotten freaky.

The picture is dropping out with Neo Geo, Genesis, and Saturn...all using different cables. No big surprise so far.

Here's the weird thing. It always drops out for 2 minutes and 50 seconds, then comes back. How crazy is that? I've timed it, with the stopwatch on my phone, and least 5 times now.

I also noticed, while the screen is blank, the iScan's picture adjustment knobs on the front have no effect on the screen (which always displays a blank single-color screen that is gray, black, blue, or red...usually a color that was present in the image at the time it dropped out.

I've also noticed, while the screen is blank, vertical pinstripes that look otherwise identical to scanlines, to the point that if I turn on my SLG HD, it makes a perfect grid onscreen.

This is just so crazy weird....does this ring any bells or give you any ideas?

Edit: Apparently the pattern is this: it plays for 7 minutes and 50 seconds, then drops out for 2 minutes and 50 seconds...lather, rinse, repeat.

Edit 2: I just went back and checked the youtube example video I did, and sure enough, there it is: http://www.youtube.com/watch?v=f6bU00fg5Pk

[Fudoh]

have you done anything yet to counter the problem ? Tried a direct component source to verify the iScan is ok ? Took on one of the RGB signals yet ?

[Billkwando]

have you done anything yet to counter the problem ? Tried a direct component source to verify the iScan is ok ? Took on one of the RGB signals yet ?

Just did a definitive test on PS2, and yup, it's the iScan. Was playing Fantasy Zone, now I'm looking at a yellow screen. I'd tried it before w/ KOF 2006, but it kept kicking into film mode, which must've been stopping the cycle from beginning. I'm timing it now with Golden Axe (using the Sega Classics disc obv).

Edit: Ya, it just went gray on me in Golden Axe too. Seemed to take 8 minutes this time, but maybe I was just slow on the stopwatch. aaaand now it's back. This is a major bummer.

I'm guessing this won't be a simple repair.

[Fudoh]

that's a shame. There's likely nothing you can do about it, if it's the iScan itself. Possibly a problem with the video deinterlacing mode, but that're all highly integrated ICs.

The only thing I would try before throwing it out: replace the PSU. Possibly the capacitors in the PSU are failing and deliver a non-steady voltage supply. What kind of PSU do you use right now ? The original 6V one ? You can use a generic 5V one, if you increase the maximum amperage a little bit.

90% of the times anything goes wrong with an iScan unit, it's the PSU.

[Billkwando]

Whoops.

I tried a 5v 2.5 supply and just got the same blank screen as shows every 8 minutes, but right away.

[Lord of Pirates]

What sucks the most is it was a no refunds auction. Bad feedback headed their way, lol.

You could still open a case if it was listed as working, though I'd try to work it out with them before escalating it.

[Billkwando]

I ordered one of these:

http://www.amazon.com/Super-Power-Suppl ... 2a+adapter

Hopefully that works. I did find it somewhat encouraging that while using the wrong power adapter (the 5v 2.5a one) it gave me the same type of screen it does when it drops out. I just find it odd that, when using it with the original PSU, it does it with such precise timing.

Any ideas why that would be or is it beyond your scope of experience?

[Fudoh]

I'm as buffled as you. Let us know how the new PSU works out.

[Billkwando]

I'm as buffled as you. Let us know how the new PSU works out.

So the power supply came, and it doesn't work. The red light just flashes. It's center pin positive 6v 2a...no idea why it's not working.

I found the same model number power supply that came with the iScan on ebay (no mention of iScan on the auction though), but I'm leery of buying a used one because if it was used on an iScan, it's probably seen hard use, since they run 24/7. The back of mine says it's a class 3 switching power supply and "+6V===2.0A". Unless the plus means something, as far as I know, that's exactly what I got. When I connect the original PSU, the red light goes back to solid.

This whole experience has been a new level of horrible for me (at least in the context of electronics), due to my obsessive nature and need to have everything set up perfectly. It would be funny if it wasn't so frustrating.


Edit: I went ahead and grabbed the one off ebay: http://www.ebay.com/itm/130949224872?ss ... 2134wt_758

It says that the estimated delivery date is Monday, so I'll have to wait in suspense til then. I really wonder why the amazon psu didn't work. Even the underpowered 5v one I tried didn't cause the light to flash like that, so it's pretty mysterious. Maybe the AC adapter was just plain defective.

Here they are side by side:


Power Struggle by billkwando, on Flickr

The square one is obv the original.

Edit: I tested the Amazon psu with a multimeter this morning, and it is putting out 6v, center pin positive, as the label states. Xian Xi from the Neo Geo forum thinks the iScan PSU is using some weird proprietary frequency maybe (the 47-63hz thing), but didn't you say you've used them with a generic psu before and it worked fine?

Also, sorry about the massive pic....when I posted it on the N-G forum, it resized automatically to fit.

[Billkwando]

Bump?

[Fudoh]

there's really nothing to add. There's no reason why the PSU shouldn't work work. The specs are allright and I've used a bunchload of different PSUs with iScan units - as said even 5V ones. The PSU does output DC, so the frequency is gone by then. You should see on your multimeter that both PSUs output the exact same thing.

Have you tried connecting the PSU to the outlet first and only then into the iscan ? Sometimes those generic PSUs are too weak to be powered up directly with the machine they power.

[Billkwando]

there's really nothing to add. There's no reason why the PSU shouldn't work work. The specs are allright and I've used a bunchload of different PSUs with iScan units - as said even 5V ones. The PSU does output DC, so the frequency is gone by then. You should see on your multimeter that both PSUs output the exact same thing.

Have you tried connecting the PSU to the outlet first and only then into the iscan ? Sometimes those generic PSUs are too weak to be powered up directly with the machine they power.

The only difference I noticed was that the generic psu put out a little less than 6v and fluctuated, whereas the official psu put out a steady 6.14v. I'm guessing the official one is regulated and the other one isn't. No idea why it's not working. I'll report back when I get the next replacement. Should be by tomorrow.

I did try plugging and unplugging it from the iScan when I tested it the first time.

[Billkwando]

So it turns out the problem is not 100% solved, but a lot better than it was before. One thing I am doing differently is not plugging the iScan in til I'm ready to use it, so that may have some effect on my testing.

With the "new" psu I hooked it up and it ran for 45 minutes before cutting out. When it did cut out, it was still gone for 2 minutes and 50 seconds, which leads me to wonder if it's not some sort of failsafe or protection circuit. I've had it go @ some odd times now. The 2nd time it went out was after 15 minutes. It's pretty odd.

I'm going to try ordering a 6v 2.5a ac adaptor, cos that's highest I've been able to find on Amazon (amps wise), and I like their return policy. I won't be able to mess with it til sometime next week, though. If anybody else knows of a good place to get AC adapters, let me know.

The seller had also contacted me to work out some kind of arrangement. I'm just gonna ask them to refund me the $18 I spent on the original iScan PSU I picked up, and muddle through until I find a fix. At least the drops are a lot less frequent.....and now I can take those recommended breaks the game manuals always talked about. LOL!

[Billkwando]

So the problem wasn't the psu....but now I have 2 spares!

The issue was overheating.

Here's the iScan board:




The metal plate with the holes is the heat sink, which is attached to a regulator. When the board is removed from the case, the picture doesn't drop.

What I did was slap a fan on there (unfortunately I ordered the wrong size, and haven't bothered to replace it, cos it works):




I just found the nearest 5v source on the board and hooked it up there. I cut some vents in the top and bottom of the case with a dremel cutting wheel.

I ran Burning Rangers for over 3 hours, yesterday and today, with no problems:)

[Fudoh]

Wow, what a journey! Gratulations on fixing it!

I'm wondering though why I never ran into similar problems. Possibly the voltage regulator on your your unit is running out of spec and simply getting hotter as it should (compared to other iScan Pro units).