shmups.system11.org

No, it doesn't really mean anything. You'd expect Line A and B to read the same but CVBS isn't a pure DC or AC signal so it's hard to draw a conclusion.

Think about DC vs AC. 5V battery is 5V DC. Can measure 5V on multimeter no issues. A 5V sinusoidal AC signal, with no DC offset, will measure 5/sqrt(2) ~3.5V in AC mode because it measures Vrms. Basically is the DC equivalent so if you had a light that glowed based on voltage, 3.5V DC = 5V AC = 10Vp-p = 3.5 Vrms. The p-p means "peak to peak" that you would measure with an oscilloscope from the highest voltage level to the lowest, including negative. So a 5V sinusoidal oscillates between +5V and -5V = 10Vp-p. I have to mention because I see PVM manuals use this figure.

My PVM manual defines Composite video as such that it takes on Line A and B: 1.0 Vp-p +3 dB/-6 dB, sync negative
The +3 dB/-6 dB is the error tolerance so the signal can be 1.4 to 0.5Vp-p and work, assuming no other issues.

Composite is a complicated video signal that carries luma, chroma and sync on one wire. Basically why it's blurry. There's no way you can get meaningful measurements with a multimeter or with Component for that matter. The sync is two square waves, the chroma is contained in a sinusoidal colorburst of its own subcarrier frequency and luma is amplitude modulated. The 1Vp-p comes from the max value of luma being +0.7V and vertical sync of -0.3V. The only meaningful way to look at it is with an oscilloscope.

But sure, -2V is an impossible value from a signal that goes from -0.3V to 0.7V. Probably a bad sign and could still have a capacitor problem but it's not proof.

Oh wait, you mean you're measuring Line A and B with the PVM turned off. Still doesn't mean anything but different reasoning. I haven't tried to mod or repair a CRT but every guide I come across warning about shocking yourself from capacitors. Can take days or even a week from power off, from what the guides say, for them to dissipate all their energy.

Skipping over explanation of RC time constants, I think that's your problem. If you played Composite on Line B a few days from now and then measured, I think you'd get -2A from there and 0V on Line A. I guess checking that the voltage is decreasing over time at the same (exponentially decreasing) rate from each Line would actually be a decent test if a capacitor is dodgy because the time constant would be different or non-constant.

As to why it's negative voltage, it's about your frame of reference. Negative voltage means it's flowing the opposite direction from your frame of reference. If you opened the CRT, could find a point to measure +2A instead. I would expect charge to flow away from you when viewing from outside.

If you open the CRT and measure things, which is not a beginner task, you'd probably want to use a special ESR meter to determine if capacitors are near end of life. If they've burst and leaked fluid then, well, no need to measure just replace.

@NewSchoolBoxer

Thanks. Just to clarify, the PVM was turned on but there was no input connected. I had a BNC -> RCA convertor on the input of the PVM, and an RCA->RCA lead connected. I measured the voltage, with my multimeter ground probe on the outer of the RCA plug, and the positive probe on the center. I thought the negative interesting enough to mention, but that's all.

I didn't expect it to mean anything. I'm pretty much clutching at straws. I had been measuring the voltage at the video output of the Famicom. I am aware that composite video is not DC, but I just was checking it wasn't wildly wrong - like I say I was clutching at straws trying to figure out why my two AV mods both starting failing after a similar amount of time. While checking this I noticed the non-zero voltage reading after I had turned off the Fami and thought it a bit strange.

Yeah sure! I like your testing approach without an oscilloscope available.

I assume since every pre-HDMI console outputs Composite that you've tested if other source has video degradation after 1 or 2 hours and on Line A and B. Just that you modded the Famicoms leaves room open to speculate. The transient response of circuits tends to be well under 1 ms so that isn't the problem. Taking an hour means something like the console heating up too much due to faulty voltage regular or heat sink. Or was fine but mod put extra heat out by consuming more power. Or console is fine but audio or video signal is over volted due to amp and on edge of what PVM can handle.

That you replaced one capacitor suggests you know what you're doing. Maybe the service manual exists online and the circuit diagrams can narrow down the capacitors you need to check. I haven't heard of a voltage regulator going bad in a PVM but definitely happens on SNES.

One device I like but lost in my move is Kill-A-Watt that measure the power used of electrical devices, even when they are turned off. More features than that but can compare stock Famicom or NES vs modded if you rule out PVM being the issue, as in, it's only Composite from the Famicoms that degrades.

If I incorporate a lm1881 based sync stripper into the output of my scart switch, will it pass through csync without affecting it?

I have 4 consoles connected to the switch that are csync and 2 that are sync on luma. Looking to clean up the sync on luma signals without affecting the csync'ers.

Sent from my SM-G955U using Tapatalk

I would say probably fine. Try it first though. https://www.ti.com/lit/ds/symlink/lm188 ... 1582111958

look on page 3 for the waveforms

Once all the parts come in, my signal chain will look like the following:

Raspberry Pi HDMI -> Portta HDMI to VGA converter -> JVC TM-H150CG (RGBHV input card)

Now I'm wondering, is there the possibility that I might damage my monitor, by sending it the wrong signal? The JVC has 15kHz horizontal sync, so it will handle 240p/480i just fine. But what happens if I accidentally try to send through a 1080p signal?

Will my monitor recognise that it's out of range, and actively refuse to even try to draw the picture? Or will it try to draw the 1080p picture anyway, and potentially damage something internally due to attempting to perform out of spec?

Yes, you can damage your monitor.

But switching over the Pi to 15khz is rather easy, since you can leave it on a HDMI display while switching the UI to 480i. Then you can switch all the emulators to 240p and test it on your HD display first (even just to verify that you don't get a picture when launching a game, but get back to the 480i menu when you quit out of it). And once you're done switching everything over to 15khz only then you can connect the Pi to your JVC.

Hmm that's a little concerning. Do you know how bad it would be? E.g. instantly damage when fed the wrong signal? Or perhaps only if it's high enough (1080p vs 480p say), and attempting to display it for long enough (a couple of seconds, or a few minutes maybe)?

I do intend to ensure that only 480i/240p ever gets sent over. In fact, that's basically how it's set up now. However just yesterday I installed Kodi on my pi, and when I booted it for the first time it unexpectedly opened in 1080p. Thankfully right now I'm using a HDMI to S-Video converter and it scales the signal down to 480i before outputting, so no harm was done. With the Portta HDMI to VGA converter though, no such scaling will take place.

I'm sure that kind of scenario will be rare in the future - especially now that I know it can damage the monitor, and these monitors are getting rarer and more expensive, I'll try to be extra careful. I guess I want to know how careful I have to be. If I accidentally send a 1080p signal through, will I have a few seconds to change the signal (or even quickly switch inputs on the monitor) before irreversible damage is done? Or is it more like, send a 1080p signal through and insta-smoke out of the back?

More electrical engineering questions, I'm kind of excited.



The second paragraph of the chip's datasheet states that the sync signal gets delayed between 40 ns and 200 ns. "This much delay will not usually be significant" if you don't add in further signal processing and you aren't. Still not ideal not to run csync into it but video quality shouldn't be affected.



This is like the 3.3V knockoff flash carts and X in 1 games on SNES where the correct voltage is 5. Everyone agrees that it could damage the SNES but no way to say how likely the damage is. 1% chance per hour? 0% chance until X part is near end of life? Is 2chip more resistant than 1chip model that had cost cutting measures? Not going to test and find out.

I can tell you that I tested around with 480i -> 480p with PS2 Rock Band progressive mode. My L2 PVM video split video into 2 parts on the screen at 480p but could run video out to SoG computer monitor that had clear video. 480p on consumer CRT scrambles video as expected. No damage but I only tried for a minute. Yes, more risk the higher the resolution.

As to if the damage would occur, you'd have check the +/- dB voltage range on the sync signal, study the video processing circuit on the specific CRT model, see what transistors and voltage regulators are being used, try to simulate the circuit etc. Square wave sync signal is like DC but the transitions have an AC frequency response. Basically, 480i uses 15 kHz horizontal sync, analog 480p / 1080i doubles that to 31 kHz and 720p / 1080p triple or quadruple. When you increase the frequency, you increase the impedance, in a linear amount in a simple circuit. You therefore *decrease the current by that amount. The voltage drop is higher so you may be under the acceptable range. Transistors may have a certain minimum base current to operate as well. Similar under volting issue to SNES then. Low risk but not 0 risk.

Are there any advantages or disadvantages of installing a Unibios if you've already got a Neo SD Pro?

Very highly doubt the csync will be affected. I've ran csync through tons of different stripping cables/devices with no issue (the exception being my old RGB-Pi, its 12v signal being the cause. But that thing is a GPIO hat, not a game console cable!)

Thanks for the guidance, folks. I incorporated the sync stripper in and it fixed my issues. Tested outputs with crt and ossc - good all around...with one exception. Output from my supergun no longer worked. It was producing the same type of scrambled image that the two sync on luma devices were producing prior to incorporating the stripper.

The supergun is a ali express $40 special. It's actually quite good. I havent had any complaints with it to date. Upon examination, I found the supergun has a ths7314+lm1881 solution on board. I tried bypassing the lm1881 on the supergun altogether, and no difference. I finally ended up using the mini din output on the supergun instead of the scart output, and it works. I cant find a good reason why it works, as all significant pins between the two are continuous, no extra or different circuitry one way or the other. But, it's working so I've stopped thinking about it. There's always gotta be at least one fly in the ointment. :/

Sent from my SM-G955U using Tapatalk

I recently installed voultar's PC Engine RGB board into a Duo-R and it seems to work great, except on my Sony PVM 14-L2 at the top there's some warping. This doesn't happen for my other systems like SNES. It also doesn't happen with the PC engine hooked up to my 20m2mdu. So I did some searching and found some threads here about sync problems that seem similar on BVMs and Ikegami monitors being fixed with an Extron RGB's SERR feature. So a couple questions about this:

Can this same sync issue happen on some PVMs like the 14-L2 too?

And if so, can the csync at 75ohm from a PCE RGB board be used on an Extron device with a passive scart to db-15 cable, or do I need something like a sync strike?

Any help is appreciated.

Does your monitor have a "VCR mode" setting somewhere within its menu? That usually fixed warping/flagging on the top of the picture from PC Engines modded for RGB.

Hooked up via the regular built-in hookups, or the optional 129x card? 129x cards can have some funkiness so definitely try it via the regular connections. Myself personally, and YMMV, I would start out by resetting the 14L2 to factory (how to do this is in the manual, write down your values beforehand if you want – and you will have to redo your geometry). I would do this just in case somebody messed with a setting they shouldn't have.



Extron requires clean sync so as long as that's kosher looks like it would be fine. TG16/PCE is one of the consoles with sync issues and I commonly run it through an Extron RGB with SERR, but like you said usually only has issues on my BVM's (PVM's are usually fine).

Also keep in mind that SCART to db15/VGA is wired differently than db15/VGA to SCART so make sure you get or build the correct cable.

I don't think pvms have the vcr mode sadly.



It's hooked up through the built in BNC connectors, don't have a bkm-129x unfortunately. I'll go ahead and buy an extron, because the picture adjustment knobs will be useful even if it doesn't fix the PC engine problem. And if that doesn't fix it I'll reset to factory settings.
I was planning on building my own cable using the pinout on the left here.


Thanks for the advice on this stuff guys.

When I replace a CRT tube with a LCD screen inside an existing chassis, what I can use to fill up the gaps between the new flat screen and the existing (curved) bezel ?

A couple of suggestions for you.

-check to see if the mod lets you switch between composite sync and c-sync

-if it's a sync issue, you would fix it by using a sync stripper, you wouldn't need the extron. in that case, buy a cheap device that has an lm1881 on it. you can even just buy a PCB from OSH Park and build your own. they are dead simple

-the extron might not work without clean sync anyway (some of them are more tolerant than others)

-if none of that works, try sync on green. raphnet has an article on it, but basically just take clean sync and combine it with the green through a 100uf cap and the green through a 680 ohm resistor. google raphnet sync on green for details. you need to clean the sync before that will work.

ideally, you build a bracket with some wood (plywood or MDF). if you don't have the tools for that, you could try either a putty that will fit in there or some foam.

if you use foam, try to get something that won't break down too quickly or turn to goo. Silicone foam strips are good for that, but a little on the expensive side. We have a mail order company in the US called McMaster that people order stuff like that from, and it let's you go through and pick the density of the foam and everything. You would want something on low density side, like medium-low. If it's too dense, it won't compress around the shape of the bezel.

What's the best way to play Wii games through a Wii U? Should I set the output resolution of the Wii U to 480p and use analog component cables, or should I just set to my display resolution (1080p) and let the Wii U upscale over HDMI?

I would think digital to digital would be best, so skip analogue output. You will need to be the one to determine whether letting the Wii U or your TV upscale to 1080p looks best to you.

For a 6ft run of component video, if I use your "next best" cable (let's say a 50 ohm audio cable), what's the expected loss compared to a proper 75 ohm coax cable ? Is there really any ? After all I think that proper (but flexible) audio cables are still better than the majority of original OEM console component cables.

My reason for asking: for test setups I can't stand those strong and unflexible component cables. For short and temporary connection I would prefer slim and flexible cables, but of course those aren't "made" for component video.

Signals are limited to 15 and 31khz, so no HD tranmissions via those.

Loss isn't so much the concern as reflection is. It probably won't be a big deal at 15kHz but might become noticeable at 480p.

Why not get something like this:
https://www.amazon.com/HD-Retrovision-YPbPr-Component-Video/dp/B07KRKM96K

flexible and 75 ohm

because I assume that the middle part isn't as flexible as I need it to be.

It's not quite that simple.

The "quality" of the cable is more about the construction of the cable than it is about the impedance. In general, 50 ohm coax is always going to thicker, stiffer and higher quality than 75 ohm coax. It's lower impedance because it's physically bigger. So, for line loss, lower is better.

However, for video signals, the equipment is expecting the cable to be 75 ohms. This sounds backwards but the way that it works is that the circuits are actually designed to operate with a specific load impedance.

If you were to graph the effect of line impedance on the signal, you are looking at signal degradation VS total power throughput. 75 ohms was chosen for video because it's a sweet-spot for signal integrity. 50 ohms is generally used for power delivery and such.

So, as maxtherabbit stated, it's about reflections. What happens is, every time you have an impedance mismatch, you lose roughly 5% of your signal. So that's 5% on both ends of the cable.

What about coax vs. stranded wire in general. The majority of original console cables are stranded, no coax after all. This is mainly where the flexibility comes from, right?

More or less. Stranding is definitely important for a number of reasons, including flexibility, but I have found that the choice of jacketing material plays a much more substantial role in the flexibility of the finished cable than does the amount of standing.

But if we are talking just solid coax VS stranded then, yes, the solid core coax is going to be stiffer. I'm sure you've used that cable TV stuff. It's hard to make that bend.

The trouble is that you can't really know how flexible one jacketing material is compared to another without trying it, unless they declare exactly what material they are using.

If you can find a cable that says it's got a silicone jacket, that has more likely chance to be flexible. That's like the white jackets that Apple uses for the iphone cables. It's very soft and flexible.

well I'm not exactly sure how much flexibility your require, but it's not as stiff as it looks

worst case you could just get some individual lengths of 75 ohm mini-coax and tip tie them together

I would like some cables that don't push around lighter gear (like an OSSC or even a Portta converter) when the cables are being moved.

This has been bothering me for decades now. I don't care for any permanent setup, but for any test setup with little space, it's really annoying. Recently a bought a bunch of cables and among them was a 6ft audio RCA to 3.5mm cable that's so flexible that it just falls in place whichever way you move or place it.