BVM-A20F1M Sega consoles Sync problems

[jwrose]

Huh. Yeah mine is an NTSC US console playing NTSC US games, with exception to Project J2. The BVM handles both PAL and NTSC.
I had the simple RGB mod performed since the N64 is an earlier unit.

Everything looks great except for those 2 titles that switch resolutions- and it’s not even the entire game- just the higher resolution screens.

[PeterWar]

Running it through the Extron RGB fixes it (just like with the AES and PCE)..

Could you please provide the name of the Extron RGB unit that fixes this issue? I presume you use it to fix the skew issue of the AES too?

[jwrose]

Yep! It’s the Extron rgb 203rxi. I use the SERR feature. It fixes AES, TG/PCE, Perfect Dark n64. It doesn’t fix SMS.

[korpskytten]

I have a found a solution for SMS. I am using a japanese Master System.




I am using two devices from Arcadeforge. Sending the signal from the SMS through the Sync strike, then generating new sync with the UMSA.
Japanese SMS > Sync strike > UMSA > rgb-scart-cable to BNC > BVM D14H1E.

I will test this on my X68K soon.

[jwrose]

Interesting! Any chance you could post an image of that chain? I’m just curious to see what cables are being used between everything since it looks like each of those devices only had a single Scart socket.

[korpskytten]

I am not at home right now to take a picture, but I made a quick sketch:



Sending the signal through the syncstrike only almost works, image is correct but with tiny flickers. The UMSA makes it 100% stable and clear.

[jwrose]

Got it! Thank you! That clarified my confusion- so, you use VGA to link them. I may have to check into a USMA and Sync Strike. I hadn’t heard of USMAs before. Anything I should know about buying them? (I.e. reputable retailers, knockoffs, etc)

Thanks again!

[korpskytten]

here is the link:
http://arcadeforge.net/UMSA/UMSA-Ultima ... r::57.html

ArcadeForge, based in Germany I think.

[RGB0b]

So, this is essentially converting RGBs to RGBHV. then back from RGBHV to RGBs. I'm surprised this works, but if that's the case, we can probably design a board that does exactly this and outputs via a dsub connector, so you can use a really cheap "VGA to BNC" cable to connect to your monitor. We'd want to verify compatibility with all consoles, as well as double check that this isn't spiking the sync voltage too high.

I also have another solution being mailed to me that's testing a different way of changing the sync. If that works, it'll be easier than this. I'm glad people other than me are still testing and experimenting though!!!

[jwrose]

I also have another solution being mailed to me that's testing a different way of changing the sync. If that works, it'll be easier than this. I'm glad people other than me are still testing and experimenting though!!!

Great! Glad to hear you’re still looking into this! I’ve been waiting in hopes of a permanent solution. I’ve been holding off for awhile on getting my SMS recapped just in case a mod became available. Keep us posted on what you figure out!!

[korpskytten]

We'd want to verify compatibility with all consoles, as well as double check that this isn't spiking the sync voltage too high.

My AES is a low serial Pal version with no issues so I can't test that. I have no chance to test a PCE at the moment. How can I check the sync voltage?

[PeterWar]

We'd want to verify compatibility with all consoles, as well as double check that this isn't spiking the sync voltage too high.

My AES is a low serial Pal version with no issues so I can't test that. I have no chance to test a PCE at the moment. How can I check the sync voltage?

My guess is that you'll want a cheap multimeter (many avaliable on amazon), I would select the 20V continous mode, put one end into the BNC sync and ground the other one.

Have you tested with the BKM-68x yet? I'm eager to learn about the results, I'm in europe and could order the UMSA+Sync strike soon for further testing if proven to work.

[RGB0b]

No, you can't check video signal voltage with a multimeter, you'd need an oscilloscope. Here's all the info I have and the first video with Ste should explain it all: https://www.retrorgb.com/oscilloscope.html

[korpskytten]

Ok, sorry I don't have an oscilloscope at hand.

Have you tested with the BKM-68x yet? I'm eager to learn about the results, I'm in europe and could order the UMSA+Sync strike soon for further testing if proven to work.

I have got a BVM A20 but no BKM-68X. What I said I would test was the Sharp X68K / X68000. It is known for strange sync signals and non-compatible with BVM-series in general. I have tested THAT a bit and it seems to work well so far. But the sync differs from different games so I will make some more testing on this.

[willis936]

Does anyone have experience with HV no longer turning on after a bad sync on a D14? I put bad signals into Skum’s 129X from an Easycel HDMI to component adapter. The monitor has 6000 hours on it and was manufactured in 2002. No burn marks or smells, no dust, no bulged caps, no irregular junction voltage drops, no blown fuses. On power on the neon bulb across the FBT flashes twice. I think the free running horizontal frequency is out of spec and trips a protection circuit, causing the HV to shut off (the second flash).

I’ve tried unplugging all boards, removing the M assembly (ISR) watch battery, discharging the parallel cap, plugging everything back in, then powering on.

I’ve tried different combinations of input cards being fed valid signals.

I have attempted factory resets. I am pretty sure I am getting into the password-locked maintenance menu because it takes an extra second to power off after I use one of the password candidates (1111) compared to the others (0000, 9999). Once in the maintenance menu I scroll down 5 times and hit enter a few times. Does anyone know if this is the correct procedure?

Sony won’t give me the serial interface software without filling out a form that includes the vendor the monitor was acquired from. They’re clearly saying they won’t help secondhand owners.

[korpskytten]

News on the SyncStrike + UMSA. PCE tested with perfect picture result. Also added a picture of the two devices connected.

[PeterWar]

News on the SyncStrike + UMSA. PCE tested with perfect picture result. Also added a picture of the two devices connected

Nice, I've seen no sync issues with my THS7374 RGB moded PC Engine DUO XR and BKM68X but it is nice to know that it is also possible to play the first PCE console too.

[jwrose]

Nice. I may pull the trigger since this seems to fix SMS, PC Engine, and (likely) Neo Geo AES. Currently I’m using an Extron RGB, but that only fixes PC Engine and Neo Geo.

[korpskytten]

I have now also tried it with an Extron crosspoint and the signal goes through it without problems. So you can have the SynkStrike/UMSA before the crosspoint in your setup.
Only issue I have got now is that the UMSA only takes 15khz signals, so I will still have to use the 203 Rxi with the SERR function AFTER the crosspoint for 30khz/24khz signals from the Sharp X68k and the Naomi. Does anyone have experience with other component Matrix-solutions like the ones from Kramer in combination with the Extron 203 Rxi?

[RGB0b]

I just tried something kinda neat:

RetroTINK 2x Pro or 2x M in Passthrough mode -> *HDMI to VGA converter -> Extron 203Rxi with SERR turned on -> monitor

That actually worked with an added bonus: composite, S-Video and component video all passed through. Composite through the M looked good too, cause of the comb filter. I forgot to test 480p through the M though, but it should work for component video sources.

We're still working on the other solution, but I wanted to pass that along in case anyone else happens to have those same components and can try.

*I just used the cheap HDMI to VGA "cable" I have linked here: https://www.amazon.com/shop/retrorgb?li ... 72N6BL06SV

[jwrose]

I have now also tried it with an Extron crosspoint and the signal goes through it without problems. So you can have the SynkStrike/UMSA before the crosspoint in your setup.
Only issue I have got now is that the UMSA only takes 15khz signals, so I will still have to use the 203 Rxi with the SERR function AFTER the crosspoint for 30khz/24khz signals from the Sharp X68k and the Naomi. Does anyone have experience with other component Matrix-solutions like the ones from Kramer in combination with the Extron 203 Rxi?

Was this with a Master System?

[RGB0b]

Yes

[jwrose]

Yes

Sorry, that was meant for korpskytten- but I do have a question on yours - did it come through at 240p for RGB?

[korpskytten]

Was this with a Master System?

Yes. Japanese MS and PCE.

[jwrose]

I don't think I posted this before, but has anyone else with an A or D series had issues with FMV sequences on an RGB modded N64 ? Its pretty much just like the AES or PC Engine issue since the SERR setting on my Extron RGB fixes it (although causes some issues for non-FMV parts that work fine normally).
Perfect Dark has this issue with its intro and RE2 has issues with its FMVs and the inventory screen about 80% of the time. Anyway, just curious if others have experienced this. All other games I've tossed at it work 100%.

[fernan1234]

I don't think I posted this before, but has anyone else with an A or D series had issues with FMV sequences on an RGB modded N64 ? Its pretty much just like the AES or PC Engine issue since the SERR setting on my Extron RGB fixes it (although causes some issues for non-FMV parts that work fine normally).
Perfect Dark has this issue with its intro and RE2 has issues with its FMVs and the inventory screen about 80% of the time. Anyway, just curious if others have experienced this. All other games I've tossed at it work 100%.

These problems may be rooted in your N64's RGB board. I came to this topic with my own BVM A-series issues, but saw this post and tested the two games you mentioned. On a BVM-A24 with BKM-68X, I did not see these issues using a console with an N64Advanced board installed for RGB output.


As to what brought me to this topic, which seems to be about A-series sync woes, is that I just found out that apparently a lot of cores in the MiSTer completely fail to sync properly with the A-series BVMs. Most console cores work well, but consoles cores like Neo Geo do not sync properly (the monitor displays a PAL signal detection and the picture rolls vertically). All arcade cores that I tested only display a garbled picture, like when a 1080p signal is input. On the other hand, the SMS core does not have the top bend distortion that it has using original hardware on this monitor. HDMI direct video, IO board VGA RGBHV or RGBS, YPbPr, none makes a difference, the problem is with the sync of these cores, not the input format.

I was surprised that the Neo Geo core in particular has this problem as I recall that original Neo Geo hardware does sync with these A BVMs.

I also have another solution being mailed to me that's testing a different way of changing the sync. If that works, it'll be easier than this. I'm glad people other than me are still testing and experimenting though!!!

We're still working on the other solution, but I wanted to pass that along in case anyone else happens to have those same components and can try.

Any updates on this solution you were working on last year?

I'm also wondering if regenerating sync using a SyncSlayerII and then recombining sync with an Extron RGB interface may help, similar to how the SyncStrike + UMSA setup helped korpskytten, though that seems to be a fix for the top flagging issue in particular. I may get a SyncSlayerII to test soon.

If anyone has any new ideas or experiences, please share!

[daty2k1]

I think the Neo Geo MiSTer core is not working because it's based on the main clock from the MVS if I remember correctly.

I have made a modified version which is using the AES clock which you can find here:
https://github.com/daty2k1/NeoGeo_MiSTer

It is not completely up to date but you can check if this is working for you. It is working fine on my A20 with this build.

I have recently purchased a real AES but I didn't get the chance to test it on the BVM. I'll report back if I have any issue.

[fernan1234]

Oh that would explain the Neo Geo core not working! I do have the core set to MVS mode, but sounds like even setting it to AES mode it would still be based on the MVS clock. Glad to know there's a modified version that would serve as a workaround.

If it works on your A20 I'm sure it works across the whole A-series, they all use the same circuitry.

Given that these problems are clock based I'm guessing that a SyncSlayer or SyncStrike solution won't help with this core and the arcade ones. What we'd need is a tolerant video processor that can passthrough 240p but at the same time normalize the timings. This would of course add a bit of lag, a worthwhile compromise in this case. Maybe something the OSSC Pro will be able to do.

[daty2k1]

It is definitely related to the base clock and thus the refresh rate. That's why all those arcade cores with their funky refresh rates are not working.

I've yet to find a solution for these. I had some success with a custom modeline on CPS1 where I was outputting in 480p but I was not really satisfied with the result, image quality wise.

Now I'm using a 20M4 for arcade cores and everything works perfectly.

[daty2k1]

Well, picture is skewed with the AES