Just got mine in today. Everything works perfectly on New Net City. Got the kick harness wired up ready for some cps2/3 games 
https://imgur.com/a/QKubPTI
https://imgur.com/a/QKubPTI
Name for JVS2JAMMA board
- Thread starter invzim
- Start date
sammargh
Professional
I'm going to guess the answer is "nope" but doesn't hurt to ask. Is it possible to decrease the time it takes for jvs to sync? I've run into an issue with Neo Geo and the Unibios 4.0 Pick 'n Mix menu replacement feature - http://unibios.free.fr/picknmix.html
To engage it you have to hold start at boot up however the jvs sync takes too long and the bios does not register the button press. I can force a crash on a mv-1c which can somewhat get me the menu but in a very broken state.
To engage it you have to hold start at boot up however the jvs sync takes too long and the bios does not register the button press. I can force a crash on a mv-1c which can somewhat get me the menu but in a very broken state.
invzim
Champion
It can be reduced, but as always there’s a trade off. JVS spec says to give io boards 1 second to power up, which is probably too slow for the picnmix you mention.
sammargh
Professional
Is it a variable that I can modify on my end or does it require rebuilding the firmware to change the timeout?
invzim
Champion
No it's not - it's part of the 'core' timings. I haven't really used unibios - any other functions where you hold a button down on power-up that doesn't require a 161-in-1 (which I don't have..)?
In other news, got the vewlix IO today and there is indeed hope
Still doing a bit of research on it, but the fix to get above is really easy and something anyone can do. I'll post a bit more when I have the facts.
Woohoo! Let me know if you need me to test anything for youIn other news, got the vewlix IO today and there is indeed hope
Still doing a bit of research on it, but the fix to get above is really easy and something anyone can do. I'll post a bit more when I have the facts.
randomdoohickey
Student
Anybody thought about input lag benchmark testing various JVS IO boards? Seems pretty easy to get an automated precision timer test going for closing an input button circuit to seeing the output on the JAMMA connector.
On a related note, given the size of JVS packets, baud rates, and typical input configs, what's the theoretical max refresh rate that could be used? I assume most JVS IO boards have some kind of hard-coded max packet and/or polling rate like 200hz or something?
On a related note, given the size of JVS packets, baud rates, and typical input configs, what's the theoretical max refresh rate that could be used? I assume most JVS IO boards have some kind of hard-coded max packet and/or polling rate like 200hz or something?
sammargh
Professional
You hold A+B+C to get into unibios or B+C+D to get into MVS menu. A+B+C+D to get to memory manager. None of them work unless you soft reset from a game.
invzim
Champion
Mailed all rev 1.4 customers with info on fix, and made a short video demonstrating how to perform the mod:
If you have the full-debug firmware installed, please update to 2018-08-20t2 version as the full debug will impact performance.
The vewlix io has a couple of 100ohm resistors in series with the JVS datalines, I have no idea why they did this or what they were hoping to accomplish - and I've seen other RS485 design that does this.
Big thanks to @Derick2k for sending me a board to investigate.
If you have the full-debug firmware installed, please update to 2018-08-20t2 version as the full debug will impact performance.
The vewlix io has a couple of 100ohm resistors in series with the JVS datalines, I have no idea why they did this or what they were hoping to accomplish - and I've seen other RS485 design that does this.
Big thanks to @Derick2k for sending me a board to investigate.
@invzim Thanks for figuring it out. Does the fix affect the jammafier working with other boards after this fix?
I cant do the fix yet, but looking forward to it. I am redoing my home office so everything is packed away. Gonna be another 2 weeks before I start getting stuff back.
I wonder if those resistors on the Taito io are also the reason why it doesnt work with certain boards/systems??
I cant do the fix yet, but looking forward to it. I am redoing my home office so everything is packed away. Gonna be another 2 weeks before I start getting stuff back.
I wonder if those resistors on the Taito io are also the reason why it doesnt work with certain boards/systems??
invzim
Champion
I've tested the mod with other IO boards I have at hand and all looking good (sega, capcom, taito egret3) and in theory it should be good all around. The resistor removed is to make the signals a little 'prettier', but not all JVS systems use it so it shouldn't have a negative impact.
Great stuff! Thanks for testing quickly, one thing is on the bench - but you never really know until it's tested live.
wigsplitta
Professional
sammargh
Professional
I wonder if someone can recreate the Sega Tower of Power....
So far this is my longest chain:
button input > JVS > JAMMAFIER > FastIO > X3 > Niko's JVS translation
Thanks to Niko's fabulous work this probably actually runs faster than a legitimate TGM3 setup since it patches the internal JVS routine to end quicker in game.
So far this is my longest chain:
button input > JVS > JAMMAFIER > FastIO > X3 > Niko's JVS translation
Thanks to Niko's fabulous work this probably actually runs faster than a legitimate TGM3 setup since it patches the internal JVS routine to end quicker in game.
invzim
Champion
I think most issues are resolved now - still waiting for little feedback from @Blackfish on test/service and audio on the Noir cabinets. Unless something major comes up I'll do another small batch (max 15) soon.
Some upcoming firmware features:
1 - Input/button test on OLED
2 - MVS remap, i.e. button 4,1,2,3 = 1,2,3,4
3 - Option to have button 4&5 on just kick connector
I'll have to implement some kind of ini file system to make feature 2&3 persistent across reboots, as it would be a PITA to set these each time the cabinet is power cycled.
Some upcoming firmware features:
1 - Input/button test on OLED
2 - MVS remap, i.e. button 4,1,2,3 = 1,2,3,4
3 - Option to have button 4&5 on just kick connector
I'll have to implement some kind of ini file system to make feature 2&3 persistent across reboots, as it would be a PITA to set these each time the cabinet is power cycled.
Blackfish
Professional
Yeah, Sorry. I have been tied up with work and haven't had anytime to do anymore testing. I will hopefully get some time this evening and then I'll report back with my findings.
randomdoohickey
Student
Was that on the older K91X1097C Taito JVS IO boards or the new K9101262B variant that jassin000 discovered?
More info:
Taito Vewlix JVS IO
Older one: "K91X1097B"
And it looks like all these IOs are essentially the same. I would love to see if the same resistors on the JVS data lines are being used in the latest iteration that @jassin000 has/had.
man you made my day with the mvs remap