SNK-NEO-GEO
Enlightened
HOLLY CAW...that looks awesome 
Shooting Gallery: Sega Type-II IR on a 108" front projection setup.
- Thread starter twistedsymphony
- Start date
The pedals I have are in pretty rough shape the sheet metal that makes up the housing is cracked in a few spots from abuse and one of them is missing the spring/spring plate. the mechanical design of these things is honestly pretty terrible I'm thinking I might try to weld the sheet metal back together and the remove the original switch and replace with an optical switch... seems like it'd be more reliable.
As for the guns, I currently have the HOTD2 guns, Virtua Cop 3 guns, Time Crisis Guns, and the GSS guns. I'm thinking I might re-pin all of them to use the same connector as GSS since that's the largest/most complex, then set them all up to use the same mounting plate. with 2 threaded holes on the plate. That way to remove a gun would be 2 bolts and 1 connector on the inside, and that's it.
I'd need something more dynamic for the holster too. The one I have now works for the HOTD2 guns but is a bit big for the VC3 and TC guns, and is too narrow for the GSS guns and definitely wont work if I ever get the Ghost Squad or HOTD4 guns.
I hear, I have a few of the rifles.
You are gonna have to get creative with gun/holster placement.
I like the idea of repinning the connectors. Maybe mount all the guns and just swap out the connections inside the cabinet/pedestal?
You are gonna have to get creative with gun/holster placement.
I like the idea of repinning the connectors. Maybe mount all the guns and just swap out the connections inside the cabinet/pedestal?
nice job
I finished designing and printing my gun holsters. The first test print I made I only printed the first 20 or so layers this let me print it in under 2 hours and make sure that it fit. it did need a few adjustments. The second one I printed in full took 12 hours but fit perfect thanks to the adjustments I made. then I just mirrored the model and printed the other side.
aside from the lock all of the woodworking was done so I took some wood putty and filled in over all the screws, once dry I sanded and then wiped everything down with a damp rag and did a second round of fill since some of the screw locations were still apparent. for this you need to see with your hands. It's not done until you can run your hand over it and not know where the screws are.
-----------
One other aspect of this project is I couldn't figure out how to hang the LED frame. the problem is I have a motorized screen and which has a very large housing that the screen rolls up into hangs wider and in front of the screen meaning I have no good way to hang the LED frame from the top. As it is the frame is already 8ft wide. I eventually had the idea to mount extensions on the top corners to extend the width of the screen another 1ft or so on either side; enough to make it wider than the the screen housing.
A while ago I had designed and printed corner brackets for the screen, they had been test-fit but I needed to drill holes in the frame pieces to finalize it. These fit very snug and getting them on or off requires the use of a mallet, but I wanted the screws on there to keep them from working themselves apart over time.
so after measuring I drilled holes using my drill press...
since the pieces were so long I used a step stool to help hold up the other end.
Also since these were difficult to slide into the end pieces I used a file to chamfer the ends
the corner pieces are printed in ABS with 100% infill, PLA would not be strong enough for this. ABS also has a nice dull/matte finish which works well in this application. I've been super happy with these brackets.
i am working on a sega ur shoting cabinet, and i have try to find some nice gun holster.I saw yours, i like it.
Is it maybe possible to maybe get the file so i can print it?
sure. I'll post the file up later tonight.
I'm not sure how useful it will be for your application since it's designed to hook over the top of some 3/4in MDF but maybe that's what you're looking for?
It's going to be perfect for what i am planing to do
here you go: https://www.thingiverse.com/thing:3133579
No payment necessary but if you are so inclined you can leave a "tip" through the thingiverse page
I've been working on a home arcade gun cab set-up for a while. It started with me buying some Aimtrak IR guns to use with mame and quickly finding out that they didn't deliver the arcade gun experience they promised...
The real arcade IR guns work a lot better but the Aimtrak modules lose all line of sight accuracy as soon as you move your arm. They are kinda ok for directional games like T2 and Operation Thunderbolt but they suck for light gun games.
Eventually I decided that, for directional gun games, it made more sense to use a real arcade fixed wired gun over the unreliable Aimtrak modules. I got an Operation Thunderbolt gun and wired the X and Y pots and the buttons to an Apac. It was a 20 minute job and works well on all my screens (including the 110" black projector screen in my living room.
For light gun games I just bought some PS2 Guncon 2 guns. I haven't downloaded the Windows 7 drivers yet but I read that they work (mostly) and, the Guncon boards can be wired to work with real namco arcade light guns (with a crt monitor obviously). We'll see.
For recoil, my OT gun came with a recoil motor already installed. For the light guns, I bought some cheap 12v solenoids that will do the job.
I'm interested to hear how accurate your IR guns are for games with no crosshairs on screen?
The real arcade IR guns work a lot better but the Aimtrak modules lose all line of sight accuracy as soon as you move your arm. They are kinda ok for directional games like T2 and Operation Thunderbolt but they suck for light gun games.
Eventually I decided that, for directional gun games, it made more sense to use a real arcade fixed wired gun over the unreliable Aimtrak modules. I got an Operation Thunderbolt gun and wired the X and Y pots and the buttons to an Apac. It was a 20 minute job and works well on all my screens (including the 110" black projector screen in my living room.
For light gun games I just bought some PS2 Guncon 2 guns. I haven't downloaded the Windows 7 drivers yet but I read that they work (mostly) and, the Guncon boards can be wired to work with real namco arcade light guns (with a crt monitor obviously). We'll see.
For recoil, my OT gun came with a recoil motor already installed. For the light guns, I bought some cheap 12v solenoids that will do the job.
I'm interested to hear how accurate your IR guns are for games with no crosshairs on screen?
So I recently jumped back into this project to see if I could get it going for this weekend.
Tidied up the wiring, swapped out the NAOMI PSU for a Lindbergh PSU and built a Lindbergh to JVS power harness. this will give me more versatility in the PCBs I can run. I also installed a nice power switch/fuse/filter/and socket: https://www.amazon.com/Suppressor-S...rds=fused+power+switch&qid=1548166054&sr=8-22
It's pretty awesome as it's got everything you need for AC power input in one piece. there's lots of fused and switched outlets but this is the only one I found that also included a filter.
I also designed and 3D printed an IO panel for audio video and LED matrix output:
So The issues I was having with the Gun tracking is really just the guns having trouble seeing the IR LEDs... if you're too far from the screen then they aren't bright enough and you get a "Screen out" even if the gun is pointed right at the screen. If you're too close then the distance between the LEDs is too much that they're no longer in the gun's field of view and you get a "Screen out". As it was about 5ft from the screen was a "Sweet spot where the guns would track, but even just pointing it towards the edges of the screen put it out of range of the other edges and made the whole thing unplayable.
I tried both the Vertical Jurassic Park Style led modules and the horizontal HOTD2 style led modules and there was no real difference. I contacted Ken @ irepairsega to see if there was as way to increase the sensitivity in the gun sensors and if he knew the LED specs so I could maybe swap to some brighter ones. This led me to some interesting discoveries.
1. There is no way to increase the sensitivity, the only way to increase range is to increase the output of the LEDs.
2. Ken worked with another person who built a projector based setup exactly like what I'm doing and they used the 4-LED modules found in some of the Deluxe cabinets and had good tracking up to 12ft.
3. the LEDs in these units are specially designed for the company that made the tracking equipment (OMZ). they operate at a non-standard wave length so they wont work with "normal" IR LEDs. Apparently Keven even found the LED manufacturer at one point and ordered every catalog led they made in the same spectrum range but couldn't find a working replacement. So the only option is to use the original LEDs.
So My best option is to find a set of those 4-LED modules PCBs.
During my testing I tested at 10ft and had decent tracking but I also had the LEDs clustered around a little 19in LCD and none of them had the tinted plastic windows. I removed the windows and in doing that the effective range for the guns increased to about 8ft and at that distance it's totally playable! still not the 12ft-15ft range I want but it's good enough to finally enjoy this setup.
one other thing I'm going to do is see if I can design and 3D print some angled mounts for the IR LEDs so the LEDs at the ends are angled in towards the player a bit, I think this will help with tracking so the LEDs are pointed more towards the guns (I'm thinking this is why my initial test range was better).
And in addition to searching for the 4-LED modules I'm also looking to see if I can find someone with a Light Spectometer that can measure the exact wavelength that these LEDs operate in. In doing that I might have better luck finding an IR LED that will work with the guns. Alternatively I'm looking for a wide-spectrum IR LED that might produce light across a large enough range that it hits the spectrum that these sensors are looking for.
Tidied up the wiring, swapped out the NAOMI PSU for a Lindbergh PSU and built a Lindbergh to JVS power harness. this will give me more versatility in the PCBs I can run. I also installed a nice power switch/fuse/filter/and socket: https://www.amazon.com/Suppressor-S...rds=fused+power+switch&qid=1548166054&sr=8-22
It's pretty awesome as it's got everything you need for AC power input in one piece. there's lots of fused and switched outlets but this is the only one I found that also included a filter.
I also designed and 3D printed an IO panel for audio video and LED matrix output:
So The issues I was having with the Gun tracking is really just the guns having trouble seeing the IR LEDs... if you're too far from the screen then they aren't bright enough and you get a "Screen out" even if the gun is pointed right at the screen. If you're too close then the distance between the LEDs is too much that they're no longer in the gun's field of view and you get a "Screen out". As it was about 5ft from the screen was a "Sweet spot where the guns would track, but even just pointing it towards the edges of the screen put it out of range of the other edges and made the whole thing unplayable.
I tried both the Vertical Jurassic Park Style led modules and the horizontal HOTD2 style led modules and there was no real difference. I contacted Ken @ irepairsega to see if there was as way to increase the sensitivity in the gun sensors and if he knew the LED specs so I could maybe swap to some brighter ones. This led me to some interesting discoveries.
1. There is no way to increase the sensitivity, the only way to increase range is to increase the output of the LEDs.
2. Ken worked with another person who built a projector based setup exactly like what I'm doing and they used the 4-LED modules found in some of the Deluxe cabinets and had good tracking up to 12ft.
3. the LEDs in these units are specially designed for the company that made the tracking equipment (OMZ). they operate at a non-standard wave length so they wont work with "normal" IR LEDs. Apparently Keven even found the LED manufacturer at one point and ordered every catalog led they made in the same spectrum range but couldn't find a working replacement. So the only option is to use the original LEDs.
So My best option is to find a set of those 4-LED modules PCBs.
During my testing I tested at 10ft and had decent tracking but I also had the LEDs clustered around a little 19in LCD and none of them had the tinted plastic windows. I removed the windows and in doing that the effective range for the guns increased to about 8ft and at that distance it's totally playable! still not the 12ft-15ft range I want but it's good enough to finally enjoy this setup.
one other thing I'm going to do is see if I can design and 3D print some angled mounts for the IR LEDs so the LEDs at the ends are angled in towards the player a bit, I think this will help with tracking so the LEDs are pointed more towards the guns (I'm thinking this is why my initial test range was better).
And in addition to searching for the 4-LED modules I'm also looking to see if I can find someone with a Light Spectometer that can measure the exact wavelength that these LEDs operate in. In doing that I might have better luck finding an IR LED that will work with the guns. Alternatively I'm looking for a wide-spectrum IR LED that might produce light across a large enough range that it hits the spectrum that these sensors are looking for.
Last edited:
Initial Research into IR LEDs and wave lengths is that most IR LEDs operate at either 850nm or 940nm, with the 940 being the more common of the two.
also most LEDs work on a very narrow band:
you can see that the intensity drops off HARD as soon as you start reading outside the specified spectrum.
so if you're looking in the 825nm wavelength on a 850nm LED it's only going to appear half as bright as it actually is.
Interestingly we can see the light output LEDs used in these OMZ light modules. And the human visible range ends around 700nm or 750nm depending on who you ask. The light is faint but I can still make it out across the room in low ambient light. To me that means the tail end is somewhere in the visible spectrum probably peaking just below 800nm.
I'd like to measure one of these LEDs to be sure, but at least that gives me a clue when researching wide-spectrum LEDs.
also most LEDs work on a very narrow band:
you can see that the intensity drops off HARD as soon as you start reading outside the specified spectrum.
so if you're looking in the 825nm wavelength on a 850nm LED it's only going to appear half as bright as it actually is.
Interestingly we can see the light output LEDs used in these OMZ light modules. And the human visible range ends around 700nm or 750nm depending on who you ask. The light is faint but I can still make it out across the room in low ambient light. To me that means the tail end is somewhere in the visible spectrum probably peaking just below 800nm.
I'd like to measure one of these LEDs to be sure, but at least that gives me a clue when researching wide-spectrum LEDs.
it looks like there are a whole bunch of DIY options for building your own spectrometers.
Most of them wont work above 700nm but this one seems to capture in that range:
http://physicsopenlab.org/2015/11/26/webcam-diffraction-grating-spectrometer/
or if I'm lazy this guy makes them: http://chriswesley.org/spectrometer.htm?ref=DIYLightSpectrometerDescription5
Most of them wont work above 700nm but this one seems to capture in that range:
http://physicsopenlab.org/2015/11/26/webcam-diffraction-grating-spectrometer/
or if I'm lazy this guy makes them: http://chriswesley.org/spectrometer.htm?ref=DIYLightSpectrometerDescription5
Well, I've got an IR spectrometer ordered and on it's way to me... lets figure out these LEDs 
you'll want them evenly spaced but it doesn't have to be super precise.
basically put the corner LEDs in the corners (they should be beyond the edge of the screen)
put the middle LEDs in the middle,
then the remaining ones just go half way between the corners and the middle.
my initial tests I just taped them to a piece of cardboard and eye-balled the spacing and it worked fine.
on the real cabinets the corners are rotated 45deg and this is mostly sothe angle of the LED points more towards the center of the screen. they'll work without doing this but, I recommend it if you can as it DOES increase the gun's visibility of the LEDs.
it's hard to find a good picture of this, but you can see how they're arranged here:
Thank you perfect!
Now i need to buy some crappy donor NUC or any other JVS cabinet.
The Light Spectrometer I purchased came in today. I used the awesome free Theremino Spectrometer software: https://www.theremino.com/en/downloads/automation
after pointing it at every light source in my office to see what it spit out, I hooked up one of the IR LED modules to a power source and got this:
the peak seems to fluctuate between 955nm and 958nm but it's got a fairly wide spectrum with the tail reaching as low as 850nm and as high as 1050nm (which seems to confirm the 950ish peak).
So now to find some super bright LEDs in this spectrum
after pointing it at every light source in my office to see what it spit out, I hooked up one of the IR LED modules to a power source and got this:
the peak seems to fluctuate between 955nm and 958nm but it's got a fairly wide spectrum with the tail reaching as low as 850nm and as high as 1050nm (which seems to confirm the 950ish peak).
So now to find some super bright LEDs in this spectrum
Coyo5050
Professional
Commenting so i see updates. Have a similar project to work on.
So the plot thickens...
I found a bag of "940nm" IR LEDs that I bought a while back and decided to test those.
putting them through the spectrometer they actually peaked at 1030nm. These were no-name "super bright" LEDs. interestingly they had zero visibile light. And while they did hit the 940nm range they were only at about half intensity for that spectrum.
so either these LEDs simply have horrid tolerances or were mis-labeled, or the spectrometer is quite a bit off.
What I've ended up doing is simply picking up a sampling of LEDs from Digikey
I ordered 850, 890, 940 and 950 wave lengths all from name brand manufacturers with data sheets available. So I'll see how these measure on the spectrometer and even if the spectrometer is off I should be able to find the LED that gives me a similar output to the originals.
FWIW the guy who built my spectrometer originally designed and built these to test LEDs and he did so because he found a lot of Chinese manufacturers mis-representing the wavelength and wanted a consistent way to verify the color output... so here's hoping those IR LEDs I tested were just junk and the spectrometer is actually right on
I found a bag of "940nm" IR LEDs that I bought a while back and decided to test those.
putting them through the spectrometer they actually peaked at 1030nm. These were no-name "super bright" LEDs. interestingly they had zero visibile light. And while they did hit the 940nm range they were only at about half intensity for that spectrum.
so either these LEDs simply have horrid tolerances or were mis-labeled, or the spectrometer is quite a bit off.
What I've ended up doing is simply picking up a sampling of LEDs from Digikey
I ordered 850, 890, 940 and 950 wave lengths all from name brand manufacturers with data sheets available. So I'll see how these measure on the spectrometer and even if the spectrometer is off I should be able to find the LED that gives me a similar output to the originals.
FWIW the guy who built my spectrometer originally designed and built these to test LEDs and he did so because he found a lot of Chinese manufacturers mis-representing the wavelength and wanted a consistent way to verify the color output... so here's hoping those IR LEDs I tested were just junk and the spectrometer is actually right on