| Main > Lightguns |
| Gun talk - Sega Type-II IR Gun setup |
| << < (10/19) > >> |
| twistedsymphony:
"Calibration" is interesting. there are 4 adjustment screws on the Gun I/O board I believe they're adjustable caps, not pots. everything I've read has said to never EVER EVER touch these, they're set from the factory and shouldn't ever need to be changed. The position of these on the board leads me to believe that there are 2 of these per gun, probably an X and Y adjustment. each one of these sits next to a tiny 50Hz inductor. The sensor in the gun itself has 3 adjustment screws, again these look like adjustable caps, not pots. and again I've been told to never EVER EVER touch these, they're set from the factory and shouldn't ever need to be changed. The thing is this whole gun assembly is self-contained, it doesn't need to be attached to anything at all to function, the output pins will happily provide whatever X and Y voltage coordinates that match where the gun is pointing. That's what makes this system so beautiful for use with a PC. The Games that use this each have their own internal calibration, which really just adjusting how the game interprets to whatever output voltage the Gun I/O is providing. Sega was pretty brilliant in the design because it means that you could swap out the IR guns for a set of mounted guns with POTs for X and Y and the game wouldn't know the difference. technically you could even play an analog based game like Super Monkey ball or even a racing game and just tilt the gun around instead of using an analog stick... or pedals and steering, again because the game doesn't know the difference. The Gun I/O has a 12V input for power, and then it uses a Voltage converter to create 5V. I originally assumed that this was just for the analog output voltage since it works within a 0-5V range, it seems odd to me that it would use this on the LED matrix, even more strange that it would use it in addition to 12V. The Circuit design on the LED board should be easy to determine, as it's a single layer/single sided PCB with mostly obvious traces. |
| BadMouth:
You have them on a 4:3 monitor. Experimented any with a 16:9? I wouldn't expect it to work, but it would be cool if it did. I guess you could put them on an oversized 4:3 bezel in front of a 16:9 monitor and the extra Y axis would be adjusted for in the in-game calibration..... |
| Howard_Casto:
The only thing I can think of would be one of those chips (probably the flip-flop) acting like a capacitor. Caps have to "fill" before any voltage is passed along which creates a tiny delay. So voltage is passed to board 1, after a delay caused from energizing the chip, the led lights up allowing voltage to flow to board 2, and so on and so forth. That would make the lights blink in sequence. But that would only need one line and probably wouldn't even need a clock, so I don't know what the other 2 or 3 wires are for. |
| twistedsymphony:
--- Quote from: BadMouth on February 10, 2016, 11:50:29 am ---You have them on a 4:3 monitor. Experimented any with a 16:9? I wouldn't expect it to work, but it would be cool if it did. I guess you could put them on an oversized 4:3 bezel in front of a 16:9 monitor and the extra Y axis would be adjusted for in the in-game calibration..... --- End quote --- I have no doubt that it will work. This setup came out of a Net City upright using a 27" CRT angled back about 20 degrees, and it's designed to drop into the "Delux" cabinet with a 50" rear projection screen and the board set originally came out of a "dedicated upright" which has the sensors pointing up and reflecting off of a mirror. Ghost squad and newer games all ran on 42" or larger 16:9 LCDs with the same hardware. I think you're right that any inconsistencies between setups just gets sorted out in the in-game calibration. Heck i just eye-balled the placement of the sensors in my carboard bezel and was still able to calibrate it pretty much perfectly. We'll know for sure soon though; I'm installing this on my projector which is a 108" 16:9 screen. I just need to build the extended LED board harnesses which is a few hundred crimps that I'm not looking forward to :lol --- Quote from: Howard_Casto on February 10, 2016, 12:58:11 pm ---The only thing I can think of would be one of those chips (probably the flip-flop) acting like a capacitor. Caps have to "fill" before any voltage is passed along which creates a tiny delay. So voltage is passed to board 1, after a delay caused from energizing the chip, the led lights up allowing voltage to flow to board 2, and so on and so forth. That would make the lights blink in sequence. But that would only need one line and probably wouldn't even need a clock, so I don't know what the other 2 or 3 wires are for. --- End quote --- I think you're over thinking this, I highly doubt there's any analog trickery going on with the LED boards, that just seems like it would just way over complicate things and open it up to falling out of calibration way too easily. This is just complete speculation on my part but if I was designing this board this is what I would do: the flip flops are used for latching, think of it like 1 bit of memory per flip flop. If we assume that the mystery chip is maybe an "AND" gate then we could use A for a pulse signal and B to be the signal for the previous board in the chain. When A+B are both positive then then toggle the flip flop and turn the LEDs on, This means that it will turn on the LED AFTER the one that is currently list. meanwhile all the other LEDs in the chain (including the previously lit LED) will have positive for the pulse on A but a negative for the value of B and thus toggle their LEDs off. This would in effect turn on 1 LED at a time in sequence. the C wire could just be a feedback to let the gun I/O know when the last LED in the chain has been deactivated. This could be accomplished by simply putting a pull-up resistor across B and C such that when NO LEDs are on C is low but if ANY LEDs are on then C is high.. that way the Gun I/O knows it's done one rotation and to start the sequence again.... this would also be helpful for starting the sequence since none will be on to start, and work as a check since it knows it should count to 10 before getting the "all-clear" and if it doesn't there's a problem. This would have to happen very fast. If we assume it wants to check gun position for every frame (and the game runs at 60FPS) then at minimum it would have to make a full rotation in under 16ms. again, this is just my personal speculation. |
| burn_654:
What kind of minimum distance does the sega ir setup allow? I experimented with the aimtrak and while I'll concede it does work very well, it doesn't work for my setup with the minimum distance requirements - I couldn't get satisfactory accuracy closer than about 2.5 - 3 feet which was just too far away for an approx. 19in screen. So I'm back at working with some crt guns. Personally, I'm working on a tool with my brother to increase functionality of the Guncon 2's on pc. The aim being proper screen flashing for the PS2 Guncon 2's using the EMS topgun driver... why not just use the Wingun driver, you might ask? Well, Wingun works great (and I'd actually recommend trying it!) with XP 32 bit but it can't support anything further. Same with the Guncon2PC driver. The topgun has a driver for x86 and x64 that is *also* compatible with 3rd party Guncons and works under Windows 7...but lacks proper screen flashing for darker games. Using autohotkey I've come up with a way to at least generate a directinput mouse click that is very slightly delayed to be during a flash (which is triggered off of a raw input). That, combined with a tailored control profile in the ems panel is starting to have noteworthy results. I'll definitely make a thread when that gets further along. This method won't work for the model 2 emulator (or raw input needs) ...but I may try a physical delay circuit for the trigger to send both an instantaneous key for our flasher program and a separate, slightly delayed key as the shot button. That would be the best of both worlds as it would still generate raw input for model 2/troubleshooter2 and a flash. |
| Navigation |
| Message Index |
| Next page |
| Previous page |