Today, I implemented your schematic on a piece of veroboard, using 2 CNY74 optocouplers. This works wonders and it’s easier to do than the Toodles mod, though a bit pricier (but time is money).
Some pics:
One question: How do you work out the 1K resistance of the schematic ?
In the datasheet forward current is specified typical at 20mA. After reading the sheet is was sure that it is sufficient to use 4 - 5 mA to switch (low currents). Working current is 5V. Forward voltage of IR led is 1.2V.
Use resistors lower than 1Kohm for working voltage of lower than 5V (playstation).
i’m new to the whole arcade stick/controller modding thing, but am i right that the sole purpose of this is to keep the pcb/wiring harness on the stick, right? otherwise you could just remove the harness and wire the stick like a normal ls32 (using 2 wires for every direction directly onto the dpad) right? or are there any technical limitations which would keep this from working?!
OK, I was unsure how much current was needed for the PCB to register the input. Thanks.
PS: The PCB shown above is for sale (preferably europeans customers since the dollar is weak and shipping is going to cost an arm). I was looking for 80 EUR + shipping. There’s a lot of work involved, some parts, PCBs are expensive in Europe. One thing which is missing is the jack connector for plugging the Xbox headset (drill a hole in your casing or make the cables run through the USB cable hole).
yes, I’ve a soldiering service in a big german forum about building stick, too. I leave the connector and the wuick disconnects and try then reduce the price to 60? plus shipping. But 80? for this luxus version is definitiv okay.
Yeah, I know hardedge, they are a good supplier of SF4-related youtube videos
But I didn’t know there was a forum associated with it. Alas, I speak no german.
I’ve had a quick look at the hardware section, there are a few things there worth of interest, I think i’ll spend some time having a look at the diagrams
OK, I talked about some potential problems of the 16-input common ground mod privately with bencao (no problem with the 5 pin connector mod, this works just fine), but perhaps someone here will know better ?
The problem is this: Some inputs report a voltage of about 0.3 V across the 2 points, I figure this could cause a problem for the current to cross the emitting diode inside the optocoupler since Vf is around this value. I checked the net but @5 mA, 0.3V is about the lowest Vf I can find for diodes (maybe I haven’t looked hard enough). Maybe it’s time to find an alternative schematic for doing the same thing ?
I have just checked on a Joytech Neo Se PCB for XB360 I have and most input voltages are superior to 0.4 V except 2 directions (left/right), the Back button and maybe RB (very close to 0.3 V).
Does anyone know where I can find some optocoupler ICs? I’m guessing that don’t sell them at Radioshack. If anyone can point me in the right direction, I’d be much obliged.
From my understanding, can i just solder the left and right ground, and the same to the UP and DOWN even though the diagram doesn’t state that its a ground?
I have problems with some PCBs which would not register some inputs. I believe the optocoupler is not strong enough to pull down the high state to low. I tried (unsuccessfully) to use transistors to force the transmission of the low state, but I think I can just achieve the same by using a high CTR optocoupler (around 1000% CTR - see KB852 datasheet, it’s PC817 pin-to-pin compatible).
bencao, what do you think about this ?
Note: Some inputs were transmitted OK, it’s just for the last two (back/xboxguide/start). I checked every voltage, everything should work (I especially paid attention to polarity) but it doesn’t … hence the idea of a meatier opto
yes, we was discussing this symptom. I think now I’m back on the way. I’m assuming that we’re searching the value V_ce_sat. V_ce_sat depends on I_f and I_c according datasheet. When you’re using 1K ohm as led resistor I_f is about 5mA. Reduce the led resistor to gain more I_f. Perhaps this results in better switching behaviour.
KB852 has Darlington output. Perhaps using this one improves the behaviour, too.
I think I resolved the issue by “amplifying” the optocoupler output with a NPN transistor (mines had Hfe around 200-300)
Connect the opto COLLECTOR to the NPN COLLECTOR and Positive side of input (“button”)
Connect the opto EMITTER to the NPN BASE
Connect the NPN EMITTER to the Negative side of input
I think all the above is equivalent to changing the normal PC817 to a darlington optocoupler (like the KB852 - but this guy’s 2 EUR a piece !). I couldn’t bother to go to the electronics shop so I tried the transistor solution and I’m happy it works !
So far, all tests are conclusive.
This is my dual-mod between a non-common ground PCB (Joytech neo SE) and a not common ground either PS2/PS3/PC PCB (3-in-1 thrustmaster). To make matters worse, the thrustmaster PCB not only isn’t common ground but has high logic for the main buttons and a low logic for start/home/select !
I had to undo quite a bit of what was planned at the beginning but I think this all history now ! Some more testing, I’ll try to remove one of the USB cables and I’ll post my finished stick ! :lovin::lovin::lovin:
I’m currently doing a dual mod with a common ground xbox360 wireless controller and a sixaxis ps3 controller. So far I have a working circuit for a cg mod on the sixaxis that uses a PNP transistor and a resistor for each button. Ill post the schematic/pics after I finish.