rtdzign, the first one only outputs 3.3V from 1-3V source, you’ll need the second one, it outputs 5V from 1.5V to 4.5V
That one should work and is the cheapest one I know of; its also the one I sent with the spark to Nerrage for S4J. But I havent tested them together, so test the crap out of it Nerrage.
I’ve already been working on a suitable DC-DC converter, so I’ll post up when I have as solid, tested DC-DC to use.
You got it, Toodles! I planned extensive testing, for certain. 
Can you sell the DC-DC converter internationally (with the ToodlesFlash) :D?!?!
The price is acceptable, and I understand why it is where it is. The important thing is that the product is of high quality, readily available, and within a reasonable price range. From the sounds of it the Spark meets all three of those criteria, while the Sanwa Flash only meets one (them things are EXPENSIVE, and IMPOSSIBLE TO FIND!).
Perhaps this has already been answered, but the big thing I am curious about with the Spark is corner sensitivity. Corners just aren’t as easy to hit on JLFs as they are on the old iL Eurosticks, and I am curious about how the Spark holds up-- what would be especially promising is if there’s any way for users to adjust that sort of thing, since there are a lot of different preferences of players out there to account for.
Basically, I am curious if the Spark has any way to solve the tradeoff dilemma between square and octo gates. On the square gate, each activation zone is the same size, but you have to push farther to reach the corners. With octos, the distance in each direction is the same, but the activation zone on corners is smaller. I’m not really sure if there’s currently a way to have equal physical distance and activation zone distribution between all 8 directions on a JLF though, other than a bizarre 8-switch PCB or PCB that digitally senses directions, like the Spark.
Im sorry Timstuff, I dont have time to go into all of what you’re asking about, but there are MANY MANY posts in this thread by me describing how I arrived at where I am, and a bunch of it describes how it works in relation to corners and accuracy.
Got everything in, just waiting on the batteries to charge to make a short install video for you.
Like a good student I heeded Toodle’s advice and did my homework. For now, most of my questions have been answered. I look forward to seeing peoples’ hands-on impressions of this stick. 
Also, on a side note I like the name “spark.” It reminds me of Transformers.
Just got my Spark kits Toodles. I installed one of them and so far it is working great.
This bodes well for everyone.
Are they up for sale?
Did some testing from both a 5V and 3.3V source, and they’re both working excellently.
However, one weird problem I had was that while testing the initial wiring, there was an odd time when the MC Cthulhu was reporting buttons 8 and 9 held down via USB. Haven’t had that problem with a PS2 with FGWidget converter > Cthulhu, so it may have just been a hiccup, or possibly something with the conversion. I just unplugged and replugged in the MC via USB, and everything was fine. I’ll test more with different restrictors to see if it may cause any overlap when comparing the two sources.
Also, I must say, this is an amazing product, and you will certainly not be disappointed by it. Everything feels spot on, crisp, responsive, another excellent product from Toodles. It may have taken a long time, but it was well worth the wait and all the effort.
Thanks for testing. Will be ordering the converters now that it has been verified working.
Let me test one more time before you order, I’m just hooking up the PS2, since I’m using the FGWiget converter from a 5V source, it may be messing things up, I just realized. Hold tight.
rtdzign, if you’re there, the PS2 is sourcing less voltage than the converter set up I was using, and it isn’t working, results up soon
Let me know any details you find; honestly, Id expect the converter more likely to have more problems when used with a 5v source than a 3v one, but my measurements said it would put out 5v even with a 5v source.
Test problem, PS2 is going badly. Here’s what I’ve got:
I have absolutely no idea what’s pulling the MC so low on voltage in these tests.
Source: USB measured at 4.11V on MC Cthulhu, Voltage measured for joystick: 5.03V, Joystick OK
Source: USB via FGWidget Converter, but using PS2 cable measured at 2.86V, Voltage measured for joystick: 4.38V, Joystick OK, but seems to have a larger deadzone
Source: PS2 measured at 2.08V, Voltage measured on joystick: 3.44V, **Joystick not working, buttons not working, PCB likely not working. **
Just installed my second spark into my other stick, currently in my avatar pic. I have it working with a Paewang PCB, and so far it is working flawless. Currently in all my Spark kit conversions, I am using a LS-55 spring which gives the needed resistance to feel about to what the JLF was with the default spring and the default switches. Maybe with the LS-55 and Spark, the resistance is slightly higher like an LS-32. Because the resistance from the microswitches have been removed, keeping the default JLF spring will feel a lot looser and the dead zone will be more pronounced. The LS-55 spring seems to make that feel a bit tightened up.
In the second test, I was comparing the deadzone against the USB results. Again, it’s subjective, for sure, but it just feels a bit bigger.
Larger deadzone at the lower 4.38v makes sense, since there’d be less light to activate the direction. But either way, the incomming voltages of 2.86v and 2.08v are way low. Can you give me more detail about the stick and setup?