I’ve been thinking about blade connectors, and how they either have to be circular or polarized, right? Well, maybe they don’t have to be either… The + and - could be interchangeable if the power was rectified. Normally, a rectifier is made with four diodes, but that wouldn’t work well in this case, because diodes have a voltage drop, so we would lose a volt or more to the rectifier, which would be bad when we start with only 3.7 volts.
Anyways, the solution is to use FETs for rectification instead. Two N-fets and two P-fets can be used to rectify power with almost zero voltage drop.
Now, the real question is; does this actually help? I’m not sure that having the ability to swap + and - around helps any since we still need a data pin too.
1 Like
From my little experience, blade connectors problems are vibration, torsional forces when inserting the blades and total amperage of pins. I fail to see how rectifying would help with either of these. May be you have a mechanical design that would be enabled by not requiring radial symmetry?
The only issue I can see is it undermines what seems to have become somewhat of an industry standard of having three ring blade connectors wired a certain way.
That said, if we’re going down that road, I would love to see four or even five ring neopixel blade connectors. This would give some useful options. For instance with a four ring, you could have blade plug charging. A five ring would also mean you could wire it so the hilt only boots with a suitable blade or blade plug fitted. This is not as crazy as it sounds as it would mean you could lose the charge port and kill switch but keep a pre-wired battery. By removing some of the hardware needed in the chassis, this would really open up chassis design possibilities.
From the wireless charging thread, you can charge from the blade plug right now. There’s no need for anything else. And with BladeID pin and Proffies V3, you could wire it for turning off when nothing is installed.
1 Like
Thanks Baldusi. I just read through that thread. I had no idea the fets could cope with power going through them the wrong way to make this work. Might have to start building this feature into my hilts. Thanks for the post. 
It’s a neat idea; I think not too useful, since this would give you non-polar (yay), but still not circular (boo). So you still have to insert the blade 1 of exactly 2 orientations.
What would be more useful would be to have a way to connect to a blade with only 2 pins. I’m thinking positive, negative, and then… hide the signal in those two? Maybe have a differential signal that wouldn’t hurt the LED power so much (just a bit of noise), but could be differentiated out into a normal LED signal? Then you could use any circular 2 pin connector (heck, even 2.5mm power connectors, or those 10A high-current pogo pins).
My current DIY connector I’m working on uses GX16 plugs with a 90 degree rotation to lock them in (no lock screw needed). It works, not too bad, but adds some complexity. If I could use a 2 pin connector, that would be amazing.
Since neopixel data signals go at 800kHz, it should be fairly simple to just composite and then filter the signals from each other, at least in theory… Not sure how well it would work in practice though. (Noise might be a problem.)