Unfortunately, the low-power killswitch circuit on the V3 board is not actually doing what it’s meant to do, which means that you still need an external FET for low-power switches. Sorry about that.
I remember seeing the wiring diagram for low-power switches with two options (one with 3 wires and one with 2 wires). Does that mean we’d need the 3-wire option to wire low-power switches, or is it now not possible at all with these boards?
Unfortunately, neither the 2-wire, nor the 3-wire option will work, the power will flow either way.
The low-power switch was a late addition, based on a low-power switch design I have in my K4. The idea was to use the reverse-polarity-protection pFET that’s already there to control the power. What I failed to take into account is that the reverse-polarity-protection pFET is hooked up “backwards” compared to a regular pFET setup. This means that if you turn off the pFET, the reverse-diode inside the FET will transmit power anyways, but with a 0.6 volt drop. If the pFET was hooked up the other way, then it wouldn’t work as a reverse-polarity-protection, because the diode would carry the reverse power…
So, unfortunately, it’s not possible to re-use the reverse-polarity-protection pFET this way.
I see. That’s a bummer, I was planning on making a chassis that would use a low-power switch. So I’ll have to go “old school” and use the standard high-amp kill switch. No problem 
If you can hide a pFET somewhere, you can still do it.
How small is that component? And how would I wire it? I haven’t used such a component before, and I tried Googling it. But I’m not sure exactly what I’m looking for 
My plan was to use a small low-power switch in a Luke Hero control box. Perhaps a pFET could be squeezed in there if I get really creative. Or design a PCB for my hilt, and learn another new skill.
pFETs exists in many different sizes.
The easiest would be to use a big one:
Another alternative would be a small one with a breakout board:
Or you could use my graflex helper board:
https://fredrik.hubbe.net/lightsaber/k4/control.html
1 Like
I designed a quick-n-dirty breakout board for the pFET:
https://oshpark.com/shared_projects/avY5rDhY
This thing gives you 4 breakout boards for $1.10
I recommend ordering it in the 0.8mm thickness, with 2oz copper.
Each breakout board is 6x6mm.
Room for a pFET on the top and a resistor on the bottom.
Wires or pins goes to battery+, Batt+ (on the proffieboard) and the switch.
The pFET would be the one in the previous post, or something like it.
Obviously it’s completely untested so far, and it may be too small to be practical… If someone needs a different design, please let me know.
I can procure a Aod403 (I’m attaching the datasheet). Would that work? It’s big enough to solder directly, but small enough not to be a problem sizewise. It’s a 30V P-Channel MOSFET.
AOD403 Support Documents.pdf (403.8 KB)
The AOD403 will work, but I a fet with a smaller gate-threshold voltage would be better.
The AOID403 lists the gate-threshold voltage MAX as -3.5 volts, which is right around the battery voltage. The DMG2301L-7 has a gate-threshold voltage max of -1.2 volts.
I see that the adafruit part I linked has a gate threshold voltage of up to -4 volts, so that one probably isn’t all that great either.
So the DMG2301L-7 would be best to use? I see that it’s rated for 3A @ 20V. But I assume it’d be fine with the li-ion batteries we use, since the voltage will be 3.7-4.2V (and can thus take a higher amp)?
The positive side of the proffieboard doesn’t need more than ~2A or so, so it’s sufficient.
(Because the pixels are fed directly from BATT+)
If you want to interrupt all power from the battery, you need something that can handle 15A+, but you can use an n-channel FET instead, which are easier to find.
Ah, that’s right. There aren’t any pixels in the circuitry that would handle these switches, I forgot about that.
It’s going to be a fun summer project learning how to design a PCB for this. Looking forward to it
@profezzorn I don’t get quite the difference between JFET and MOSFET and P vs N channel. Please correct me if I’m wrong, but a MOSFET that works in drain mode is equivalent to a JFET (i.e. only three contacts). And I don’t get why for a JFET you need a P channel one, but for a MOSFET an N-channel one.
What resistor should we use for the DMG2301L-7DICT-ND? It doesn’t seems too small to me. In fact, if it could be arranged in longer mode (so the three contacts and the pFET are all inline) it would be ideal. Since for sabers length is usually a lot easier to handle than width.
So my little board has space on the back for a 0603 100k resistor.
I’ll have a go at designing a thinner/longer board tonight.
Can’t have too many options, right?
2 Likes
What’s the use of the 100k? Is it needed for it to work?
Without a resistor of some sort, the pFET gate is floating, and can be in any random state; on, off, or anything in between. The 100k keeps it off when it’s supposed to be off.
The alternative is to wire the middle pin of the switch to the gate, then wire the other two pins of the switch to GND and BATT+, that way the gate will be always be pulled high or low. (Unless you leave the switch in the middle somehow.)
1 Like
Longer/thinner version of he pfet breakout:
https://oshpark.com/shared_projects/PZinyOen
9 for less than a dollar.
Does the switch pad go to the battery negative, or to the middle pin of the switch itself? I think it’s the latter, but wanted to be sure