Most flashlights with an electronic/momentary switch have that switch at the side, where the driver and the LED sit. However there are a few flashlights with such a switch at the tailcap, like the ThruNite Scorpion, or the NiteIze IQ-switch for minimags. But how does that work? The main problem for electronics in the tailcap is that it does not have direct access to the positive battery terminal, so how is it powered?
I've seen a light with a special 3*AAA battery carrier which got both terminals to the tailcap, so that's easy to understand, but that's not the case in many other of these lights.
The IQswitch may draw some milliamps through the bulb when off, but when on it is supposed to have a low resistance, i.e. to drop almost no voltage, so how does it get powered then? Also it does work with LED lights where pulling some milliamps through the driver (when off) might be difficult.
If it's a momentary at the tail cap, isn't it just like the side switch where it makes/breaks the negative circuit? Actually, I haven't had any drivers with momentary besides yours and the Nitecore ones, but wouldn't it just be a momentary on for the negative path?
The switch itself is just a momentary switch, but it can't simply connect the negative terminal, because the light can still be on after you release the switch, just like with my drivers.
There has to be more back there than just the switch, some sort of circuit that acts as a solid state relay. There's voltage there to run the tailcap electronics, measure between BAT- and GND and you see full battery voltage.
If the light has a driver up front, instead of all of it inside the tailcap, the front driver will work same as a driver for a conventional clicky switch driver, not like one specific for a momentary switch.
There are some electronic tail cap switches that have a small coin cell in them driving the gate of a high-current FET. A member here (Mattaus) is building one that even has a processor in it…
I modded a Sipik 58 to have an electronic switch in the tailcap. I used the metal outer casing, the switch boot, and the spring. Everything in between was replaced.
I used a small momentary pushbutton switch I purchased from radioshack to serve as the switch. I mounted it on a couple pieces of plastic sized for the inside of the tailcap and put the spring on the other side.
I used some solder braid as my negative contact to the outer casing. That bypassed the switch and was soldered directly the to the spring. A second wire was soldered from the solder braid to the the momentary pushbutton, and from there to a contact plate that sat around the top of the spring, but did not touch the outer casing of the light or the spring.
I filed a slot along the inside of the battery tube and ran a wire along it from the 4-star on my DrJones driver to a contact ring at the back of the body just below the screw threads. This contact ring was held in place with arctic alumina. When the tailcap is on, the contact plate touches the contact ring closing the connection and allowing the switch to function.
Other lights have much more elegant solutions. My Liteflux LF2XTs have essentially 3 tubes around the battery: There’s the outer casing which serves as one contact, then a middle plastic insulating tube, and then an inner metal tube to serve as the second electical contact.
You could probably use a much higher resistance pull-down resistor on the gate. 100K ohms will pull 30 microamps from the battery. 3 megohms would reduce the drain on the battery to 1 microamp and increase its useful life by 30 times… It would increase the turn-off time of the FET, but that should not be a problem.
I’m watching this with great interest, I was planning to do something like this on my Apex 5t6 but didn’t think it was possible without finding a way to connect to Battery +.
The button cell for power is a neat idea, thanks for the schematic!
I’m completely ignorant when it comes to specing mosfets so I need some help.
Would anyone be able to recommend me a FET that could handle around 16A (ideally up to around 20A) and would function in this kind of circuit?
Is that even feasible?
I’m interested in a very compact (flat) switch that would allow me to use an 18650 battery with a 18500 battery tube. Right now, the only way to have this setup is to have a twisty switch (Oveready). So in order to have a push button switch rather than a twisty, the push button switch needs to be super flat.
I know DrJones hasn't had a chance to reply yet but I can vouch for him and say he knows all about FET switches - he wrote the code for my Smart Switch and even helped me refine my hardware design to ensure minimum power consumption by the MCU :)
I think what the good Doctor is struggling with is a momentary style side switch that is in the tail instead of up near the neck/head of the light (like Zebralight etc). I was recently exploring this issue and didn't get very far. This one should be familiar with DrJones - my Smart Switch now uses PWM to dim and brighten the light (by controlling the current supply to the actual driver of the light) when the coin cell battery is getting too low to function safely. Doing so stops the driver from switching modes during this alert as the first firmware blinked once quickly which meant if you turned the light on high, but the coin cell was low, it forced a switch to the next mode. You could expand on this by simply programming the MCU to control the FET just like you would for a normal driver. The problem with this is instead of being an analog on or off (as per my switch), it's now a PWM signal and small coin cells won't last very long AT ALL doing this. To make this approach truly usable you'd need to route the main battery supply to the MCU. You could achieve this by switching to a P-Channel FET and reversing the battery orientation so the positive is now pointing towards the tail.
Battery+ > MCU, FET, & Switch > LED > Battery-
It is effectively a direct drive light but it wouldwork...to my simple mind anyway lol. You could add modes very easily to a whole host of older lights this way. It is something I'm looking at doing down the track although if someone else took it on I'd be happy let them do all the work haha.
I think actual constant current lights that use a tail mounted momentary switch are routing the signal from the switch (in the tail) to the driver/MCU (in the head) via it's own signal path, normally via a single thin wire embedded in the host itself. Only one wire is needed because normally the switch signal is recognised by the MCU as a sudden connection to GND on a particular pin. The possibilities are endless with this sort of switch.
[quote=cncyana]
[quote=cncyana]
I'm interested in a very compact (flat) switch that would allow me to use an 18650 battery with a 18500 battery tube. Right now, the only way to have this setup is to have a twisty switch (Oveready). So in order to have a push button switch rather than a twisty, the push button switch needs to be super flat.
[/quote]
I don’t understand what a Pfet would do better…? You always need power- and + to the tail, if you have electronics in there….always…
I also think that one thin wire as a connection for the momentary would probably work best, maybe a thin piece of metal like on the protected cells isolated with kapton tape would work.
If you think about protected cells the positive connection is only some mm away from the tail…
I’ve been planning a fet switch mod for my modded Apex 5t6 and if you have a soupcan 4x18650 style light it’s actually quite easy to get battery positive to the tailcap.
The contact board/driver in these types of light usually have a circular contact patch in the middle of the battery positive contact ring. This is either already connected to + or easy to connect if it isn’t, and then you just have to use a long brass standoff with a spring soldered to the top to bring this connection to the tailcap. There’s enough space in between the 4 cells to accommodate the standoff/post and it’s also easy to bolt it to the center of the spring board in the tailcap.
I’m using a 60mm brass standoff but you can also stack shorter ones to get the ideal height.
-
Fascinated by the non connection methods you guys are bringing to the table though. Single cell or long tube lights are going to be harder to connect up directly. Cool stuff
Yes, I was aware of the coin cell solution and those FET switches, should have mentioned that.
With the PWM and capacitor solution, I guess there must be some bypass for some low current in the head (resistor parallel to LED?) to get sufficient voltage towards the back (and without lighting up the LED when in stand-by).
I ordered a SmallSun ZY-A629 a few days ago and will check it.
... Ah, just searched the forum for it, and indeed, there's a resistor bypassing the LED :) -> link (and review)
So they use the resistor to create a high resistance path to the tail. This allows the circuitry in the tail to see the full voltage (rather than what's left after the voltage drops across the LED). Most of the current will flow through the LED however so everything still works.
It still requires what is basically a 100% custom light, rather than a simple tail swap. But who knows, that's what you might want :)
how do hds and novatac do it? the novatac has power running through the switch even without the spring inside. i put the battery in without the spring and the light gave me the battery detect but no access to mode switching.
