Yes I potted to save wires from high centerfire recoils. I use AK-47 C (high, mid, low mode). This is the only one of 50 that failed by my mistake. I just wanted to report that to you guys to be more careful than I in building or modding :)
Bats are Pana NCR18650PF (there is no protected version of this cell on the market). I put this cells on discharge test year ago and now this happened and it survived :) . I don't know how but obviously it has some kind of internal protection even without added pcb's.
I will have to scrape fujik of driver to see what happened.
I managed to redo fujik treatment. Everything was in place although artic alumina under FET melted during fan de soldering. It seems that FET died since it looks and smells smoked but I did not want to bother replacing with another one. Curious thing is that driver still works on one weak mode.
With this test I found that FUJIK can withstand heat better than artic alumina when we do plastic to pcb gluing like I did with FET before leg and wire soldering.
For metal to metal I don't even doubt AA is far better than Fujik.
So if curing time is not issue I suggest that you put fujik under small FET. Than wait for 24hours for firm fujik cure in warm room, and after that you can finish driver leg soldering and adding +- wires.
You made me just recall something. I once modded a Nanjg 105C similar to what you did above and it only had High mode too. Turned that the FET Drain Pin (Big Tab across the back of the FET) was touching the inside of the pill and thus taking ground straight to the LED - Lead.
EDIT: I must be having a brain fart. Is the Gate soldered to Ground and the Source to the PWM feed? That can't be right. The OP shows the same setup though. I thought these N-Channels in this package were bottom-left pin = Gate and bottom-right pin = Source (when looking from the top). I must have something mixed up.
I don't know how to properly express myself but everything is connected as Djozz said. Like this:
I just scrape solder mask of those 4 vias and fill vias with solder for even better + contact to spring side.
My situation(one mode only) happened after shorting +- upper emitter wires with reflector ( i had only one layer of kapton tape on the bottom of reflector as protection and while I was pressing reflector +- wires somehow managed to dig hole into kapton tape, and then smoke out happened :) )
My drivers are potted so no way anything contacts with pill.
I haven't had any driver failure so far and they are in use for more than 3 months by various people.
It is dirt cheap and it works. It works even when smoked :)
I would not say this driver is winner over my diy Djozz driver guys.
My driver has High - mid - low without memory effect so it always starts on high. And I would not change that for new banggood driver unless they made similar modes.
Yes BG driver looks good but modes does not sound good to me.
Edit: - diy driver has reverse polarity protection
- It has a kind of low battery warning protection (light drops on about 1-2% at 2,8V of battery)
- 4.3A current draw of single ncr18650PF and all that stays in nice regulation for about 10 minutes on high and than another 10 minutes of 90-80% power, and then another 10-15 minutes around 80-50% output and then after some time it steps to low light warning protection which works for about 1 hour until it drops to 2,5V battery voltage.
- It is safe from start to complete battery discharge. So for all driver creators please try to make simple fet driver without memory, and user interface and with High - mid - low mode. Why? Try to sell flashlight to hunters and military guys and they'll say that they want instant full power action without any user interface.
Driver makers please try to make this mentioned version of driver and I guarantee that it will be your bestseller. You have nothing to loose.
I'm not very knowledgable in FET's, I just use what others use. I do hear that one FET is not the other, and that the specs highly matter. So if you post which exact FET you use, I hope that someone who does know this stuff will chime in.
I pulled it off my zy-t08 driver and was too eager to try it to look it up. I’ll do some poking on the t08 driver to see where each leg went.
edit
turns out to be the opposite of the FET used in the OP. Either needs to be rotated 180 degrees or mounted right-side-up. I chose right side up and isolated the back/bottom pad of the fet with kapton tape.
Always a good idea to search the code on an unknown FET. Not all or even close to all of them are appropriate to this usage. Look for the PDF data sheets with pinouts, drain source resistance, and gate voltage. If you don’t understand them then post a link to the PDF so it can be checked.
i have been thinking about trying this mod as well, but my programming-efforts so far have been limited to buying the hardware.
I was thinking about the mosfets mtn have, would one of them be good for this mod? and if, which one? AMC7135
That looks just like one I have. Guys I don't know even what FET is :) but I know how to build this driver(thanks to Djozz) so I really don't know what would be better.
Or Ebay one: http://www.ebay.com/itm/2pcs-New-NXP-4O30AL-403OOAL-4O3OAL-PH-4030AL-PH4030AL-SOT669/401027338093?_trksid=p2047675.c100011.m1850&_trkparms=aid%3D222007%26algo%3DSIC.MBE%26ao%3D1%26asc%3D34492%26meid%3D765e20dd42dd4e97bfe8b04e06bc3636%26pid%3D100011%26rk%3D1%26rkt%3D9%26sd%3D400900238110
I mean if I need to solder differently than mentioned on this thread than it can be problem?
and TK’s A6 firmware.
