Reverse Polarity Protection with the DNA20D

[mamu]

All I could find that might be doable are either dual channel or SMD P-FETs. ugh.

I emailed Vishay explaining my build and the voltage drop under load and requesting a recommendation. Am waiting to hear back.

 

[mamu]

Update: the Vishay rep recommended si4477dy or si4463cd for reverse polarity protection with the DNA20D. I ordered si4477dy and have been doing some testing with it.

I'm happy to report great results with reverse polarity protection plus reasonable voltage drop under load (except with the single 18650 batt under load at higher watts).

Single 18650 AW IMR 2000mAh batt:
@10W = .27v drop
@15W = .42v drop

Dual 18650 AW IMR 2000mAh batts in parallel:
@10W = .14v drop
@15W = .22v drop

The rep suggested I combine the 3 source pins (pins 1,2,3) and treat it as one pin for source and to combine the 4 drain pins (5,6,7,8) and treat it as one pin for drain.

It is an SMD part, but at a 5mm size it is doable.

si4477dy

 

[DrMA]

Wow, si4477dy looks like a very nice FET for this application.

The Rds(ON) graphs tells it all IMO.


Can you compare the voltage drops with the FET and without the FET in-line?

 

[dr g]

where are you measuring the voltage drop?

 

[mamu]

The pins on this part are fragile and break easily when soldering the wires directly to the pins, so I modified a 3x3 piece of veroboard and mounted the FET to it.

...Can you compare the voltage drops with the FET and without the FET in-line?

I did and voltage drop is essentially the same (the same or 0.01 - 0.02 volt difference) with the FET vs without. So this is definitely great news as it means the FET isn't affecting load and is only providing reverse polarity protection.

It also means the AW 18650 batt doesn't hold up well under load with the DNA20D running at higher watts. At least using a mod with the DNA20D and a single 18650 batt. A voltage drop >.4 @ 15W is significant. Dual 18650 batts in parallel, of course with the doubled max discharge, hold up well under load at higher watts.

I now have it in my Denali DNA20D mod that has dual 18650 batts and also my Sam mod with a single 26700 MOLI batt. Both are working great with this FET! I am so in my comfort zone now with having reverse polarity protection for these mods.

where are you measuring the voltage drop?

No load vs load input - at the battery end.

 

[txredxj]

dang it. looks like i have another part to order lol. this goes after the fuses?

 

[dr g]

Is that an AW 1600mah imr?

edit whoops I see it is a 2000mah. those dont seem to be quite as stout as the 1600mah under relatively high loads, but 15w is sort of moderate ... do you have other batteries that perform better under that load?

 

[mamu]

dang it. looks like i have another part to order lol. this goes after the fuses?

I wired the FET after the fuses in both my Denali and Sam mods - have tested both for reverse polarity and aok.

... do you have other batteries that perform better under that load?

only high drain lipo batts.

The only AW 18650 IMR 1600mAh batts I have are old and tested poorly. So need to order new ones.

I also tested the 26700 MOLI batt and it tested the same as the single AW IMR 18650 batt under load at higher watts and it has a 40A rating.

 

[bapgood]

Mamu do you have a few cycles on the 26700 moli?

I noticed a major increase in performance with the MNKE 26650s after they had cycled a few times.

 

[mamu]

I have maybe 6-7 charges on the MOLI. Also have 3-4 charges on the MNKE. The MNKE still cut out at 3.8v-3.9v when running at 15W while the MOLI keep on going til 3.5v before cutting out.

 

[dr g]

in practice does the battery drop matter?

 

[mamu]

in practice does the battery drop matter?

I think it does. With a batt voltage sag under load, I don't feel 0.1v or 0.2v is a big deal, but when you get voltage drops higher than that, the batt is being stressed (resulting in possibly a shorter life cycle or more serious issues) plus you're going to get shorter vape times with that batt with it cutting out at a higher unloaded voltage.

I just vaped a complete charge with my Denali that has dual 18650 batts. With the watts set at 15 the batts cut out at 3.67v. The voltage drop with that set up is 0.22v. If there were no drop under load, I could have vaped with those batts longer before they cut out. If the voltage drop were even higher, I would have had the batts cutting out at a much higher unloaded voltage.

 

[mamu]

I received my order of the AW 18650 IMR 1600mAh batts today and did some testing.

No significant voltage drop difference with the FET vs without the FET.

--------------------------------------------
Single 18650 AW IMR 1600mAh batt:
@10W = .15v drop
@15W = .22v drop

Compared to...
Single 18650 AW IMR 2000mAh batt:
@10W = .27v drop
@15W = .42v drop
--------------------------------------------
Dual 18650 AW IMR 1600mAh batts in parallel:
@10W = .11v drop
@15W = .16v drop

Compared to...
Dual 18650 AW IMR 2000mAh batts in parallel:
@10W = .14v drop
@15W = .22v drop
--------------------------------------------

Not a significant difference in voltage drop between the 1600mAh vs the 2000mAh when using 2 batts in parallel, but def a noticeable difference in voltage drop between the 1600mAh vs the 2000mAh when using one batt.

The 1600mAh batt claims a lower internal resistance and a higher C rating than the 2000mAh batt and based on the specs between the two and these results I feel is a better buy than the 2000mAh batt, especially when using in a DNA20D mod that takes one 18650 batt.

 

[dr g]

When you look at the discharge curve of the 1600mah it's pretty surprising. Holds voltage very well until it's dead. That's why it's hard to beat as a mech battery.

However IMO with a DNA 20d you can make use of a lot of the sub 3.5 voltage that longer-life batteries offer. You really need to look at the whole discharge curve.

It's not the .2v from 4.2 to 4.0 that matters in a DNA 20 application, it's the .2v from 3.4 to 3.2.

 

[bapgood]

Awesome news Mamu!!!!

dr g....I think you a mamu are basically saying the same thing.

I just got some of the fet's to try....now I need to get some of the 1600 18650's


Mamu do your numbers include PTC's in the circuit?

 

[mamu]

...Mamu do your numbers include PTC's in the circuit?

No. I should test that though I'm curious too. I did test with and without the fuse in my Denali DNA20D mod and the voltage cutoff was a bit lower without the fuse than with the fuse.

rtdvapor.com has good prices on AW batts.

 

[Littlefeather]

Want to say first off, big thanks to Mamu, Bapgood, Retird and all the others on this forum that are so willing to share information and help others.
I’ve decided to get my feet wet as it seem much more fun being in the game than on the sidelines so to speak. I haven’t made anything this small in probably 25 years. Back then it was scale models and miniatures. I am amazed at how short my arms have gotten in that time! Thank god for opti-visors and simple hand tools (non-motorized)! Ya don’t cut things as fast but then again, ya don’t cut things off as fast either.  Ask me how I know that? LOL!
Need some help with the wiring connection of the SMD reverse polarity chip! Just wanted to make sure these old eyes were going at it in the right way. Please see the attached photo and let me know if I'm heading in the right direction. As I said, it's been a lot of year since I worked on anything this small and I'd hate to bugger it up.
Thanks again for all the help.
Littlefeather

 

[mamu]

You're good to go with the wiring, Littlefeather.

Just be careful with the pins on the FET - they're fragile and break easy.

 

[Littlefeather]

Thanks Mamu

Excellent! Yeah, I'm going to mount the FET on a PC board.

 

[ch2468]

Correct me if I'm wrong, but I believe you had your diode installed wrong.

http://i.imgur.com/WLaqzLq.png

It's no wonder you were getting a voltage drop... The diode is meant to divert current away from the return path of the pcb input terminals.

edit: and if you're worried about the battery shorting under reverse input conditions, you can put a fuse between the positive terminal of the battery, and the diode. (Or a current-shunt resistor)