ZebraLight AA, Nimh 550 Lumen Boost Driver?

10 posts / 0 new
Last post
moderator007
moderator007's picture
Offline
Last seen: 2 months 4 weeks ago
Joined: 12/23/2012 - 04:47
Posts: 3739
Location: North Carolina
ZebraLight AA, Nimh 550 Lumen Boost Driver?

ZebraLights SC5w Mkll can do 550 lumens from a xp-l2 using a AA, Nimh cell. According to TA's test here with the xp-l2 that should be about 1.25 amps to the led which should be slightly above double that on the input side of the driver. I have searched many times for a boost IC from a single AA, Nimh cell but I have never found anything that even gets close to the output of 1.25 amps. Does anyone know of a IC that can do this or maybe even seen the circuit of the driver? Any info would be helpful!

Edited by: moderator007 on 05/20/2020 - 20:13
iamlucky13
Online
Last seen: 7 min 15 sec ago
Joined: 06/22/2018 - 09:18
Posts: 965
Location: USA

We discussed this topic a bit at the other forum when the mk II AA-powered lights were new. A member received an answer from Zebralight explaining that the boost circuit of the SC5 II uses 4 inductors to share the work of boosting the voltage and provide relatively high amperage.

I doubt any manufacturer other than Zebralight has put the electrical engineering effort into getting that much output from a single NiMH cell.

Bob_McBob also documented that the H53 and SC53 each only have a single inductor, consistent with Zebralight’s distinction between those drivers and that of the SC5, and with their lower maximum output.

I don’t think he’s modded an SC5 II so far.

Also, I don’t think anyone has so far measured output numbers quite in line with Zebralight’s claims, so it might not really be 550 lumens.

moderator007
moderator007's picture
Offline
Last seen: 2 months 4 weeks ago
Joined: 12/23/2012 - 04:47
Posts: 3739
Location: North Carolina

That pretty much answers my question, I was going to try to couple together a couple of PAM2803’s to see if it would actually work.
That’s how they are doing it, I just dont know the IC used or if there’s alot more going on I don’t know about.
I’ll test to see what coupling two together does excatly.
Anyone had a SC5 apart that might know the IC’s used or if they sanded the markings off?

loneoceans
loneoceans's picture
Offline
Last seen: 2 months 23 hours ago
Joined: 01/08/2017 - 00:18
Posts: 283

moderator007 wrote:
That pretty much answers my question, I was going to try to couple together a couple of PAM2803's to see if it would actually work. That's how they are doing it, I just dont know the IC used or if there's alot more going on I don't know about. I'll test to see what coupling two together does excatly. Anyone had a SC5 apart that might know the IC's used or if they sanded the markings off?

There are a few ways to do this; here is one (not really cost optimized) example: https://www.analog.com/media/en/technical-documentation/data-sheets/3425f.pdf.

www.loneoceans.com/labs/

- Next-gen Switching Drivers: Lume X1 and Lume1
- High Power Boost Drivers: GXB100 GAN 100W, GXB172 17mm 50W
- Older: GXF22, GFS16, GXB17 & GXB20

moderator007
moderator007's picture
Offline
Last seen: 2 months 4 weeks ago
Joined: 12/23/2012 - 04:47
Posts: 3739
Location: North Carolina

That maybe the IC they use. Apparently they’re using 4 inductors and that IC rquires 4 inductors to run in 4 phase boost.
From the example layout of a pcb in the datasheet it looks like getting this down to a 17mm or smaller 2 sided board might be possible.
I looked and looked for something similar to this and you just pop in and lay it in my lap. Thumbs Up
.
If you think you might consider another driver project, I vote for this one. Thumbs Up Big Smile
It be awesome to have some of these small AA lights pushing 2 amps or more from a single Nimh cell with a driver swap.
I would be one of the first to build one.

sidehack
Offline
Last seen: 3 months 2 weeks ago
Joined: 03/03/2020 - 21:30
Posts: 53

My cousin just asked me two days ago about a single-AA-sized light. Might be worth looking into.

waxing twilight
waxing twilight's picture
Offline
Last seen: 2 months 3 weeks ago
Joined: 01/21/2016 - 08:34
Posts: 34
Location: Somewhere near the Mason-Dixon Line, US
moderator007 wrote:
If you think you might consider another driver project, I vote for this one. Thumbs Up Big Smile It be awesome to have some of these small AA lights pushing 2 amps or more from a single Nimh cell with a driver swap. I would be one of the first to build one.

+1!

moderator007
moderator007's picture
Offline
Last seen: 2 months 4 weeks ago
Joined: 12/23/2012 - 04:47
Posts: 3739
Location: North Carolina

I finally got around to tying 2 PAM2803 boost drivers together to see what happens.


Something very strange happens. Instead of getting double the current at the led which one would expect you get 3 times the current.
At about 1.0v in with one board I get about 300ma to the led (xm-l) with both boards paralleled I get about 900ma to the led. At 1.2v in I get about 1100ma to the led with both boards.
A fully charged eneloop using a xp-l2 and two PAM2803 circuits paralleled should get around the 500 lumen mark.
I think getting 2 circuits on one 17mm board with one mode is very possible with maybe needing fewer components.
Making enough room for a mcu for modes might be a little challenging and I havent tested yet to see if 2 boards could be master slaved using only one mcu.
Interesting that the current triples instead of doubles, I haven’t a clue as to why. I’m no EE so I learn by trial and error (magic smoke), so now I know two PAM2803 boost circuits can be paralleled for more lumens from a single cell. Big Smile

calipsoii
Offline
Last seen: 3 months 2 weeks ago
Joined: 03/23/2015 - 11:25
Posts: 24

The reasons you're likely seeing triple the current:

  • Two inductors doubles the magnetic field available to store energy when the FET switches are closed
  • Two inductors splits the power path, making it less likely that either inductor will saturate
  • You've tied the outputs together, doubling the capacitance
  • The PAM2803's are likely switching at very slightly different frequencies (due to manufacturing tolerances) decreasing the off-time when the LED is fed solely from the output caps
  • You've tied the outputs together, putting the sense resistors in parallel and halving their resistance (1/x), which decreased the feedback voltage at the switch IC and increases output.
hodor
hodor's picture
Offline
Last seen: 2 days 46 min ago
Joined: 07/11/2018 - 03:58
Posts: 1057
Location: UK

I don’t know what any of this means but am looking forward to the day an efficient/regulated, high power AA boost driver is (finally) available with customisable modes. Very good and carry on Beer