Amutorch AX1 Quick Look

I just received a newly-released flashlight from Banggood. It’s called the AX1 and it’s made by Amutorch.

It has a thoughtful, attractive design, and the overall execution is great. But the example I received has a few flaws that really detract from the experience.

First though, some quick observations:

  • Unusual, attractive, well executed, two-color design.
  • Attractive body form, with thoughtful, unconventional, functional details.
  • Large, 22mm, reflector for a 18650 tube light without significant added exterior bulk.  I measured the outer diameter as 25mm, which is just ~1mm larger than a Convoy s2+. The inner diameter of S2+ reflector, though is ~17mm, 5mm smaller than the AX1s. As another point of reference, the Zanflare F1 has an ~20mm reflector, while the OD of the bezel is about 27mm, and the max OD of the whole light is closer to 30mm.
  • Good action on the reverse-clicky switch.
  • Modes seem good. The AX1’s low is pretty close to the low on a Convoy S2+ with an 8×7150 3/5 mode driver. The high/turbo is similar to other ~3A ~1000 lumen lights. There is strobe, but it is pretty well hidden. It has mode memory.
  • The beam pattern is quite nice. Much less floody than a standard S2+ with an XM-L2 emitter and an OP reflector. At 10-15′, the hotspot and spill are much closer to an S2+ I modded with an XP-L HI and a SMO reflector, but the AX1 has nicer smoother, corona with less tint shift, though not as nice as it would be with an OP reflector.
  • I’d call the color temperature neutral white and without an objectionable tint.

Now, for the big problem with the AX6 I received: The surface texture on the blue head and tail pieces is visibly, glaringly, uneven. I had some trouble photographing it, but it is quite apparent when you have the light in hand. When I first saw it, I thought/hoped that it was some surface schmutz. It isn’t, it’s the metal.

Note the machine marks on part of the piece. This is one of multiple areas where the surface texture is inconsistent/in-complete.

It looks like it was supposed to receive some sort of uniform finish, like bead or sand blasting, but some areas were barely touched, so the machining marks are obvious. It doesn’t seem very even over the length of the head either, nor is it consistent around the axis. The problem is most obvious on the head of my light, but the tailcap also has similar but more subtle problems.

In addition, there were a few nicks on the body. The photo above shows the largest of them. These are unfortunate, but the truth is with a weeks use, I probably won’t be able to distinguish them from other wear and tear. The inconsistent finish is going to be obvious for a long, long time.

So, bottom line, this is a nice flashlight for $20-25 if the surface finish is as it should be. Mine isn’t. So I’ll be asking them to replace the faulty parts, if not the whole flashlight.

Digoo IP Camera Teardown and Links

I was just looking at unfinished posts and noticed that I’d taken, but not published, a bunch of notes I’d made earlier this year in hopes of hacking better firmware onto the Digoo BB-M2 WiFi PTZ Security Camera.I gave up on the quest, but here are my notes, with minimal editing.

Someone mentioned that the Chinese language page for Netcam360 has a link to the IPC-SDK. When I downloaded it and looked inside, I saw client-side code, but there was also a self-extracting archive called “HSmartLink Win32 SDK” and I remembered the PCB marking started with “HSL.” Searching for HSmartLink brings up hsmartlink.com, which, among other things, has IP webcams! The i9812 looks like a good match for my camera!

Unfortunately, no sign of any firmware updates. I checked the .cn version of the site too. It doesn’t seem as up to date on products (i9812 isn’t listed), and while there is more info in support section, it is still quite sparse. Page that looks like it is intended to link to downloads hasn’t been updated since 2015

Company is “Shenzhen Hsmartlink Technology Co. Ltd”

FCC Database listing for company

So what is the relationship to NetCam360 (check whois & IP ) and they mysterious APKLink?

Nmap Pobe

Starting Nmap 7.40 ( https://nmap.org ) at 2017-02-06 18:38 PST
Nmap scan report for 10.31.1.124
Host is up (0.0043s latency).
Not shown: 998 closed ports
PORT STATE SERVICE
23/tcp open telnet
81/tcp open hosts2-ns
MAC Address: E0:B9:4D:8F:61:6C (Shenzhen Bilian Electronicltd)

Nmap done: 1 IP address (1 host up) scanned in 0.49 seconds

Teardown Inventory

Photos

Ingenic

Module with Mediatek MT7601UN

Unknown

  • 15UDN8WY, ULN2803AG 18Pin in 2-row SMD
  • Darlington Transistor Array: http://www.ti.com/lit/ds/symlink/uln2803a.pdf
  • Probably used for driving PT motors

Atmel 542

  • 24C02N, SU27D
  • AT24C02
  • Two-wire serial EEPROM (2K)

Vertical PCB

  • SF1810-002, www.sufeitech.com. PCB antenna?

oosilicon or dosilicon?

Other

Balight 21W Folding Solar Panel USB Charger Partial Teardown

I picked up a 21W, 3-panel Balight folding solar panel-based USB charger from Amazon for ~$36 a couple of weeks back. It uses high-efficiency SunPower Maxeon cells much like similar 20-21W panels from AukeyAnker and dozens of obscure brands. All of them have the same basic construction. They are all made from nylon ballistic cloth. Each fold has a panel made from two SunPower cells encapsulated in a flexible waterpoof sheet. The panels provide power via two 5v USB ports, which presumably have some sort of voltage regulator.

I wanted to know more about how the chargers worked. In particular, I wanted to know if they were wired in series, or parallel because I wondered if it was worth trying to tap into the raw output, before the USB regulator to reduce power conversion and resistive losses for some applications.

I thought I’d be able to get the information I needed by finding someone documenting a teardown of their own panel on YouTube or a blog post. Despite the dozens of variants from dozens of brands and a handful of manufactures though, I didn’t find what I was looking for.

So, I decided to dig up a seam ripper and open my panel far enough to get a look at the wiring, and tap in to it upstream of the voltage regulator.

The panels appear to be wired together with some sort of woven wire conductor. I had some hope that all the cells would be wired in series, to give a nominal panel voltage of 18v. Based on what I could see, and measuring the voltage before the regulator in full sun, it looks like each panel is wired in series, for 6v nominal voltage, and then the panels are wired together in parallel. I was disappointed at first, but this arrangement makes sense in upon further thought.

Using a 2s3p configuration means that the input voltage into the switching regulator should be pretty close to the 5v (actually, 5.2v with enough sun and a light enough load) output of the USB power regulator, which will typically have higher conversion efficiency than 12 or 18 volts. It also means that the manufacturers can stock one converter for everything from a 7W single-panel charger, up to a 28w 4 panel charger without the converter having to support a wide range of input voltages. Perhaps most importantly, it means that partial shading of one panel shouldn’t have a disproportionate impact on the power output of the entire array.

The only downside is that resistive losses in the cabling will be higher with lower voltage and higher current, but that the interconnects aren’t more than a foot or so, the resistive losses shouldn’t be too high.

As for the converter itself, I may look at it more closely and add some more details, but, a few initial observations:

  • The PCB design has extensive ground planes on top and bottom, tied together with vias.
  • Both outputs are served from a single buck-converter (step-down) power supply based on a Techcode TD1583, which is a 380 KHz fixed frequency monolithic step down switch mode regulator with a built in internal Power MOSFET.
  • It looks like only port 1, at the top right in my photo, has the data lines connected, which suggests that it is the only one with fast-charge coding.
  • IC U2 looks like it has its markings sanded off. I notice though that one of its pins is connected to the enable pin on the TD1583, leading me to think that it is responsible for cycling the output to make sure devices draw as much power as possible when the panel voltage rises again after clouds or an object reducing the light falling on the array pass. I don’t know if it is a MCU, some sort of timer, or comparator, or what, though.

There you go. I can’t be sure that other folding solar arrays like this one are wired in the same way, but if they only support a 5v output, I suspect they will be. I hope this proves useful to someone besides me.