DANIU/Banggood Swingarm LED Magnifier Review: Hackable and a STEAL at ~ $25 US

[mnementh]

            

Daniu 33cm/105mm LED Magnifier Lamp on Banggood (affiliate link)

5 weeks ago I bought this LED magnifier on "preorder"; with Canada Post Direct shipping it came to CAD$32 or a hair under US$25. I'm certain the same magnifier lamp is available elsewhere under different names; shop carefully and look closely for the spun aluminum head with 330mm/105mm advertised size. There is ANOTHER model that looks very similar, but it is like the "midget version"; smaller head and much shorter swingarm. This model is a very close approximation in size & construction to the swingarm lamps we all had on our dormroom/drafting desk back in the day.

 It shipped a week later than promised so I missed it for Christmas, but now they're in stock. It arrived in a pretty beat-up box, but the unit inside was unharmed. Bubble-wrap and a custom cut EPP foam brick around the head. 

First impressions: Good usable size, all metal arms, cast alloy base, even metal knuckle joint for the lamp itself.  Base was not user-friendly in my case due to my desk having unmovable brackets in the way; however I had a resin base that I could get to clamp down. Cast aluminum base is better-than average build quality; in fact, overall that's how I'd describe the whole thing.

The LED ring/lens head is made from a single piece of spun aluminum; inside is a dual-LED illumination ring with Blue-White & Warm-White LED channels. You can turn on and dim either color channel or both at once. The lens is held in by a threaded ring, not clips or adhesive. 

However, the illumination level is "borderline"; those of us used to using one of the old-school versions with a 22W T9 Circline bulb as both magnifier and swingarm desk illumination will be sorely disappointed. For most uses as illumination on what you are viewing, however, the LED ring is fair to good; I found that by removing the diffuser ring as seen in the video below, illumination improved greatly and even was useful for lighting up the desktop, though still not even close to a common 6W flood bulb in a desk lamp.

Demo video showing lamp/lens assembly: https://cloud.video.taobao.com/play/u/3388301779/p/1/e/6/t/10301/234633372172.mp4

They do not include a USB power pack; this saves you money but you have to source your own rated at 5V/2A. I'm using a 1.5A rated one and it does not seem to load up at all; after running an hour it was barely warm.

The lens itself is a nice ground glass unit 105mm/4in diameter; however the 5X advertised is entirely wrong (I knew this just seeing the lens in the video, though); measuring the thickness it is more like 5-6 diopter, or approx 2.5X magnification. This is what I was expecting, so not really disappointed there; a 5X lens this large would tend to distort the subject so much as to be pretty useless for my purposes anyways.

All in all, excellent build quality right up there with many of the US name brands; not Dazor quality, but definitely a whole lot better than the pricetag.

   

And finally, here's the glamour shot; my Christmas present to my bench doing its job.  Good reach, good enough lighting, plus a STEAL at approx $25 US.

HACK/MOD TIME:

I like this Magnifier Lamp a lot but it seems a little dim, especially considering the BOM and LED density. I've spent some time puttering around with it, and I have come to a not-so-surprising contusion: It has constipated electrons.

      

This magnifier's LED circuit uses 32 3020 white LEDs; these typically drive at from 0.2-0.5W each, with a 60-150mA current at 3.0-3.4V_Fd. Even the cheap, single die/package versions are nowadays rated to dissipate 60mA/0.2W per LED at operating temps up to 80°C. TL/DR Takeaway: We should be able to get approx 6W out of EACH lighting channel in the lamp, and there are two; one for WHITE and one for WARM WHITE.

Some quick measurements indicate a different story, however; plugged into a 5V/1500mA rated USB charger from my old smartphone, I find that with 5.2V measured at the jack under load, the entire thing is drawing .847A, with 3.657V actually making it to the +5V rail at the LED ring. Investigating the intervening wiring, I find the wiring from the LED Ring to the controller is 22ga and the main power/USB plug is only 24ga! 

Popping the covers off the controller (pleasant surprise: it's held together with 3mm press-fit pegs and no glue ) I find that under load (single color) at MAX BRIGHTNESS it's getting 4.75V at the controller IN & sending 4.67V out to the LED ring. This means that we're dropping .08V across the controller; the rest of that 1.54V is being dropped in the wiring. 

This lamp is driving a 4S8P configuration 32-LED Array for each lighting channel at 2.93V_Fd/.847A with each 4S LEDs driven through a 6.8Ω ballast resistor; the WARM WHITE channel draws a few mA less & slightly lower LED V_Fd, but not statistically significant variance. Plugging these figures into whatever online LED calc you have handy will reveal that they're being way underdriven (possibly even to the point of actually losing luminous flux efficiency rather than gaining) and the total wattage reflects that:

MEASURED:
Source (E) =3.657V
Ballast (R) = 6.8Ω
LED V_Fd  = 2.93V

CALCULATED:
LED (I) = 0.1069A (÷4 LEDS/Ballast) = .0267A or 26.7mA
LED (P) = 0.0783W x 32 LEDs = 2.506W Total LED Output/Channel (White or Warm White)

This is less than half what I estimate it should produce, based on the design and the 5V/2A power spec.

At this point I decided to do some quick reality-check testing; powering the controller directly from my CV/CC power supply, I do a number of tests starting at 4.5V and ramping up to 7V in. This last test yielded approx 4.5V at the 5V rail on the LED ring, and LED 3.05V_Fd/64mA. However, as you would expect from the 2.5V drop across the wiring to the LED/Lens head, that wire got noticeably warm after only 30 minutes.

So... I dismantle the LED head and install some 18ga silicone wire (affiliate link) made for mini-quadcopter battery leads; I decide to relocate the controller to the arm by the LED head, so those wires only need to be ~200mm long & easy to work with.

The results are actually pretty impressive. First, using my little homebrew CV/CC Power supply, I test the newly rewired LED/Lens assembly at 4.5V set/4.25V measured at 5V rail on LED ring with both channels lit. This yields 2.16A total current draw with 3.005V_Fd/46mA at the LED. This calculates out to 8.85 watts total with both channels (64 LEDs) lit!

Illumination at this setting is excellent; perceived brightness is more than double previous tests, and after 2 hours the LED/Lens assembly is just warm to the touch and I can hold the back of a finger to all LEDs & resistors comfortably. Quality of light is excellent as well; the White channel is pretty blue and the Warm White channel is pretty pink (to be expected with the cheap no-name LEDs probably used here), but driven in parallel the color is a good approximation of Daylight white. 

Encouraged, I bump it up to 5.2V set/4.85V measured at 5V rail on LED ring with both channels lit. This yields 3.15A total current draw with 3.09V_Fd/65mA at the LED. This calculates out to 12.86 watts total with both channels! The perceived brightness is again double previous test, while after another 2 hours the LED/Lens assembly is warm to the touch and I can hold the back of a finger to all LEDs & resistors, with the resistors just uncomfortably hot but not blistery. This is driving the LEDs & ballasts a little harder than I like, but the test does does prove concept.

Adding the Controller in circuit these figures remain approximately the same at 5.2V CV setpoint for single channel operation; output of either WHITE or WARM WHITE channel is approx 6.4W. However, when both channels are enabled the PWM controller is programmed to reduce max output to approx 60% Duty cycle on both channels; this drops total output to approx 8.4W @ 2.16A total current. meh. But it DOES validate the 5V/2A power supply design spec. 

This indicates to me that my original thumbnail guess of this lamp being a 6W/Channel design is pretty close to correct, and that the lackluster performance is entirely due to losses in the cheap wiring. Running in single or dual-channel mode, even with 7V input in my first tests, the FETs & driver semiconductor on the controller never heated up; they were barely perceptibly warm after 2 hours of runtime in either mode at Max Brightness.

A good experiment; it shows you can get excellent illumination from this LED magnifier simply by rewiring with 18ga using the original controller and a good regulated 5V/2A power pack or wall-wart.

More importantly, however, this lamp's easily dismantled design lends itself to modding for a completely DIFFERENT objective: Low-noise illumination.

If you've been watching Dave for any length of time, you'll know how he loves/hates (loves to hate...? ) the little PWM controllers that are used in all this sort of LED lighting; they create massive amounts of common-mode noise and in many cases will function as unexpected/unwanted RF noise radio stations all around your bench!

My testing above indicates that if the controller & associated wiring is removed entirely, and the LED/lens head powered directly from a nice, clean linear-regulated PSU good for up to 4.5-5.0V@3.5A, with minimal assache you can get excellent EMI-free illumination from this LED Magnifier for only $25 plus the cost/time investment in rewiring the lamp with 18ga & building a half-decent PSU. w00t!

Tinker-ily fun AND great value!!!

Cheers,

mnem


DISCLAIMER/WARNING: I use some clearly identified affiliate links in this review.
If you buy from them it costs you not a penny more, but I get a few cents worth of points which I can use to help pay for weird sh...tuff from banggood, which is exactly how I bought the items linked here.
If you don't like this, please delete the affiliate suffix on the link in your browser or go to banggood.com (or your favorite similar website) and search the item up the hard way.