I’m talking about bench power supplies.
I just tested with one of these CNQ drivers, with an R200 stacked on top of the stock R100. I also have some thermal padding that I sandwiched between the inductor/toroid and the board:
I got:
Input-side: ~7V @ 1.92 amps
Output-side: ~3.2V @ 2.72 amps
So, so far we have the following:
==
wight:
Stock CNQ driver:
Input:
7.10v @ 1.08A
Output:
3.10v @ 2.00A
Efficiency:
~80.85%
==
ohaya:
Stock CNQ driver:
Input: < 10V @ ? A [Sorry, I didn’t note the exact input V and I :(…]
Output: ~3V @ ~2.3 A
R200 stacked on top of R100:
Input: ~7V @ 1.92 A
Output: ~3.2V @ 2.72 A
Efficiency: ~64%(?)
Earlier, I had the 1st CNQ driver in a light, but it’s been underperforming (lower than expected current), and I’ve been trying to diagnose why.
I was doing some testing of the pill with the driver in it, and I had made a copper collar to try to improve the negative connection, ala comfy and also Ima4Whlr:
but when I tested, I noticed that if I put pressure on the driver (pushing it into the pill), I’d have this strange behavior where it’d go through the modes, but when it got to high, after a sec or so, voltage (on the PSU display) would drop and current would jump way up, like to 5+ amps.
This happened a few times, then I finally decided to take the pill apart (the driver was soldered into the pill), and it took me awhile, but I think I found the problem. Some pics:
Can you see the (potential) problem :)??
It looks like that R200 resistor on top of the stack is cracked :(!!
I’ll be replacing that obviously, then (ugh!) trying to re-solder the driver into the pill again…
Jim
Some updated info:
The following were slightly different in that I was using triple XP-G2 emitters on Noctigon instead. Also, input was a pair of red laptop pulls in series rather than my power supply. Output current was measured with a clamp meter:
2xR200 stacked on top of R100:
Input: 2 x 18650
Output: ~3.7 A
3xR200 stacked on top of R100:
Input: 2 x 18650
Output: ~4.72 A
Also, a couple of observations:
- There’s noticeable audible whine in medium mode
- The low mode is actually pretty low, almost a moonlight, which I think is pretty nice.
EDIT: I have some thermal padding material sandwiched between the inductor and the top (component side) of the driver board.
I wonder if the ld 40 has the same key solder pad as some of their other LD-xx drivers
No. Those drivers have an MCU, this one pretty clearly does not.
If that is the case then how are the different modes generated? And how are the different mode orders from different vendors set up?
An integrated circuit generates basic PWM at different duty cycles. Note than none of these things have functioning memory. That’s because there is no NAND (no flash), something which is required for memory to function. Next mode memory does not count since it isn’t remembering anything - just moving forward (and also it resets given enough time… because it’s not memory).
When there are flashy modes you’re looking at an integrated circuit which skips the PWM and just generates pulses of different lengths, no doubt using the same one or two timers.
All very basic and something that can be built from discrete components with a few timers. It’s not really my thing, but what we are discussing is probably multiple orders of magnitude less complex than an MCU. Basically I think we are talking about a small pile of something similar to 555 timers and flip-flops forming a bunch of latches I suppose. The “modes chips” used in drivers like these have very simple behaviors.
Ah, I think ill pick up a couple.
I received my LD-40s today. Can someone confirm whether the PWM is visible on low or medium with this driver? I seem to read a post stating PWM is visible, but I cannot find them now.
If PWM is ok, then I might reconsider using this driver than the stock Dipper D19 and stack two R100 on top of the existing R100 to get me a theoretical 7.2A to the LED. I’ll be using an MT-G2.
My LD-40 from CNQ has visible PWM on low and medium. I also has quite annoying whine on medium and high.
I bought this LD-40 from IOS, link here.
Put it in my UF 1405 and it zapped my XM-L2 when I turned on my light! Replaced another XM-L and it was zapped again. 
Definitely the output voltage is too high. I went to IOS website and they didn’t say it is suitable for one LED operation. It only describes “You can power 2*XP-L / 2*XP-G2 / 2*XP-E2 in series, MK-R 6V, MT-G2 6V or other higher Vf LEDs while using 2-3*18650 or 2-3*18350 cells in series.”. Is the IOS LD-40 different with CNQ one and its output voltage is always above 6V?
I would be getting directly in contact with Hank and post his answer back here. Possibly there is a short somewhere causing a direct drive to the led?
This driver will certainly drive 3-4v emitters without issue. It’s not in the description because it’s intended to be implicit (note that he specifies that it will work with 1x18650 - since it’s a stepdown driver that must mean that it will work with LEDs which have a Vf < 4.2v). If I were writing the description I would have explicitly pointed that out - or if I didn’t, I’d correct it later. 
I agree with MRsDNF, contact Hank @ IOS. It sounds like you got a bum driver.
Take a close look at the negative emitter lead and the inductor it solders to: they should not short to GND, but they are close to the GND ring. Also visually check for shorts to the GND ring at the 8205s which is next to the R100 sense resistor. While you’re at it confirm that your sense resistor is marked R100 and that it does not short to the GND ring. If all of those are good, I’d assume it’s a defective QX8205, but who knows.
Thank you guys. I just checked the driver visually and seems like all the points are fine, and it has R100 sense resistor.
Now if I want to test this driver again is there any method I can use without using any LED? For example can I hook my DMM on the driver’s output leads and measure its output voltage? Does the driver know what kind of voltage it should put out by doing this? Because I don’t want to lose any more LED.
It’s a current controlled driver. When current is lower than the “set current” (the current dictated by the sense resistor) it will increase voltage until it maxes out or until current reaches the set current. If you measure it with nothing attached to the outputs you should see approximately full battery voltage (minus a little bit).
You could try hooking it up to a 2s1p battery and an MT-G2. That wouldn’t tell you a lot though. If current was much above 2.0/2.4/whatever amps with the MT-G2 you’d know something was wrong… but you already know that something is wrong. (it has killed 2 emitters, something is wrong!) The normal 6v MT-G2 can survive direct drive with 2s and I think it survives pulses from a 3s1p battery (I have not done this), so there should be no danger in doing that. There is also no point in my opinion, your driver is bad and you’ve inspected it for all the types of repairable manufacturing errors I could think of. Maybe someone else will think of something.
Also be sure that the two legs of the inductor do not short together. Again: only a visual inspection is appropriate.
No.
I don't think you can test it with just your DMM without something else because the driver is current regulating, not voltage regulating. I think you need some sort of a resistive load.
edit: great minds think alike (or the blind lead the blind 
)
hahahah, that’s another way to put it! 
That's about as technical my mind gets. Trust me I've asked more than one member here lately what to use to simulate a load on a driver to prevent leds going poof.