Has anyone used this driver? How bad is the PWM?
CF 17mm 0.9V - 3.3V 4 Groups 2 to 5-Mode Boost Driver Circuit Board
I am looking to rebuild a very old 1xAA light and want to add the ability to use a single 14500 as well.
The other concern is that it is only rated for 3V which would limit use to an AA or 3V lithium primary, not sure it would take a 14500.
I have the CF 17 in a mini-mag mod and it has been working well thus far.
It definitely uses PWM on Low and Med that I can detect if I vigorously swing the light back and forth but itās not noticeable (to me) in normal use.
On 1x14500 it continues to have all 3-modes (each brighter than when using 2xAA) but I normally run 2xNiMH so I canāt say about longterm. On the plus side itāll even run on 1xAA!
āHas anyone used this driver? How bad is the PWM?ā
I have used that driver. I am pretty happy with it. The mode spacing is not ideal in my opinion. Low is about where I would like medium to be. You also have to wait a painfully long time before turning the light back on in order to avoid advancing to the next mode. The PWM is audible to me but I have pretty good ears so Iām not sure how noticeable it would be to most people. It is pretty noisy to me though. I donāt āseeā it though. I have run it a fair bit on an 18350 and single Cr123s and had no issues thus far. I did not run it long on a single AA but I did some when outside of a host for testing the modes. It seemed to run great on it. For the price I am happy with it and donāt mind living with its quirks.
Edit: Sorry I donāt know how to do the usual in-text quotation.
pc_light and Superstocker, thanks for the reply! I am targeting using it with a single AA. Sounds like it might be a decent option.
KD, I see that youāre offering dedomed XHP50.2 emittersā¦but just a single tint.
I would be interested in a dedomed variant of
XHP50.2 6V/12V 3000K CRI80 J2.
Would that be possible?
I have a few of this varrant and I have never noticed the pwm. But I dont think there the exact driver in your link.
This should be using the PAM2803 boost ic. If the battery voltage is over the vf of the led it runs direct drive with the current running through the inductor, FB resistor and diode. They are all in series tied to a Fet which is tied to the led. The mcu controls the Fet for modes. All these components in the leds path has some resistance not just the FB resisitor. So it is limited in current in direct drive. I have never had one stop working with both battery types but if you wanted to add a little more assurance you could change the FB resistor to a little higher value. Pic looks like a R120, for insurance you could change it to R150 or more.
mattlward, moderator007's comments about the CF 17mm 0.9V ā 3.3V 4 Groups 2 to 5-Mode Boost Driver Circuit Board are on track. I didn't mess with it, but it looks about the same as the typical driver found in run-of-the-mill SK68 flashlights (same R120 sense resistor too), which in my experience works well with high drain TrustFire IMR14500 cells (same high drain cell as everyone else, check for example Vapcell INR14500 650mAh (Blue)).
Oh! I once tested the SK68 boost driver I speak above and it didn't feature any low voltage warning for li-ion cells, and while the above driver isn't the same this likely means it will run 14500s dry unless precise care is taken or protected cells are used. 
Wed, 11/13/2019 - 19:01
Need to add negative experiance with KD. Last 2 orders came with missing (several) items and after sending several emails, even PMing BanL, all iāve got (every time) āWhatās the problemā. And now happend with 3rd order also.
FFS, level of not giving F...K is amazing.
that being said, do not order more than several items at once.
You will never get a reffund or Sorry.
They even marked with X missing items...
Recently purchased and received KDLITKER QUAD 4QP-20 DTP COPPER MCPCB FOR CREE XP SERIES only to realize they are not compatible with Carclo 4up TIR lenses.
Does anybody know which TIR lenses work with these quads?
Its right at the bottom of the page you linked. 
If it had been a snake, you be bit. 
I purchased a couple a few weeks ago.
Oh boy do I feel silly! Thanks for pointing that out.
The real answer is none. Donāt buy them. Look back a few pages and save your money.
Or if you want to mod your parts before you even get to mod your light go ahead I guess 
I wish I read those posts before ordering. I made the bad assumption that a 20mm quad mcpcb is standardized and clearly it is not.
I just got in one of the E6 P60 hosts. It is very well build, feels good in the hand and parts fit very well. I do have one complaint with itā¦ The battery tube is a whisker to long, unprotected flat top cells do not work well or depending on the cell do not work at all. I could maybe add a longer spring to the tailcap, but the spring on the dropin is already really long. KD uses some really long springs on dropins.
This is not a huge problem as most P60ās do not draw tons of current, but for me I have very few cells that are long enough. So, I need to find longer cellsā¦ suggestions for either a good protected cell or for a longer 18650?
Well, iām happy to read there are finally some quad boards available for the Manker / Astrolux quad optic.
A pity that they donāt fit each other too well thoughā¦
The fiddling and finicking never stopsā¦
And i hope one day most drivers have proper mode spacing againā¦
But itās often due to that silly moon mode that they sacrifice a low or low mid mode, so iāve noticedā¦
Modes should increase by 300% (current) every step.
Look at the giant jumps most drivers make from moon to low or from low to mid.
Dāt think so.
Thatās 1 - 3 - 10 - 30 - 100 %
Those are not too big and not too small steps.
I like those steps...
And when the number of lumens triples, it only appears to be 73% brighter because of the way the human brain interprets the amount of light.
Yeah, something like that.
And for a 1000 lumen light 3 modes like this is quite enough.
Youāll want an extra 3% low mode for 3000 lumen lights.
But often one of the steps is sacrificed for a moon mode nowadaysā¦. :person_facepalming: