Thank you for answer, my english dont are very good.
About S2+
In here mentioned, that “All Convoy lights with 7135 drivers use 4500Hz PWM” I think that taken from candlepowerforum so S2+ 4500Hz and E2 3900Hz
nice to see my shower waterdrop test to detect and measure PWM
“E2 3900Hz” it’s not PWM, it’s pulsation (~20-25%), light output don’t drop to zero.
Convoy S2+ (8x7135) PWM is not visible to me, but it’s clearly audible in medium and to a lower extend in low mode. An audio spectrum app in my phone shows a peak at about 4.6Khz (and multiples) if I put the head of the flashlight next to the phone, which is consistent with the frequency of the PWM.
That’s pretty cool so your phone app could identify flashlights by their sound.
That’s how I had to measure it before this fancy Fluke showed up. Well, that… and just eyeballing it. You know you’ve been playing with flashlights too long when you can tell the approximate PWM speed of pulsed light up to 10 kHz just by looking at it.
true. correct, and true
the key point is the light is NOT Constant Current, it is pulsing, strobing, flashing, or whatever other nonConstant adjective you prefer
Does anyone know which Nichia is used in the 4000k E2? I got mine a few days ago and it’s really yellow compared to a Lumintop IYP365 (also 4000k) and not rosy at all. On “high” the hotspot looks almost creamy. I’ve read the 219C has more yellow in its tint, so maybe they chose this and not a 219B like in the 5700k version?
The tint isn’t bad at all, but not what I expected.
Btw, I can’t detect any PWM. It isn’t visible in waterdrops nor can I hear any noise from the pulsation. I was a bit concerned when I read all the post about it, but the actual implementation seems to work fine.
My only other light with PWM is a Nitecore Tube and I can detect it in both “low” and the various “mid” modes.
Falling water isn’t a very effective way to detect PWM. It can work for slow PWM, and it looks really cool under a party strobe, but it’s not good for checking PWM above a couple kHz. And on true constant output lights, falling water can still appear to pulse because the droplets change shape as they fall.
It should be easier to see by waving an index card or similar rapidly through the beam.
Using that method, here’s one at 488 Hz (slow) and then one at 4.5 kHz (what the Jaxman probably has):
I know I should investigate this further just for the sake of it and to give a bit back to the community, but for the moment I’m glad I can’t detect it. 
I’ll try in a few days with a DSLR.
Reminds me of my younger days….
It was 1997 one of my race buddy had a problem with his brand new Honda NSR150R bike. He reported that the temp always went up to 3/4 hot or more. I checked, tested it for several days, and nothing was wrong. Later, he told me than the problem occurred only at the traffic stops (this bike didn’t have cooling fan behind the radiator), the only time he ever look at the panels at the dashboard. This guy never pay attention to the meters during riding, not even the speedo! And the traffic can be real jammed, no moving air to cool all the engines there.
It was the time when fanless water/air cooled engines were the norms. Most engines were overheated during heavy traffic, but only few noticed that because….temperature sensor/meter was something unusual for production bike in that era. I remember when the light turned green, almost all the overheated bikes accelerated with heavy smoke and knocking/pinging sounds!
I explained all the theory behind engine cooling to him and guess what he asked me to do? Please remove this meter out of my dashboard!
OMD! hahahaha :person_facepalming:
- Clemence
We’ve come full circle because the cars I work on now have no temp gauges at all. Try finding a temp gauge in a modern Mini or BMW (non M). You’ll get a check control message as the car is overheating meanwhile the car is already in fail safe mode. BTW I bought one based on this review, although I ordered a 5k neutral I got a 4k warm white and can’t be happier.
Hm, I bought a Jaxman E2 early on.
Mine has no solder on any of the three stars,
and it’s giving Low-Medium-High, with no memory.
Hm, is there a good way to temporarily connect the gap in the split stars to find out what they do on my light?
Maybe a #2 pencil will do it? Nope.
solder and desolder. It’s easy.
You can use desolder vacuum pump followed with solder wick for (almost) clean result
About the UI: I think they (he?) should consider removing strobe / SOS from the default configuration. After using it for a few days, it’s definitely a bit annoying. I can totally understand why most manufacturers prefer a simpler, non-programmable UI without hidden modes. That’s probably what most of their customers want and expect, but I think the assumption that the majority of their customers also wants strobe is false.
The flashlight manufacturers should think of it the same way watchmakers do: They don’t call it features, but complications. Every mode or setting comes at a cost in terms of UI complexity. If the “cost” of a setting / mode is higher than its usefulness, than it should be dropped for a general purpose flashlight in my opinion.
The Low-Mid-High configuration is probably a safe bet and won’t alienate neither casual nor power users.
Overall I’m still happy with the E2 Nichia 219b (4000k). The tint has really grown on me after using it for a while.
I haven’t had time to test the PWM, though.
Sorry for the necro but figured this would be the most appropriate thread (search ranking, this comes up first on Google).
Jaxman has been having problems with 5700K 219B supply. Gearbest, Fasttech, HKEquipment, Jaxman store on AliExpress all out of stock/discontinued for the 5700K Jaxman E2.
Joson Bo from Jaxman store on AliExpress replied to me that they have 5700K 219C’s on hand and that they are preparing the E2 with 219C LEDs instead.
Which is a kind of a bummer in my opinion 
But at the very least there will be more 5700K E2’s shortly but with 219C instead of 219B.
I got some SW57 219Bs from Kaidomain a month ago. They turned out to be very different from the one in the Jaxman. With a (presumably) Carclo optic they were quite green and the only way to make them usable was to put them in an Astrolux S41 with a copper head as a mule. The warm tone helped a lot and now the tint is perfect.
So my thought about the LED compartment colour affects OTF tint or even CRI is valid. In my case with TIR optics, painting the whole space white improve the overall output (especially in moonlight modes) and colour mixing. Thanks Maukka.
Actually putting the optics in to the S41 head didn’t change the tint or CRI at all at 0° compared to the MCPCB+optics being out of the light. It’s only when used as a mule the copper had an effect. And from steep angles the effect is even more pronounced when the light reflects off the side walls.
The looks (when you look directly to the optics) and output do changed. I can’t confirm about the rest, possibly it’s just my perception.