Test results: Convoy S2+ 365 nm torch

enricosavazzi · Post by **enricosavazzi** » Sun Mar 05, 2017 4:33 am

My test results:

http://savazzi.net/photography/cheap365nm.html

If you want to see only the bottom line, skip to the end of the linked page.

Lou Jost · Post by **Lou Jost** » Sun Mar 05, 2017 4:46 am

What a thorough test! Thanks for that.

ChrisR · Post by **ChrisR** » Sun Mar 05, 2017 6:35 am

One S2+ is giving me 0.71 Amps at battery cap.

Battery voltage 4.04V
LED voltage 3.56V

Led 2.53W

enricosavazzi · Post by **enricosavazzi** » Sun Mar 05, 2017 10:00 am

ChrisR wrote:One S2+ is giving me 0.71 Amps at battery cap.

Battery voltage 4.04V
LED voltage 3.56V

Led 2.53W

Thanks for your measurements - there is something very wrong with my original current measurements. I now got .21 A for the MTE and .46 A for the Convoy. I am recharging the batteries, then will try again.

zzffnn · Post by **zzffnn** » Sun Mar 05, 2017 10:15 am

Thank you, Enrico, for your very helpful test results.

MacroLab3D · Post by **MacroLab3D** » Sun Mar 05, 2017 11:43 am

Thank you for your research. Very interesting.

ChrisR · Post by **ChrisR** » Sun Mar 05, 2017 12:01 pm

something very wrong

I usually find I've got the meter probe plugged into the wrong socket for measuring current, or something

Charles Krebs · Post by **Charles Krebs** » Sun Mar 05, 2017 6:51 pm

Enrico,

Thanks for doing this comparison and linking to it!

enricosavazzi · Post by **enricosavazzi** » Mon Mar 06, 2017 12:57 am

ChrisR wrote:
something very wrong
I usually find I've got the meter probe plugged into the wrong socket for measuring current, or something

That is my suspicion too, the +10A input is separate from the general + input, but instead of giving no reading, the meter gives strange readings when using the +10A input for something else, like V. I also recharged the batteries overnight, and open-circuit battery voltage now reads well above 4 V. Maybe new batteries need "training" for a few charge/discharge cycles?

Readings with the UV meter remain pretty much the same with both torches in spite of the higher voltage, so the drivers are doing their job.

johan · Post by **johan** » Mon Mar 06, 2017 2:13 am

Looks like a good torch, what a bargain!

enricosavazzi · Post by **enricosavazzi** » Fri Mar 10, 2017 1:52 am

The page is now updated with current and voltage measurements, and a few other corrections.

It turns out that the MTE 301 is specified byMTE to emit only 450 mW of UV radiation, while the MTE 303 emits 1030 mW. So the 780 mW emission specified for the NCSU276A U365 LED of the Convoy S2+ is completely consistent with my measurements and visual observations, and there are no "alternative facts", miraculously high UV emissions or fabricated product specifications in this case. It also appears that the LED in the Convoy S2+ is not driven above its current limits, although heating during protracted use may still be a problem.

ChrisR · Post by **ChrisR** » Fri Mar 10, 2017 3:52 am

A small thing-
"but unscrewing the reflector of the Convoy S2+ disconnects the LED PC board from the torch barrel"

I unscrewed the brass cylinder (using the outer slots) complete. It brings the board with it.

the picture in the other thread

is what I saw. I replaced the glass with a ZBW2 filter, replaced the reflector, and screwed the brass back in. I can't remember if I tried it without the reflector but I'm sure it would light up.

I've removed 3 leds from torches (not these), and found no thermal transfer paste under the 20mm plate in 2. The other one had a tiny dot as though used only to hold the board in place while soldering. That's plain dumb. The picture above shows some paste by the white wire, I don't know if there's enough for proper thermal contact. Power ratings for semiconductors are usually quoted ONLY for a specified heatsink with a "Degress per Watt" figure. They're meaningless otherwise. Running the leds in the upper half of their operating range without screwing them firmly to such a heatsink is asking for a short life.
Reducing the current small amounts can have dramatic effects on life. It would typically involve replacing SMD components on the driver board - OK if you haven't had coffee.

One detail - the ends and threads of the Convoy aluminium tubes and end caps are variously covered in black coating, so the electrical contact is via whichever parts are clean. Other torches have clean ends and threads. The coating can get in the way.

enricosavazzi · Post by **enricosavazzi** » Fri Mar 10, 2017 4:35 am

Thanks Chris - when replacing the transparent window with a UV-pass filter, it might be a good idea to see if it does not take too much work to put some thermal conducting paste between LED carrier disk and head of the torch.

I have a few ZBW2 filters on order, but it would be good to know whether they block VIS to a sufficient extent. Do you see any blue/violet light if you look into the filter with the S2+ on (of course using polycarbonate goggles, not naked eyes!)?

ChrisR · Post by **ChrisR** » Fri Mar 10, 2017 5:14 am

The paste goes harder with age, so the boards therefore donn't pop up like they do when there's NO paste.
I didn't have any handy (I have now) so I didn't lift the board.

The ZWB2 filters block a lot of the visible light. I honestly don't understand why eg my hand is perfectly visible in very- filtered light

I can still see the chip glowing perfectly well, but my wife can "hardly" see it.
Spectacles stop it almost completely.
http://photomacrography.net/forum/viewt ... 534#206534
A Samsung phone can see it through the filter.

I'll upload a picture if Virginmedia ever works. Mostly it doesn't work when you want it to.

ChrisR · Post by **ChrisR** » Fri Mar 10, 2017 7:51 am

With Samsung S3 phone.
"UV" is a little brighter to the phone than it is to me. Looks the same on the hand though.
Self explanatory I think-
one filter in torch , then
a pile of ZBW2 filters

ZWB1 filter - supposedly:

Does skin fluoresce that much?!