I was getting kind of annoyed that Simon wasn’t replying to my message, but it appears he had actually replied within 4 hours and I didn’t receive an alert notification from aliexpress. :person_facepalming:
He offered to send me a new piece of glass.
I’ve noticed that you don’t always get notifications from existing message threads on Aliexpress.
So i have 2 Convoy C8s, an old one and a new one.
Yesterday i braided the tailspring of the old one.
It has a brass hat (in which i drilled a hole for copper wire), it has a high current switch (parts of thick metal)
Now i’m gonna do the new one.
It has no brass hat but only a conical spring, it has a rather tiny Omten switch with thinner metal parts.
I would say, not clearly better.
In the latest version it is the same as in my new one.
But maybe (apparently) it is good enough.
Any thoughts?
The little Omten can handle much more power than you would think. People have taken them over 20amps in real-world use. There is a thread of tests floating around by Djozz as well.
This^. I have used the stock tailswitch in an S2+ (same as C8 I am pretty sure) to over 17 amps and had no issues besides melting the spring before I braided it.
IIRC in the Djozz test the little switch lasted longer then the large one and only had something like 100mv of voltage drop @ 15 ish amps (been awhile since I looked it up).
Did the switch itself fail or just the springs?
Alright, thanks for the reactions.
It’s good enough then.
Cool. :sunglasses:
Both. It was my very first failed switch due to high current iirc.
If you have a spring bypass in place and get a short circuit, you can then easily melt down a switch internally, maybe even externally. That’s the thing we need to keep in mind when bypassing springs, where’s the weak link? A direct short and at some point the cell is the weak link and then we have big trouble. On really big current drain lights I like to use Arctic Alumina Thermal Adhesive on top of the contacts on the mcpcb just to make sure that if I heat it up enough to flow the wires off the board they’re still not going anywhere. 
Decided to put in a bigger Omten anyway…
Have some in stock.
Skipped the switch PCB too, the braided spring soldered to the long contact strip (braiding inside the spring).
Only have to figure out the best way to isolate the long strip from the tail-part yet.
(edit) done. (/edit)
Did i already say how much i like the deep matte anodizing? It’s :+1: :sunglasses:
A quick question - is that the correct .hex file that I could use to flash old convoy drivers to have new UI? Or is it adopted to the new HW already?
http://bazaar.launchpad.net/~toykeeper/flashlight-firmware/trunk/view/head:/ToyKeeper/bistro/biscotti/biscotti.hex
That’s the one I used to flash some old Nanjg 105D and 105C drivers and it worked fine.
Correct. The updated firmware for the new red drivers is not available yet. When TK posts the revision that’s adjusted for the new driver both versions will remain available. There will be some info added to help people figure out which one they should use based on which driver they have.
Thank you very much for quick answer!
Wellp, a 1A switch can probably handle 5A-10A or so without failing catastrophically (melting, fusing, etc.), but it absolutely does affect lifetime. Every time a switch is, well, switched, small arcs form and make small pits in the contacting surface. This ever-so-slightly increases the resistance across the contacts. And the more pitted and more irregular the surface, the worse it gets with each actuation, and the faster it degrades. That’s why mercury-wetted switches are a must for hi-rel circuitry. Thee Hg “reforms” the surface each time.
It’s like LEDs. Their lifetime isn’t rated ’til they just up and go pouf! Instead, it’s about lumen-maintenance, and when the light output decreases to L70 (formerly L50), or 70% (50%) of its initial brightness.
So when everyone’s bypassing springs and going crazy trying to squeeze out every extra microohm of resistance to maximise current through the LED, the switch will start climbing up and up in resistance, from milliohms to fractional-ohms, and max current will drop until the switch is replaced.
And yeah, that’ll happen probably over hundreds (or even thousands) of cycles, not dozens, but it still happens. By the time it gets to that point, though, unless you’re flicking that switch a lot each day, it’ll be time to toss that light and get an XM-L5 in a Z6 brightness bin.
—
Eh, here ya go. Not complete, but I just took a quick look around…
So, in other words, worry more about the carpal tunnel surgery than the $0.30 switch replacement. 
I’ve seen people here grexing about limited current through their LEDs, swapping cells, bypassing springs… and they they measure a too-high resistance in the tailswitch. Well there ya go…
At 10:1 or 20:1 overcurrent, that’ll happen for sure, and every mode-change counts as an actuation.
But hey, those LEDs will be happier with lower current, so have fun!
I saw the new clear C8 at gearbest .
I found a few mistakes in description. They writed everywhere XP-L H1 instead of XP-L HI where HI means high intensity.
The modes described the old firmware not the new one, but on the pics you can clearly see the red driver board.
Yesterday I received my new clearly better clear C8 and it doesn’t seem to me as better, it’s nice looking thou, but I have all this issues already mentioned: dirty lense from inside, large particles of what seems to be plastic material. I bought several old style C8 flashlights for friends and except occasional tiny reflector imperfections all was good.
I will upload my pictures later but it’s all as described by other members.