Today I got my 2.5 mm brass screws delivered (very fast 11 days shipping), but also disappointing cracks in the screw heads.
Would you go the negative feedback route?
XP-L2 and XP-L W2 are different creatures. I have both, they are obviously different. The XP-L2 is along the lines of the XP-G3, the inverted die with no bond wires. The XP-L W2 has visible bond wires and the dots on the phosphor not seen on the XP-L2.
I was as surprised as anyone to see that 10,212 lumens, believe me!
When I re-flowed the XP-L2’s onto the Q8 MCPCB I didn’t realize the thermal pad was sunken. I didn’t add solder paste so I ended up overheating the LED’s. I think that’s why I am now “only” seeing 8728 lumens instead of the 10,212. If I’d had more of the XP-L2’s I would have used new ones again after I cleaned the MCPCB and used a solder paste mask to get the right re-flow. I re-used the LED’s that had started glitching, and here I am.
It runs fine. Long click off, no new e-switch specific features yet like on-press shortcuts or lockout. Not that it matters for the Q8, but it fits on an attiny25 still. Download includes precompiled hex and click to compile and click to flash scripts.
I find it extraordinarily amazing how a year’s worth of speculation and conjecture culminates with a finalized item that is scrutinized not for it it’s excessive delivery of refinements but for the simplest of choices… the screws used in assembly!
Lot of this stuff was already posted, just hard to find, and not centralized and confirmed. I didn't organize this myself til just now. I forget what I posted, what I took pics of, what I wrote down notes for, etc.
But if it's the main thing to critique, that's not bad...
My concern was the way the PCB was warped by overtight screws, and the poor contact
made with the battery tube because of the warped PCB and rough metal edges on the battery tube
Understood Hank, in the first batch of 500 there were those issues, but they paid attention and corrected the problem such that the 3 Q8’s I saw were not affected in like manner. By having a slightly larger hole in the PCB of the driver, the mount screws don’t lift it or tilt it, firmly securing it to the shelf. Same as the spring’s pcb in the tail end of the Q8, the screws used mounted it securely and were properly applied.
As Tom intoned, if the screws are the biggest problem we are doing really good! Much more output than originally intended, better copper MCPCB, excellent emitters, great reflector, lighted side switch, and the driver components stick with the tried and proven values as prescribed by Del and Tom, so the successes here are larger than the failures, lessons were learned and adaptations applied. Win win, great overall interface with a known Manufacturer and certainly an encouraging joint effort.
I am really liking the batch 2's. This is the 3rd one I modded, but for the first two, all I did was the tail mods (bypasses, etc.) and FET replacements. Now that I got in the brass screw replacements for the batch 2's, figured I'd go further with the mod.
Did my usual 20 AWG bypass method with tossing the inner springs, tinning the rings around the screw holes. See where I'm making the pads on the traces? I'm backing off a little further from the screw hole, hoping the traces will be my acting fuse:
Replacement brass screws M2.5 x 5 mm shown. I'm still sanding down the shelves the screws thread in to, and using light coating of NO-OX-ID:
Here's the brass M2.5 x 10mm's and the M4 x 10mm shown. Notice the shiny copper of the MCPCB? It's sanded from 400 to 2000 GRIT, and so is the shelf, then using Arctic MX-4. I got quite a bit squeezed out shown there, more than I thought. I screwed in all 3, but cranked down on the center reflector screw first. The LED's in this light are the best centered I've seen so far in any Q8. I always struggle with batch 1's to get them centered. The stock 18 AWG is replaced with better 18 AWG. The stock ones are stiff, less strands, while the replacement wire I got from IOS a while back: 200C rated and 150 strands. The stock wires in batch 2 are much shorter at 45 mm, batch 1 are 62 mm. I made the replacements 47 mm, but if I were to do it again, I'd use 48 mm. I think 48 mm will be about right for the shortest possible, but still able to get at things easy enough:
How here's something I haven't see noted yet. The Batt+ contact ring is not flat - it slants down inward. So I got about 2/3's of it level by sadning using 400 GRIT paper initially, then smoothed out to 1000 or 2000 GRIT. You can see below the shiny portion is sanded level:
This Q8 got the Infineon FET, shown below. I'm still thinking the SIRA20DP is slightly better, but I got 10 of these to use:
In addition to the above mods:
use a bigger drill bit now to bevel the drill holes if not done so already for the critical screw connections
stayed with the stock driver screws - no point to replace them
this one has regular glass lens, no AR coating
sanded smooth the battery tube edge that makes contact to the driver, to 2000 GRIT
So what about the results?
On 30Q BT's @41.9V, lumens: 7510 @start, 6970 @30 secs, 70 kcd taken at 5m (529 meters)
It's the best results I got from a Q8 so far - new record for me.
I made mechanical reverse polarity protection ring. I cut a piece of plastic and thinned to fit. The height of the brass ring is 1.5mm and thickness of plastic is 1.8mm. I did not use any glue, just used a little more pressure to get plastic into the position.
For those who have access to the 3D printer I have made a 3d model.