If they only had the PCB stars separate, I need to try to source a couple for the lantern project testing. (easier to switch out the PCB than it is to re-flow the LEDs back & forth on the single one)
- Do any one know where i can get or source two bare PCB stars for the Q8? I need two for LED testing in the BLF Lantern project when i go on my east coast trip in a couple weeks. (BLF lantern project is here: *BLF LT1 Lantern Project) (updated Nov,17,2020)
Same issue, overseas shipping to Canada will take longer than i have before the trip, and that board is not compatible with the 3535 XP-L or LH351 series chips.
Damn late to the party, everybody has moved on. I’ll just throw in my fifty cents…
Did me a few mods to the Q8, naturally the tail spring bypass. Had tried some braided silver to a center tie point. That didn’t work out as the silver is really conductive and just sucked up the solder making it a solid mess. After some thought and as I wanted a clean look, designed a solid copper coil (spring-like but not really springy) that I fitted within a steel cylindrical spring. Copper diameter .030”, 5 turns, which should be good enough for some 6 amps without too much resistance. I had tested the conductivity (subjectively) of the original spring and the contact point to the battery is a line at best (the end loop of the spring). By soldering a cap there is more contact area. This added feature gave a noticeable increase in the light output.
I had trouble testing the current drawn at the tail end. Conventional method thru DMM has too much loss through the internal shunt. By using a standardized external shunt calibrated to 50 mV @ 50 Amps - tested using a 1500 W resistive load and my trusty AC/DC power supply, and with my own tweak on a 250 µA Moving Coil bi-directional meter. Turbo result: some 17 Amps. Yes, this method is not accurate as a clamp on DC meter (UNI-T UT210E Digital Clamp Multimeter/Ammeter). If I were to be testing often or many, the purchase is unavoidable. Anyways, the results are in the ballpark.
I had made a good electrical path thru the body via the mounting threaded hole (removed anodization at hole bottom + brass 1/4” - 20 screw with dog) and a center tie point on tail PCB. There are losses through the wires (although 12 gauge ~ .090” dia.). Tested light output versus shunt by-pass (by twisting body tube to contact driver PCB) via ceiling bounce. Very comparable (as the eye can see).
Also changed the tail and driver PCB’s screws for #4-40 Beryllium Copper. To align the body tube flat with the switch/logo I added a circular shim between the driver PCB and the head shelf. This is not seen when changing batteries, all very clean look inside as well as out. Used flat head screws to clear the center ring. They should make very little contact with the PCB, it’s not required to conduct to the body. Yet, my setup does give a pathway as I’m reversing the circuit in shunt mode. Can’t figure that one out. An internal short would have direct conductivity and have no switch modes (n channel driver).
Finished light estimated 6500 lumens. From data XP-L V6 (BLF post #39 - djozz) would give 1625 lm driven at 4 1/4 amps per emitter.
Quite happy as is. A picture with Jr. (TK-15)
Many thanks, djozz. After my first posting, ’ I suspect my Q8 doesn’t output all 5000 lumens ’ ImA4Wheelr (post #39) suggested using an external shunt. This got me thinking about my earlier builds of some power supplies. After some temporal jumps, I found some industrial shunts and my archived meters. Tested and tweaked the setup against some 800/1500 lights (tungsten filament - some induction but mostly resistive). Results were encouraging and then worked some secondary issues as electrical pathways.
I made due with what I had at hand, and that resembles more of Edison’s warehouses, post-WWII. ‘Old school’ taken as a very fitting compliment!