750+ Woaw Woaaaaw Waaaaaarg g g!
Good job The Miller!
I hope the very noticeable interest in this project and the SRK style flashlight type will help motivate a better then decent manufacturing process and put it back into the glorious position it used to sit.
Yep, totally agree. It appears to be a tight beam for a flood light, but overall much better looking pattern than the SRK clone. I have to do some comparisons to better make a judgment on the width of the beam - will know better tonight. I though a video would show the differences much better than still shots.
I'm in the process of writing up a detailed post now on the prototype.
I like the SRK beam better myself, it appears to be a tighter hotspot. The rose peddle design is something that doesn’t bother me when actually using it outside.
More tech details on the ThorFire BLF Q8 prototype
After taking some pictures and measurements, it was time to tear her down and found some interesting things. We are cautioning everyone not to rush to judgment. This is the first prototype and has several issues that will be addressed.
Has a real nice one piece DTP copper MCPCB, maybe a little on the thin side at 1.24mm thick, but nice 35.8mm diameter
the LED shelf is a solid 7.05mm thick and integrated with the head, but has an oversized center hole of 15.5mm
outer dimensions are dead-on with standard SRK's at 59.0mm head, 50.0 mm tube, 52.0mm at tailcap, 135.1mm length
LED wires are just two and are heavy gauge of 18 AWG - this is as we requested, and probably a good part of why it performs so well
driver was a surprise - they used our design, parts list, and firmware as a model to clone. So, they used a different MCU and FET. but still a FET+1 7135. This is not acceptable, and explains the issues I noticed in the operation of ramping, etc. We are working with them again to review our driver proposal.
it has an operational Indicator LED under the switch cover - we did not expect this and not sure if it will stay in the production run (but I do like it!)
So we had some good news, some bad news. There's been some re-org with the engineering team at ThorFire, so think we got a good foundation to continue with, and hoping the electronics is not as issue (may have been cost and availability of parts, not sure).
So we decided to swap out the ThorFire driver and install one of our own OSHPark drivers to better match up with the eventual production run. here' some pics:
Stock proto ThorFire driver, a bit messy:
Cleaned up:
The LED's/MCPCB:
Reflector base, showing the cutout for the wiring. Alignment pieces appear to be glued in and beat up - again, we are asking to get this fixed up because the LED's aren't properly being centered - alignment pieces are a little over sized:
Nice brass retaining ring for the driver:
The top stack: Bezel, glass and o-ring:
Tail PCB w/springs. PCB has wide traces on both sides which is good for lowering resistance, but those springs gotta go - hopefully better quality double springs will be used:
Just to show, the one cell is held in place well, so no problem running the light on 1 cell, up to all 4:
Showing the real copper of the MCPCB:
18 AWG wires:
Shelf the MCPCB sits on, all cleaned up (anodized, oversized hole that can be much smaller):
Closeup on the reflector:
Brass ring removed:
Cleaned up, and reflowed to the OSHPark driver using high temp hot air station:
The "real" BLF Q8 driver goin in :
I left in the LED's the prototype came with, though they are CW XPL's - should be NW 3D's. I'll be swapping in V6 3D's next, after this.
Good news after the driver swap, lumens at 30 secs bumped up from 5,320 to 5,624 - 5.7% gain
Could be within a margin of error, but made sure I used the same 30Q cells slightly lower @4.19V.
Also I found out the body threads match to other SRK clones I got (2 I tried), so I could swap in the spring bypassed battery tube base, and measured on the same slightly depleted cells: 6,018 lumens @30 secs, so another 7% bump, even with the cells at 4.15V.
I also measured parasitic drain on this new setup:
- with Indicator switch LED on: 0.22 mA, so 1 Ah drain in 6 months
- with indicator switch LED off (in lockout for example): 0.037 mA, so 1 Ah drain in about 3 years
Basically, even with the indicator LED on 24/7, it would take 6 years to fully drain a set of 3000 mAh cells, and the GREEN switch LED is pretty bright. Again though, can't guarantee the final version will have the LED installed.
Yep there is still work to be done and Thorfire is very responsive, so for now we thnk th work can and shall be done.
Woowzers on the 6K lumens, 50% above specification, how is that
Throw was slightly less than the 4X SRK clone, but so was the lumens with the ThorFire proto driver. Should be up now with our OSHPark driver, didn't measure yet though. The LED's are still not center aligned well though, so might be hurting throw.
Interesting, overall I would say it is a success for a first prototype. I would have been surprised if it was better then this honestly.
Are the LED’s the same distance as your SRK clone? Is it possible to swap reflectors?
If they insist on the basic MCPCB and tailcap (although with a new guy to talk to maybe he would get the dual footprint MCPCB idea back on track?), then you could suggest that the entire tailcap PCB be a solid piece of copper on both sides with some vias to share the load. Same cost but more copper = win.
Do you mean they’re using a different Atmel MCU? Or they basically cloned the UI and wrote their own code for a PIC or something else?
Is that shelf hole the size of a 18650? Looks like maybe they drilled it with the same tool used to do the cell holder. Not that big an issue as far as I’m concerned but unless there’s a good reason for it (cost), it’s certainly worth fixing.
The whole thing looks really good already Keep at it guys.
I think it is good that the center hole 15.5mm because it is convenient for the modification, we could insert a piece of copper or aluminum as heat sink.