Fireflies E07 preview

you see the mcpcb did just touch the body on a small circle (1mm)

looks like the floating mcpcbs from the sky ray clones ;)

mine had leftovers from pushing the holes and shape of the mcpcb too which also prevebted the copper from touching the head

Ahh, ok Thanks! I'm starting to get it now. Sorry, takes me a while...

All this stuff is so much more critical on high powered lights. I've seen this many times before so when I have a light apart, I just make it a habit of checking the MCPCB and shelf flatness. For an MCPCB, sanding the bottom on a flat surface is the ultimate test - it can tell you a lot about the surface. For the shelf, sometimes you can see the problems, sometimes it helps getting a straight edge on it, and get one as close to the width of the shelf as possible. P60 pill surfaces are notorious for being unlevel, but can be leveled with a flat ended punch easy enough - few taps.

Seems like FF is a newbie at all this. Always check your vendors for quality, insist on them meeting detailed specs and do their own quality checks to ensure all spec'd dims are met or exceeded, penalize them if not done so. Might be difficult with typical sourcing over there though.

any suggestions on how to protect the LEDs on the mcpcb while fixing it?

they shouldn’t need any special protection.

After you pull the star out, run a handfile around the edges to chamfer the lower edge of the star. Keep the file perpendicular to the star so there is no chance of it slipping and hitting the LEDs. Should take no more than 10 minutes with a small file to make a 1mm chamfer all around the lower edge. A quick few swipes with a file over the bottom the star with a file should be enough to get any burs at the bottom of the holes.

Course, it might be harder if you plan to run a razor blade over the bottom like a previous poster did. Safest way to protect the LEDs if you plan to do that is to remove them all, fix the star, then reflow them all afterwards. I think that’s overkill though.

So the full diameter at the top is fine? Just take an angle off the bottom edge?

Yup. Upper diameter is fine. Just chamfer the lower edge slightly, flatten the bottom a little and add a thicker layer of thermal paste.

My E07 heats up pretty quickly. Probably no thermal paste like Martins. I mean I don’t care if you’re an amateur building your first P60 pill, this is basic stuff. I have 4 ROT66 with no issues, 1 PL47 with broken AUX LEDs, and now this E07. I’m not even looking at his stuff any longer; I give him credit for customer service I will say that much but at some point you have to do your job better, not worse.

7 low voltage LED's, heavy teflon coated short LED wires, decent FET, decent springs, the LED+ wire goes directly thru the driver spring, combined with a decent 21700 cell -- this light pulls more amps than the way bigger Q8, so it will warm up fast even with the poor contact the MCPCB has.

I just got mine fully apart, except for removal of the switch. I used 600 GRIT paper and sanded smooth the MCPCB, then got it mirror smooth with 2000 GRIT, so sanding on a flat surface, I'm pretty confident it's flat now. I used an Exacto knife to trim the big wire hole edge - don't want sharp edges there to fray wire coverings. The switch wires are long enough to tilt the driver out, but not long enough to clip/re-program the MCU - sooo close though. Might fit with flexible silicone coated LED wires, not sure. The driver was mounted with the glue that's like rubberbands - flexible, stretchy. I poked it out with a solder pick tool through the wire hole - broke the bond fairly easy, no damage but I was careful, no hammering needed. I don't see any problem with the width of the MCPCB in mine.

Considering how super thick the MCPCB shelf is, the head is pretty light weight. It's nice to see a light built this way, specifically designed for low profile TIR optics but the style is more conventional with the elongated head so it can be packed with aluminum, all in one piece with the outer shell - I think that aspect is quite unique, and for us, really nice! They probably could have reduced the length by 5+ mm or so and still had a decent thick shelf. I think the Emisar's went overboard with priority given to as min length as possible. Their pill space is super tight and not much metal in the heads, while the E07 least has some limited space in there, plus lots of metal. It's all a tradeoff of course.

If I can make a suggestion….

I like military watches (doesn’t everybody?). Everybody and his dog makes a model but what very few do is to raise the bezel just enough to avoid contacting the glass. Just a few millimeters.

If I was designing it, I would explore how much I could raise the body or sink the switch to make it less liable to self turn on if it hit something. Not much, just a few millimeters. Maybe 5. Of course there is no guarantee that it will never happen, it just reduces the probability by a a substantial margin. Plus the rubber on the switch is probably the second most vulnerable thing.

Just a thought……

There's been lots of discussions on the switch height and protection on just about every light with a side switch here. Not sure, but thought ThorFire did set it a bit lower, but of course, not low enough. There were issues like this we just couldn't get them to budge on for various reasons, so we eventually settle, like we did with the screws securing down the driver - really wanted a retaining ring. Also wanted a better FET, didn't get that either. It's all about what they can do or can't, pricing, priorities, adding further delays, etc.

I’m not sure why they would resist. It’s a matter of machining and that is about the easiest thing to change. Literally ZERO cost. They have to mill a hole, they can very easily mill a countersink. Just programming.

Perhaps I can help on the next build with machining and manufacturing knowledge.

I can't recall, others might know better. Not sure if we gave it priority, and might have been a slippage issue - when you are running say 6 months behind, and they tell you it would delay another month of this or that, you tend to back off. Of course it doesn't really happen till another 10 months, but of course we didn't know that at the time -- hindsight thing...

If it was a specific resistance to countersinking it then it had to be the switch itself. Either they had stock, wanted to maintain commonality to their other products or simple unavailability. As is, the switch may not be thin enough or the threading wouldn’t close to hold the switch if there was 5mm less of material thickness.

On the other hand, they could accommodate that if they wouldn’t remove off as much from the top, where the fins are.

I’m looking at the Thorfire Q8 and they could of left the ridge/fins on there and countersunk the switch. As it is now, it’s about 3 mm proud of the machined plane. It looks good, but I always chose function over sexy. On the other hand, I’m not selling flashlights.

Anywho….next time.

Black paint removed from h2r nova clip to use on E07. Much better! Ordered a silver Olight S1R/S2R 2 way clip but it’s much shorter, so I did some pocket clip swapping around today. Zebralight clip went to the h2r, new Olight clip went on my old zebra h600w, and the h2r clip was sanded and put on the E07. Still have some polishing and clear coating but it looks good and much better than original clip :+1:

The clear ano eo7 looks very nice!

I received my clear E07 today and one of the main LED’s wasn’t working. I already planned on taking it apart for a Luxeon V2 swap so I wasn’t concerned and got to work.

This is what I found:
D7 was the one not working. Absolutely zero solder on the cathode pad.

And the bottom of D7 is dry as well

But I’m happy to report the L1V2 emitter swap went perfect and it’s all buttoned back up!

Light box readings (raw numbers):
Stock SST20 4000K = 35100
Luxeon V2 5000K = 47200
A 34% gain (but that’s to be expected with it only having been running 6 of the sst20’s)

Guess this one missed QC huh…

And don’t forget the SST-20 4000k and below only come in 95CRI, so it is a bit unfair.

How did it not reflow though? Don’t they use skillets to specifically prevent this?

I’m guessing it was a paste applicator error. I can see maybe the center pad not flowing with how much mass this MCPCB has but the cathode and anode pads flow easily even on this big honking thing. I just can’t imagine it not flowing when the 20 other solder joints, including all 7 thermal pads were good unless it simply didn’t get paste.

Either way literally 1 second of QC should of stopped it from going out the door like that…

No issue with my, love it!

What made you choose the luxeon v2 over all the other swap options available? And what method did you use to reflow?