Haikelite HK04 groupbuy is ALIVE

Commercial Sellers' Spot BLF User Negotiated Deals
81

No, I can't yet. On Q8's to measure current, I had to mod the rear spring PCB to force the electric flow through one path, then added a heavy gauge wire loop on that path so I could use a amp clamp meter. I don't know how others do amp measurements on these 3 or 4 cell parallel setups, but another thing you can do is test current on just one cell - that's easy to do but would probably result in lower #'s than you get with 3 cells, so not realistic. Another option may be to remove the spring PCB from the housing, and with the battery housing removed, hand hold the PCB in place, using a thick gauge wire for the batt- return, then a clamp meter on that -- I never tried this method though - you might need a couple extra hands for that.

For the 35E's, I'm not surprised at all. I've gotten the same results many times with different lights. The 35E is an outstanding cell, frequently under rated for it's ability to do relatively high amps.

Come to think of it, what I've done more recently is use a single cell - the Samsung 30T, measure the lumens with it, and measure the amps. This way I have some legit #'s to compare, and might be close to the results to 3 good 18650 cells, least for the initial turn on.

82

Has anyone ordered and received the xpl hi 5000 lumen version? Can comment on the beam profile width, amount of spill and throw distance?

I ordered hxp 50.2 but they sent me a wrong version ( with sst-40). It is 9k lumens with good batteries , the throw distance though is less than 850m. I still need to measure the candela but I can see the hot spot is far less bringht than my haikelite mt40’s which throws to 850meters. Also not crazy about the super wide overall beam - blinds too many people on the side.
Wonder how far the xpl-hi version throws.

83

I rigged up a method to measure current on this light. I ordered 4 of the real deal SONY VTC5D's from Walmart (believe it or not), and taped them up as below. I got a 14 AWG wire there and measured with the clamp meter -- it hit 37.1 amps at the peak.

Testing on the fully charged, new VTC5D's:

  • 14,620 @start, 12,920 @30 secs (my calib), or 12,900 @start, 11,400 @30 secs (Mauka calib)
  • 79 kcd measured at 5 meters

After taking a close look at the springs, they appear very much like blue springs, BeCu's.

The 37.1 amps was taken with the setup shown below, so couldn't really get a decent compression on the springs - the tape holds things in place so they don't go flying around, but doesn't apply any pressure. I'm quite impressed now with the performance and overall quality of this one. They even used brass screws (which I prefer and modded on Q8's), and the tail PCB is a solid plane of copper under the green, so it maxes the conduction. I scraped off some of the green to solder the wire in place on the top.

Probably could get a bit more out of it with spring bypasses, but might not be much of a gain - the springs look and feel like true BeCu's.

The lumens and throw were taken with it full assembled. After the testing of about 60 secs on max, I used those same cells, partially depleted, to measure the amps. So given that, probably would measure higher amps with a better setup and fully charged cells.

84

Thank you so much for your research. I read it with interest and received a lot of useful information.

85

I was curious myself. I may re-do the amp test on fully charged cells and after spring bypasses. With the XHP50.2's, it's a compact monster for sure. Comparatively, I hit about 30 amps on a single GOLISI 26650 in my modded triple XHP50.2 3V Amutorch X9.

For the springs, Martin specs it out in post #2, but I certainly can confirm. I just did a little research on 6063 alum a bit. Most lights are 6061 T6, while this one claims 6063. 6063 is very close to 6061 in terms of hardness, price, thermal characteristics. For a flashlight, I would be most interested in surface hardness, how well it takes anodizing, and of course thermal/electrical conductance. For China sourcing, these variants are just way too close to figure out which is better. I think we would see variety in quality on just 6061 sourcing, so probably the sourcing of the material in regards to tempering (T6, T5, etc. ratings) and how well and precise the alloying elements are controlled, are more important. Of course we all should not assume the specs are accurate.

Lights like the ROT66 claim they are 7075 alum, which has a superior hardness, but lower thermal and electrical conductivity, more corrosive prone, and at a higher cost.

Update:

I added 20 AWG spring bypasses, and on fully charged VTC5D's I got an initial reading of 15,810 (my calib), 13,980 (maukka calib). It dropped quickly - apparently I didn't set the temp regulation limit high enough. So it got about an 8.1% boost with the bypasses. But additionally to the bypasses, I did some modding at the contact at the screw holes of the PCB to the battery housing. I sanded off the anodizing around the screw holes of the housing, added some solder on the PCB side, all to improve contact surface area. The screws and the immediate exposed copper tracing around the screw holes are the only contacts passing the current from the batt - to the housing.

Also just fyi, the VTC5D's I got are from here: https://www.walmart.com/ip/Sony-Murata-IMR-18650-VTC5D-2800mAh-25A-Rechargeable-High-Drain-Flat-Top-Battery-4-Pack/675299181. They are hard to find, and in stock, and at a decent price. It's a good deal for this price, I might just order another 4. The only other VTC5D's I had were the black labeled Vapcell's, but the raw SONY's seem to have the batt+ contact better exposed, lacking the extra wrapper. They are flat tops but the type that protrudes, similar to SANYO GA's, and seem to work fine as-is in the HK04.

86

I finally was able to get the driver out - used the SP36 technique of soldering a thick wire to the batt+ ground ring, and leveraged some force to pull it out. So no damage, and the glue was the soft silicone type. The switch PCB is an entirely nother problem - tough glued, really bad. The driver design is somewhat unique for a ATtiny85 design. It uses parallel higher powered FET's for the high amp channel, and what looks like a low power FET for the low power channel, usually a single 7135.

Wish I knew what Anduril config files they used, or if they made/modded their own, they should be accessible somewhere in accordance with the public license. It's disgracefuly really... Their site is still down.

I may just try a FET+1 Anduril config file and try it - I can crank up the temp max which this light needs.

87

Please tell me, is the stainless steel bezel pressing glass glued or can it be loosened freely?

88

I didn’t try to remove the bezel so as not to scratch but it seems to me it can be removed as it has the grooves on the inside to use the special took to turn to unscrew it.

89

Haikelite and Imalent are under built and over priced in my opinion. I bought the DN 35 and 70 when they first came out and gave them both away. I could not find the switches in the dark to save my life.

90

Haven't tried yet. I need to get back to this light.

91

Oh, sorry, just got to this. I kind of messed up my HK04 a little:

  • Nasty scratch by the switch
  • in trying to get the switch PCB out, broke it, damaging one of the 2 LED's. It's glued badly - worse I've seen on a switch

So got frustrated with it and set it aside. But yesterday, I got back to it and programmed my latest Anduril in it, using it as is.

The bezel unscrews easily by hand, least for me - didn't need a tool. The stack is:

  • SS bezel
  • flat hard plastic o-ring
  • glass
  • soft o-ring
  • reflector

92

Thanks you!

93

Can I have the code for the HK04 xhp 50.2, thank you.

94

its not on stock

95

Has anyone received these lights yet? I’m looking to order one from Fasttech and can’t seem to find any reviews about it.

I was going to attempt this group buy but the price is better from FT….

Maybe I’ll review it when I receive it?

I’m most interested to see the possibility of removing some material to fit 21700’s and also reflowing warmer led’s (Specifically XHP-35Hi’s).

96

I’ve ordered a while ago from fasttech (xhp50 sand) and I’m quite happy with it. It is awesome powerhouse and beautiful light. Only thing that bothers me little bit is color temp, I like 5000K and lower. For xhp35 you would must change the driver because it’s 12V but I can imagine that with 4x xhp35 4000K, the beam would be just awesome. If you can do it, please share the results :wink:

97

Awesome to hear, and I think a lot of us feel that way! I’m most happy at 4000k.

I actually didn’t realise this doesn’t come in a 35Hi version.

Maybe they are already using 12v xhp50.2 or they are wired 2S2P. Would you know what type of driver? I’m gonna be looking towards mtnelectronics for a replacement, maybe SRK driver will fit?

98

It's a FET based driver. I believe it's two channels - a small FET (plus some circuitry) and a double large FET. Thought I posted pics, maybe not yet. I'm still working on modding mine - it's a pretty decent quality small SRK style light (3 batts), stock with the quad XHP50.2 3V's - so yes, its for super high amp 3V LED's with the MCPCB designed for XM footprint LED's (XP-L2, SST-40, XHP50.2, etc.).

Here's some detail shots of the driver. I don't know who designed it. It's similar to a high amp TA driver but instead of using a bank of resistors for the low channel, a small FET is used. Of course on this driver is the support for USB charging as well. The MCU is our beloved ATTiny85 to run Anduril (or NarsilM). The two big FET's (Q1 and Q2) are in parallel, and the Q7 I suspect is a small FET with at least 1 resistor connected before it.

I busted my switch PCB and that's what I'm stuck on - trying to replace it. They used a ridiculous amount of nasty glue on the switch PCB. Like to keep the USB charging LED indicators but I don't have a dual LED control switch PCB -- wish someone had one. Might have to simply repair it but need one of those dual LED parts.

I'd also like to re-flash Anduril but there's no info whatsoever about the exact Anduril config they used. Could experiment -- the thing I mostly need to know is how the config the ramping table for the low channel.

99

I ordered from fastech and they messed up my order. Sent me the 9000 lumen version instead of 5000 lumen. I then ordered the 5000 thru this thread’s link from banggood and they sent me the correct one. I then ordered the 13000 lumen version from here also and after waiting for 2 months banggood has no idea where it is, no tracking number shown. Anyway between the 5000 and 9000 I’d definitely go for 5000. The lumen level i measured and it’s about right for both but the 5000 version throws farther,( although I’d like the hot spot to be larger, but it throws slightly further than sofirn q8 ), lasts longer of course, and visibly the difference in luminance is not that evident believe it or not . The only thing I don’t like is how wide the overall angle is, I hate blinding everyone on the side.

100

Thank you! The 3V setup makes much more sense since they can use the same driver, you’ve shed light on that for me. It also makes sense then why 5000k is the lowest color temperature they can offer.

You’ve been so helpful already, I don’t mean to bother, but what diameter do you measure that PCB to be?

Also any chance of cramming some 21700’s in here? I would be willing to mill out some material (if it’s there…that is.)

It’s over my head but i’d love a battery bank function in this light, does that seem possible to add? I’d most likely have to disassemble a light with that circuitry already to learn about and borrow components.

Good luck with the switch replacement, and it seems tracking down the firmware can be quite difficult.