Q8 modding

Thanks for the video on swapping the leds djozz. You have far more skills than I soldering like you did. I may have to buy new speakers as I could only hear heavy breathing throughout and no cursing. :slight_smile: :beer:

In production.

I did pop off 2 outer springs out of 20. They are a pain to re-solder since the coating seemed to have peeled off on the bottom, and without the coating, nothing to solder to.

I shall work on the entertainment value of my video’s in the future :cowboy_hat_face: , perhaps even editing in some faul words :innocent:

Tom, you may be able to solder the steel springs (plating peeled off) quite easily using a strong active flux.

This is what I use, if I can’t get the solder to wet with this stuff, I give up. But having strong acid etc. you must wash it off afterwards, or at least clean thoroughly using e.g. IPA and a bristle brush (always good practice, even leaving “no-clean” flux on PCBs is all very well, but I like everything squeaky clean).

Definitely not to be used on electronic components, just metal parts:

https://www.fasttech.com/products/1260004

Applied using one of these:

https://www.fasttech.com/products/1155702

Usually I junk the steel springs and use these “carobronze” springs instead, provided they fit (they are quite small). They perform very well, and don’t usually benefit much from a bypass:

https://www.fasttech.com/products/1347100

Or you can use Lötöl-ST from Felder. It is spcially made to connect solder with steel. Works perfect! But be sure to clean the joints after you finnished. This stuff will corrode your joints it it stayes on.

PS: For newbie modders, here a couple of essentials for getting started on SMD parts:

Fine solder: https://www.fasttech.com/products/2529000

Rosin flux paste (apply with jewellers screwdriver, unbent paperclip etc.): https://www.fasttech.com/products/1784101

These are just what I have bought myself and found good, but Fasttech are well stocked for many other such stuff and a good supplier. They also do sample packs of SMD resistors etc.

https://www.fasttech.com/category/2257/soldering-supplies

I have also found this solder paste is good (keep it in the fridge, as with any solder paste): https://www.fasttech.com/products/1261003

Wow, no idea flux like that even existed!! Too bad it's out of stock, but guess I know what to look for. I got a ton of those carobronze springs - yea, djozz tested, many times hard to find a usage because of the size, and they tend to collapse a little.

As I wrote earlier I’m testing my own firmware in the Q8. Didn’t intend to use the switch LED at all but out of curiosity I added some code for it. Different to Narsil I’m using this LED also in lock out. Made this work with sleep mode and watchdog timer and now the switch LED flashes for about 30 ms with an interval of 4 seconds. Power consumption is less than 50 uAh, little compared to the 115 uAh when constantly on in non lock out mode. Just FYI.

What’s wrong with stock springs ?

Can someone write a few words or give me a link comment with the explanation ?

Thanks

As far as I know nothing but high resistance. Up to now.

I think the answer is: Nothing, but has that ever stopped a modder from trying to make it betterer?

I’d like to chime in on working with solder. Honestly I’m quite an amateur and haven’t had much experience but after reading up on the matter I found a few issues that worked for me.

First, use a good solder. There’s tonnes of cheap solder available on ebay/aliexpress, etc but the problem is you don’t know what’s good and what’s not. Even a cheap solder that worked well for someone else that reviewed it a year ago could change or get copied and resold with an inferior substitute. I use Kester 44 rosin core solder (US brand). I bought 2 partially used rolls off of ebay and though they are years expired they work perfectly well.

I also use Kester flux. Acid flux cleans the components for the solder to stick to. It was so frustrating that when I heated up the components the solder would only pool on my iron tip and not stick to either the board or the component. (Lots of swearing involved). It may be a bit redundant since rosin core essentially means it has flux in the middle but I use it just in case. Make sure to clean it afterward since it does continue to work eating away at the metal I hear.

Someone else mentioned to clean off the existing solder, and that’s not a bad idea either. Their solder paste might be a good one time use with a hot air station but when you’re resoldering parts back on, it may not work as well as you’d hope. I always thought I was doing it wrong or it’s just me but in this case it is the tools you use. A good clean tip on the soldering iron helps too. I ended up replacing my worn out iron with a cheap chinese no name brand. I only spent $8 on it and bought some additional tips just in case (which I haven’t needed yet). A narrow tip of course helps since the components are so small. Good luck.

OK, here's the deal - where I'm at on this spring issue. I fully expected production Q8's to measure in my lightbox at no less than 5,300, averaging about 5,500 based on the round 3 prototypes. I asked for BG to post the 5000+ lumens claim on their listing page to play it safe, low ball number. As it turned out, they range from 4,700 to 5,400 or thereabouts. So what happened? Springs were changed - thought they'd be improved, plus glue/loctite/sand remnants where they shouldn't be, poor screws, poor driver mount because screws were changed from M2.5 to M3, etc.

So out of all this, what the real cause? Well, I think it's a sort of combo. Because of working on so many bypass's over the years, you can telll the difference in quality of springs - some chip off the coating rather easily, some not so easy - there's a great deal of differences. Many times in the past I replaced springs simply so I can get solder to stick to them when applying bypasses, or keep the coating from peeling off. The IOS/MtnE springs have always been good ones, and CNQ or KD also not so bad, even FastTech have been ok. Generally the springs bought separately are superior then what I'd see in many budget lights.

But without testing the stock springs for resistance, I can' be certain but getting a solid 1,000 to 1,200 lumens boost from 5,300 is about 20%, so that's a pretty significant bump for a bypass - I'm thinking these springs are fairly high resistance, and the inner springs did next to nothing - I know they peel the very thin coating easily. I really thought double spring would be a lot closer to results you get with bypasses, as I thought we seen before with other dual spring, FET based lights.

I just chime in to note that a high resistance spring is not neccessarily a bad thing. Apart from that they act as a safety feature (they collapse and break electrical contact when a short occurs), they also keep the led current in a more efficient region. With the stock springs in place (and if needed after cleaning some stuff), the current through the XP-L leds is probably about 4A per led. If you look at the output graph of the XP-L that I made a few years ago (I think that V6 bins are used in the Q8), 4A is still a fairly efficient region, good for runtime, good for keeping temperature in control for a while.

With springs bypassed you get 1000 lumen extra, with the leds at 5.5A, so 30% more power (heat!) for 20% more light which is perceived as 10%. (all numbers are estimates).

I think for people who plan to often use the Q8 as a work/camp light the stock springs without bypass are perhaps preferrable. And the rest of us who simply want as much light as possible, bypassing the springs is the easy mod that gives the single best bump in output possible.

Oops, it was showing in-stock when I posted the link. Someone must have ordered the last bottle. “Discontinued means it’s gone now (from Fasttech at least) Maybe still available elsewhere.

Perhaps alternatives are available at sensible price ?

Key ingredients are ammonium chloride (sal-ammoniac) and hydrochloric acid,

Be interesting to see how these springs perform in a Q8: http://www.mtnelectronics.com - nickle coated

I just got a few in - they are little shorter but much stiffer. Dunno how conductive they are but I'm thinking they might survive the jolt test. I tested the stock Q8's - if you tap them hard enough on a counter or wood bench on the tailcap, they do shut off, so the beefing up of the springs they were supposed to do maybe improved it a little, but didn't solve the "jolt test" issue.

Agreed, I see the Q8, as delivered as a nicely balanced thing, with pursuit of further current being into diminishing returns (efficiency does matter to me) I’m actually considering piggybacking on another 7135 so I can mostly run it in linear mode at light levels I mostly use, but would that badly unbalance the smooth ramping I wonder ?

Yes and no.
One low Vf Led powered by one high-rate cell is not very efficient combination with FET driver. One cell to three or four leds is much better :smiley: .
But we shouldnt forget that any extra resistance means extra heat. P=I^2*R=25*0.05=1.25W per each spring (or more). I dont know how will it (spring) feel after hundred of 20C-80C-20C cycles (some springs stop being spring and become just a piece of wire).

Depends on your definition of “badly” but I guess it’s better to adjust the ramping table.

My SRK clones all have that jolt problem as well, i.e. shut off with a smart slap on the tailcap. That’s the problem with not having springs at both ends. I hoped the Q8 would be immune, maybe it’s just a matter of enough decoupling to the MCU so it can sail through a few milliseconds outage ? Not possible on my clones because the MCU runs from the same rail as the LEDs, so no amount of decoupling would hold it up, but is the MCU on Q8 not behind a reverse-protection diode, so maybe stacking on a big tant. decoupler could help ?