@sharpie and emarkd: We have discussed the fusable trace idea but stopped where we could not decide on how narrow and long they should be, what effect on the performance it would give (added resistance/lowered current in the flashlight), and if it would work at all. Could one of you work out what the shape and dimensions of useful fusable traces in the tailboard should be (4 in total, one per cell, blowing at say 20A), and (because I’m lazy enough to ask instead of calculate it myself) what the added resistance would be?
djozz and TM: I’m not experienced enough to claim to be able to design the thing; I was just pointing out that a proper fusible link doesn’t have to affect performance in a meaningful way to be effective. Getting Thorfire to build it to spec is another issue altogether and I’m not close enough to the project to comment on that. If that’s a concern then I can see why you may want to avoid it.
Ok, well, so are we 
I hope that Sharpie has some idea.
@Sharpy, those are numbers that we can ask Thorfire, but I would not expect that they use different stuff than other SRK-manufacturers, so we can measure it ourselves too. Or is there a ‘standard thickness’ known to us all?
1 great to see its basic (thus) easy for you
2 thanks
3 (the edit) thanks, a bit too technial for me, glad for your point 2 
4 it is all new so lets go with the oshpark specs since that would be the place for us to get prototypes. so:
1oz copper on both sides (1.4mil, 35um)
It would be fun to have a test-board made by Oshpark (assuming they use the same copper-alloy and tickness as Thorfire will) with on it several designs of fusable traces. I can not make the boards but I’d be happy to test them (voltage drop and blow-current, my powersupply goes up to 20A, should just be enough )
I’m not a modder, but I do have an Electrical Engineering Tech degree. Designing circuits with a safe fail point is part of the job. As Sharpie said, the other SRK clones have that safety built in by accident or (possibly) on purpose. Having a trace that is designed to fuse at a known high current level is the responsible thing to do. It is NOT a matter of proving that it has not happened and is therefore not needed (or vice versa). I despise the politically-correct nanny-state as much as the next guy, but there really is a liability issue. It is impossible to anticipate EVERY possible failure point. In hindsight, both of the recent designs that failed violated some pretty obvious good practices. This is proof that sticking to good practices, such as including a fuse trace, is important. This is not “stashing the trunk full of spares”.
What other methods are there in this design of limiting damage? (I won’t say limiting current, because around here most people around here take that to mean “limiting lumens”.)
To be clear, I’m not, and I don’t think Sharpie is either, suggesting that the trace be calculated to fuse at five amps or some such silly low number. Furthermore, including an easy way to bypass the link SHOULD be included. This ‘mod’ would be as simple as soldering a wire between a couple vias, or even solder bridging a couple of pads. Someone (DavidEF?) said that this is to be a “Mod-Friendly, not Mod-REQUIRED” design. The design point of the fuse should be such that the only time it is needed to be bypassed is when OTHER mods increase the current draw beyond the stock performance. Inevitably, this will start to affect the current and therefore performance near the top end, but hopefully not by much.
Yes, you CAN replace a fuse with a coin, or a chunk of 1/4” steel rod. But doing so means you run the risk of the failure happening somewhere else with unknown consequences.
[edit] figures, you guys had a whole conversation while I was typing this up. fooey. [/edit]
The use of fuses all over the world, from low power electronics to gigawatt level transport infrastructure suggests it is not so. It’s the last line of defense to prevent other things from going up in smoke. If it’s going to be in my hands, on a shelf in my room or in my backpack, I’d very much rather have the fuse than a few extra lumens.
Besides, it should be easy enough to disable with a blob of solder if that is planned for. It’s far easier than a spring bypass.
As an electrician, I know exactly where you’re coming from with this statement. But, let’s not forget that BEST and CHEAPEST are not synonymous terms. A lot of times, manufacturers do the cheapest thing they can that serves the purpose for them. They are not at all concerned for performance. All they want is to sell the thing. Also, in the case of home electrical wiring, controlling the voltage and current is all the engineers, manufacturers, and builders can do, because they don’t really know what will be connected to that power source. They don’t get to control anything but the current, so that’s where they design their safety mechanism.
If a fusible link that doesn’t significantly hinder “stock” level performance can be designed by Sharpie or others, I’m all for it. I understand that if one is modding already, then it should be no big deal to mod over the fusible links “while they’re in there.” 
Hey miller! I do not know what it is called and yous have probably already answered this but the PCB where the batteries meet the head are you going to have any thing that makes contact on that side? I do not have a SRK clone but i have a solarstorm T4 and it is similiar and i cant use flat top cells in the light very well some work and some don’t. I have seen these SRK clones with bumps on the PCB like copper buttons that make contact. That is where i have seen it. the Palight boss has this design!
I will try and find a picture of what i am talking about!
Mine looks like this!
!
!
The back here…
For the tailswitch thing I handle it in the way that if a light has a lockout through a twist(like the m6) I don’t need a switch. But if a light has unanodized threads I add a switch(like the courui)
Just for the protocol, i dont See a need of a extra fuse trace…I want maximum performance… :smiling_imp:
I also don’t see a risk of shorting the batteries on the driverside.
But of course it would be interesting to see a back on the envelope calculation of the currently used traces and the corresponding resistance and when they will blow…
Lazy-R-us you are making it clear that I have to think about this different, thank you.
And apologies Sharpie for failing to see your valid points.
I now get it because of how Lazy-R-us worded it.
I am once again very happy with how we discus the ins and outs of the Q8 with BLF and grateful for the input it generates.
Thank you Sharpie for looking in to a fusible trace design to make the Q8 have the basic safety it needs.
@everydaysurvivalgear, yes it is for button tops either flat or a ring (I think ring adds costs so probably flat) no brass things sticking out like the Boss1 has.
You’ll need buttons mate
I would like a fuse if it doesn’t affect performance in a very significant way and can be added to the design easily.
I once shorted a single 18650 cheap chinese light with a 30Q battery in it. Multiple components burned out instantly. Dang, there’s a lot of energy in one of those things! Let alone 4…
However if you fuse the driver, you can still get a short cirquit like the cometa experienced. That shorted before the power even reached the driver. Only protected cells would help against that, like someone pointed out here before.
A fuse would make it a lot safer for most failures offcourse. (e.g. self unsoldering LED wires, drops leading to a short somewere, water intake, metal flake residue from the manufacturing process leading to a short etc.)
Agreed, but this being Budget Light Forum, I doubt we can get the “best at any cost” in this light. You can’t get much cheaper than a PCB fuse :]
My dad’s an electrician too so I’ve heard a few tales of objects dropped in the wrong place being vaporized.
I personally don’t think it needs a tail lock out switch. If I’m gonna store it for a long time, I’m just going to take the batteries out of it. It seems like extra stuff that isn’t needed. I also don’t think we need a fusible link. If anyone is concerned they could use protected batteries. I’m in the “keep it simple” camp.
No tail switch
Thijsco is emailed to the posts about clearance and surely can check the CAD drawings how current tube and tailcap compare.
But it will be a modding effort to install one.
If fusible traces can indeed be done as simple as Sharpie says and does not cause current limitations for operation of the Q8, yet a failsafe that does not cost anything more in production I am all for it.
Just waiting on the real input on this from Sharpe, since it is way out of my league to design such traces. Glad that he spoke so confident about it and only needed the before provided info on copper thickness used.
The calculations design of such traces is way above me
Cool, thanks!
+1
For maximum performance, this Monster must comply with KISS principle.
A fuse may result in compliance with Murphy’s Law.
I like the idea of a fusible link with an optional bypass for those who must have every mA. As was mentioned acell like a 30Q contains a whole lot of energy. I have seen a 30Q melt a power pack spring with ease. I guess that as the packs fusible link, now that I think of it. Not that it was designed with that in mind.
I think its would be nice to be able add a forward clicky for tactical, momentary blasts, or signalling purposes. But it would require driver changes (like 30 clicks puts it into tactical mode) or another driver (future proofing).
But after reading your reasons for not having one, I’m can live without it. Maybe a tactical momentary mode using the momentary switch can be added to Narsil?
I do like Millers PCB drawings making it possible to change to 2S2P by cutting traces. This adds to future proofing and modding potential.