Q8, PMS SEND TO THOSE WITH ISSUES BLF soda can light

[Texas_Ace]

Great tests Tom on the cells, they basically mirror my results although I didn’t test as many types of cells. I suspect that the final XP-L will pull a bit more, the V2 HI’s I am using don’t generally pull quite as much as the V6 HD’s from what I have noticed. I would still be surprised if it pulls over 20 amps though as even with my 8x SRK builds I have trouble making it over ~25-26 amps.

Now if you used 219c’s or XP-G3’s (which I have considered as a high CRI light), then you would have some SERIOUS amperage.

Hate to quote myself but I wanted to see what people thought of using pre-made fuses instead of trying to rig something up with traces? It seems like the better option to me and the results would be easily known as it lists the resistance in the specs.

@Texas_Ace

If we are sure that a fuse is needed (which I do think is a good idea). How about actually using a fuse?

I started thinking and realized they must sell SMD fuses. I found a wide selection such as this:

0685H9150-01 Bel Fuse Inc. | Circuit Protection | DigiKey

They come in all shapes and sizes and amp ratings. I am sure from china in bulk the price would come down to a reasonable price for this situation.

This seems like a much better option then guessing with traces. Plus if they are properly placed then you can simply remove a fuse to set up the tail PCB for 2s or 4s operation.

If traces are still considered the best option I did run across the paper that is supposed to talk about this subject but I have not read it yet: http://www.ultracad.com/articles/fusingr.pdf

[schlockproof]

I just enlisted to enlist. Please count me in for the Q8!

[The_Miller]

Real fuses have come up bit we’re dismissed due to the extra complicated tail PCB, sockets and fuses.
Adding costs because of parts and manufacturing stages.
If the M6 and Meteor can do without I see no reason why we should do it

[Texas_Ace]

@The Miller

Real fuses have come up bit we’re dismissed due to the extra complicated tail PCB, sockets and fuses.
Adding costs because of parts and manufacturing stages.
If the M6 and Meteor can do without I see no reason why we should do it

Well in this case it would just be the fuses and solder them directly to the PCB but you still have the parts cost. I tend to agree they are not needed but then if they are not, is a fuseable trace needed?

I do have an idea on HOW to do it but the math of what to do is the issue. I also worry about the resistance it will add to the circuit if the trace is made small enough to actually “fuse” when overloaded.

If the Q8 uses normal springs like on a SRK with no bypass (I like the SRK springs that go all the way through the PCB and are soldered on the back BTW), then the spring is already a fuse of sorts and I have a feeling it would still be the first thing to go if a battery was to short out.

To make a trace have more resistance then a spring would take a silly small trace and would add resistance to the circuit greater than that of the un-bypassed spring. If it didn’t have more then the spring, well it would not work, the spring would be the one to melt first.

Which is a good point, you would basically be adding another spring to the circuit.

[Tom_E]

@schlockproof

I just enlisted to enlist. Please count me in for the Q8!

Cool! Glad you joined in! I'm liking your BLF ID.

Yes, bout the springs, think we are still planning on double springs, like many of the Manker lights have, that I originally seen done on the BLF X5/X6's. Depends on quality of the spring used, but double springs do (should?) perform better, but makes a less usable fuse, maybe? So, plan is bypass's will be the left to the buyer/user. So actually in stock form, the Q8 will probably fail at the springs I would think in case of a short? Double springs may still be the weak point?

Ooops -cross posted with djozz: I'll defer to him

[djozz]

TA, what do the specs say on the resistance of the SMD fuses? If I remember well, they have been considered but add more resistance than we wished for. If we would go for (SMD/trace) fuses, IMO they only make sense if low resistance (copper alloy) springs are used, because double safety=unneccessary double resistance as well.

[Texas_Ace]

@djozz

TA, what do the specs say on the resistance of the SMD fuses? If I remember well, they have been considered but add more resistance than we wished for. If we would go for (SMD/trace) fuses, IMO they only make sense if low resistance (copper alloy) springs are used, because double safety=unneccessary double resistance as well.

The one I pulled up in a quick search above showed a resistance of 0.0026 Ohm. Which would work out to about a .013v drop in voltage. It can also be replaced with a 0ohm resistor if desired.

If we will be using double springs and/or low resistance springs, then a fuse would be a wiser move as the springs are then far less likely to melt.

That said, if fuses are not an option then I think that a normal thick steel spring that is strong and more worried about keeping contact with the driver would be a better move IMHO.

Each cell by itself should only be producing ~5amps at most, so a steel spring should handle that. Heck I have SRK’s that are running over 25 amps on stock springs and don’t have issues.

I was doing some reading and there were people saying they did use traces as fuses, and that they worked. They also said it was worth the cost to use fuses as once the trace burnt it was hard to repair and thus generally the whole PCB had to be replaced.

[The_Miller]

@schlockproof

I just enlisted to enlist. Please count me in for the Q8!

Welcome to BLF!
Will update OP list later

[fixed_it]

@Texas_Ace

The one I pulled up in a quick search above showed a resistance of 0.0026 Ohm. Which would work out to about a .013v drop in voltage. It can also be replaced with a 0ohm resistor if desired.

That’s about the same as 4 inches of AWG 18 wire. It’s also about the same as the FET IIRC. And way, way less than any spring can do. Also consider that when putting one fuse per cell, you should divide that number by 4 to get the global resistance. It’s not significant IMO.

@Texas_Ace

If we will be using double springs and/or low resistance springs, then a fuse would be a wiser move as the springs are then far less likely to melt.

That said, if fuses are not an option then I think that a normal thick steel spring that is strong and more worried about keeping contact with the driver would be a better move IMHO.

Each cell by itself should only be producing ~5amps at most, so a steel spring should handle that. Heck I have SRK’s that are running over 25 amps on stock springs and don’t have issues.

I was doing some reading and there were people saying they did use traces as fuses, and that they worked. They also said it was worth the cost to use fuses as once the trace burnt it was hard to repair and thus generally the whole PCB had to be replaced.

I’m not even certain we need to work very hard to get a PCB fuse. It may well be that the current layout would already not survive the 50-100A of a cell short. Anyone feel up to ordering a PCB with different trace widths and test at high current? From what I can figure out with an online calculator, on a 2oz PCB, a 2mm wide trace which is 2mm long would have a resistance of about 1 mOhm (insignificant next to the springs) and almost certainly blow up at 50A (it would dissipate 3 watts then).

[Lazy-R-us]

I’m in favor of some sort of fuse. There was concern that the cost of the fuse(s) would be prohibitive, so trace fuses were suggested. I would rather see SMD fuses to allow replacement and bypassing for the extreme current chasers. We have recently had a post where an experienced BLF member reversed one cell in an SRK (can’t remember where I read it). This light will be sold by ThoreFire to the masses, I think it is necessary for us to design in a fair degree of safety. If we use a 7.5A fuse for each cell we are limiting the overall current available to around 30 AMPS, is that REALLY a limitation?

Yes, there is a voltage drop across the fuse, and that will impact the available output. Will anyone be able to tell, in a side by side comparison without instrumentation? I doubt it. For those who are looking for every last lumen, swapping the fuses out can be done while you are putting in new LED’s.

[Texas_Ace]

@Lazy-R-us

I’m in favor of some sort of fuse. There was concern that the cost of the fuse(s) would be prohibitive, so trace fuses were suggested. I would rather see SMD fuses to allow replacement and bypassing for the extreme current chasers. We have recently had a post where an experienced BLF member reversed one cell in an SRK (can’t remember where I read it). This light will be sold by ThoreFire to the masses, I think it is necessary for us to design in a fair degree of safety. If we use a 7.5A fuse for each cell we are limiting the overall current available to around 30 AMPS, is that REALLY a limitation?

Yes, there is a voltage drop across the fuse, and that will impact the available output. Will anyone be able to tell, in a side by side comparison without instrumentation? I doubt it. For those who are looking for every last lumen, swapping the fuses out can be done while you are putting in new LED’s.

This is kind of my thinking, if you do anything to prevent a disaster then you are thereby taking responsibility, now Thorfire being a china company that is not a big issue but if an American company you sadly have to go all the way or say it is 100% the users responsibility.

Fuses are the only way you will get true safety fuses.

I did just skim through that paper on fuse traces and it was interesting. I also found this program: https://www.saturnpcb.com/pcb_toolkit.htm

I am going to mess with it and see what kind of results we would get from a fuseable trace.

[Morpheus]

Sound’s great, how can I get on the list?

[Schku (Schkü)]

Please count me in for one.

[pilotdog68]

@Morpheus

Sound’s great, how can I get on the list?

You have to pay me your $40 to get on the list

.
.
.
.

[The_Miller]

Morpheus you just did (well when I have updated the OP )

Hmm there is a manual stating how the cells must be put in.
Button tops are very easy to see + and -
Would Ford/Chevrolet/etc be responsible if somebody puts the wrong fuel in?
Mountain Electronics sells both Supfire M6 (with versions up to extremely hot rodded) and Meteors to the general public.
Neiher have fuses.
As a matter of fact is there any light in the long list of parallel lights using fuses?

So what I see now is 4 options to pursue
1 do nothing, as all others making parallel lights do.
2 add costs complexity by adding fuses, the Q8 the first flashlight with fuses.
3 lower complexity by choosing the proper spring configuration.
4 add no costs by having a part of the tracing for each cell act as fuse. If somebody blows this it needs repair but it is not like it happens all the time and a little repair seems much better then a possible dangerous situation.
4 needs calculations how thick and wide the tracing should be and as before when Sharpie was sure this could be done this seems to be a calculation beyond us

So option 1, does not feel 100% right yet is the industry way of doing it.

Option 2 seems heavy overkill and just trying to convey this to Thorfire is gonna take a lot of time.

Option 3 with all the posts about how to bypass springs this does not seem all that negative.

Option 4, unless a calculation can be made no use talking about it since it all boils down to this calculation.

[Flashy_Mike]

Fuses - Groundhog Day …

An engraving “protected cells only” is cheaper and they protect against more causes of damage.

[pscal]

I think the fusible trace is undesirable compared to a proper fuse in terms of actual safety. The manufacturers of fuses know a lot more about breaking a circuit than BLFers do. Even if BLFers design a good trace, who knows what it will actually do once the part has been created? Fuses are a known quantity. It may well be cheaper to design for a fuse from the get go than to mess around with testing fusible trace prototypes.

I think the fusible trace is a bad idea for the reasons above.

Edit:

My preferences in order

1. No fuse at all
2. actual fuse
3. Fusible trace (which i will bypass)

[nqcken]

If no one else in the flashlight industry is using fuses or fusible links, why don’t we follow best practices and just put a BOLD IN CAPS warning front of the manual?

Just a thought…

[strango]

vote for “1 do nothing, as all others making parallel lights do.”

[RollerBoySE]

@strango

vote for “1 do nothing, as all others making parallel lights do.”

Not true, most either make it impossible to put in the batteries the wrong way (mechanical) or use springs that will melt upon a short, or both.