Q8 modding

try to make the world idiot proof and the world will come up with a better idiot….
Has it been discussed what could happen if one would put different cells at extremely different states of charge in the Q8?

Running cells in series presents virtually no risks with a single discharge cycle such as in a flashlight.

The issue comes when you never remove them from the pack and charge them in series as well. In that case they most assuredly can become unbalanced (although much less then it is made out to be with quality cells). This is a non-issue for flashlights though since we remove them and charge them individually.

The only thing you need to do in the case of series cells is set the cutoff voltage slightly higher then you might otherwise with parallel cells (say a 3v cutoff instead of 2.8v if you really wanted a margin of safety from killing the cells).

Seeing as the big manufactures use almost exclusively cells in series, I am going to say that the benefits far outweigh the possible side effects. There is a reason they run them in series.

There is virtually zero fire hazard with series over parallel cells as long as the cells are installed properly and you can install fuses to prevent the worst case scenario.

Why?!

You wanted to make a photoflash in SRK body and you did it! Great job! And now I want to do a good flashlight in this body :slight_smile:

Be nice, now. A big Thank You for getting it to where, someday soon, it’ll be available to play with.

I’m hoping those who got the early samples may have already modded them and care to show their work.

Yes.

Just need to decide - four triples or three quads.

Please read one more time.
Parallel config is safe and well-known - one rear pcb with thick traces and 4 springs, one thick brass ring soldered to the driver and button top cells.
4S or 2S2P mod require springs from both sides, one contact plate should rotate, other should be fixed, this is not safety decision.

Big brand lights often have battery carrier.
Also

I still dont see any difference in self-cost with boost drivers.

Well Bistro and Narsil offer such wide setup options and easy too, not matched by any manufacturer, these feel like the most modern UIs

On highest modes less bright does not seem like an issue for high end lights might not get less bright because of voltage drop in cells, they too won’t sustain max brightness because of stepdown.

No choice in UI feels outdated IMHO
Habkng manufacturer decide when or at what temp stepdown is instead of user feels outdated IMHO
Whining PWM feels outdatelights

Used a little Olight the other day, nice regulated output till BAM darkness without warning, first thing popped in my head, “dang the Q8 would have stepped down as warning and I would have had several hours to a day of useful light” that a brand method felt outdated IMHO

As a kid I needed the full amount of cells get get toys running, that was a long time ago, being able to use a light with empty spaces where cells ought to go feels so nice.
And besides all can say how safe series is, when I joined it wasade clear extra care should be taken for series lights and the exploding lights are series lights.
On longer lights the cell at the head gets more heat and thus wears different over time, I swap each charging cycle, this of course is less of an issue with the cells next to each other.

Oops, you’re right, I missed that little word, boost somehow. It was right there all along. Sorry. :person_facepalming:

Still, I’d say buck is better for one reason: it doesn’t strain the cells as much as boost, especially for high output. AFAIK, for cell strain, boost is the worst, then direct drive and/or linear, then buck the least.

How do you all calculate this?
I (in)=I (out)xU(Vf)/U(Vcell)хE(driver eff.) (fixed)
With same eff., there is no difference.
In real world processes are not so simple and input power is not linear and have some “ripples” but both boost and buck have this effect.
Also, how boost can be worst, if peak current will be always lower than DD with same cells-leds?

Well, maybe I don’t understand it well enough. But, I’d think that in converting voltage, the max current is also converted in the opposite direction. So, in a boost circuit, the max current to the emitter(s) is necessarily lower than the max current from the cell(s). For a buck circuit, it moves the other way. Since the voltage is being lowered from cell(s) to the emitter(s), the max current available to the emitter(s) is higher than the max from the cell(s). So, for any specific desired current draw at the emitter(s), a boost driver draws more from the cell(s) and a buck driver draws less.

As the cells deplete, the difference is even greater, as boost drivers have to convert harder, which works the cell(s) harder, at the same time that the cells themselves are getting weaker. Alternatively, buck drivers are actually working less the lower the input voltage goes. The cells have it easy the whole time, while the buck driver can keep the output to the emitter(s) in perfect regulation longer as well.

Like I said, I may not understand this well enough myself. Texas_Ace makes these statements all the time, though. Perhaps he can step in and explain this difference better than I can.

.
.

EDIT: Just wanted to address this last statement of yours as well.

In the world I live in, boost is only needed when the cell voltage is too low for the emitter. So, to use the same cell/emitter combo, you would either have to drive the cell harder to keep the same current to the emitter, OR else drive the emitter less hard to keep cell drain the same as DD. There’s no way to keep the same input current as DD while also keeping the same output current as DD. Am I wrong?

To have 80W output, you need 90W input (example).
Please show me difference between eating 90W of power with buck or boost drivers.

Not talking about watts, but current. If you want to talk about watts, the difference is made up in the voltage. So, 4V times 10A is 40W but 10V times 4A is also 40W. The latter case is easier on the cells.

Once again.
Four cells. 90W.
4P=90/4*4=5.6A each cell load
4S=90/16=5.6A circuit load
:student:

Yeah, there’s your difference. :wink:

Correct, sooner or later we will also. Till then we have to make do with what we are given.

The GT just might give us a good carrier to work with for future lights.

He is correct, the current draw from the cells will be the same in either case. The difference is what happens in the wires and traces.

For example in RMM’s mega M6 where everything is in parallel he has to use 3 FET’s, thick bypasses, multiple 18awg thick wires just to reach a max output of ~150w (about 45A) and he said he could bearly get the driver back in place with all the wires crammed into the cavity.

On the other hand my 4S Mega M6 uses a single FET, 22awg bypasses iirc, a single 18awg wire to connect all the LED’s and I then had to add 6 feet of 20awg wire to act as a resistor to reduce the current. I still am putting out ~200w (~60A if they were in parallel with high drain cells) while trying to reduce power.

This vastly improves efficiency and opens the possibility of using a buck driver for even larger gains. The TA buck will handle 15A but a buck driver to handle 45A would not fit in the flashlight.

Hey, we are talking about cells config. All agree that series is better config for leds.

I finally figured it out. Sorry kiriba-ru. What can I say? You caught me not using my brain. :person_facepalming:

No prob.
Im not arguing for buck or boost or else. I just know that it is very hard to short-circuit 4P SRK and much easier to do same with 4S cells without carrier.