# need help understanding an HKJ Battery Comparator graph

#### seery

##### Flashlight Enthusiast
If a light draws ~4-5 amps, (in terms of output and runtime) what does this graph tell us?

Sanyo/Panasonic NCR18650GA 3500mAh vs. Samsung INR18650-30Q 3000mAh

Battery chemistry and understanding graphs is not on the upper shelves in my wheelhouse. Any help understanding this graph would be appreciated.

This is from HKJ's battery comparator.

Last edited:

#### ChrisGarrett

##### Flashlight Enthusiast
It just tells us that with a 5A current load applied to both cells, the 30Q Samsung will hold a higher voltage down to about 3.3v (30% capacity left?,) where the higher capacity Sanyo/Pannie NCR-GA then prevails and goes on a bit further.

Whether that added voltage matters, in current regulated lights, I can't say? I have a light that Ven gave me that is direct drive on turbo only and the 30Q might give me a bit more output, in the beginning, than the NCR-GA.

Direct drive is just running off of what the battery can provide, in real time, like older lights, as the voltage drops, output drops in concert.

Regulated lights (current regulated) have a chip that controls the current flow down to X, Y or Z point (power/watts: voltage x amps) and then either shuts off, or drops down to the next lowest level.

Chris

#### Phlogiston

##### Enlightened
As an extension to Chris' post:

If your light is direct drive and draws something on the order of 5A, the 30Q cell will give you higher output at first, but the GA will give you more total runtime. You can't say much more than that from this graph, because the graph is constant-current, but direct drive current drops with voltage, which means that the light will produce a long tail of low output that the graph can't be used to determine.

If your light has a linear constant-current buck driver - the usual type - which runs at 5A, then you can use the graph to work out how long you'll get before the driver falls out of regulation. To do that, you need to know what cell voltage the light needs to stay in regulation. This varies from one light design to the next, between samples of the same design, and even with the level of corrosion on your sample's contact surfaces!

Eyeballing the graph for a rough example, if your light needs at least 3.8V to sustain regulated output, then the GA will hold that level for the first 0.5Ah. 0.5Ah ÷ 5A = 0.1h, or six minutes. By a similar calculation, the 30Q will hold that level for 8 minutes, 15 seconds, giving you an extra 2 minutes, 15 seconds.

In this case, the 30Q's ability to sustain a higher cell voltage for longer means that the constant-current driver can stay in regulation for longer. If you value constant output for as long as possible more than you value absolute runtime, the 30Q is for you.

On the other hand, if you value the ability to produce at least a little bit of light right up until the end of a longer runtime, even after the light falls out of regulation, then you'll want the GA cell for its higher capacity.

As another example, if your light can still sustain 5A constant-current regulation at 3.4V, then you'll get 22 minutes, 30 seconds from the GA cell, and 24 minutes from the 30Q, giving you an extra 1 minute 30 seconds from the 30Q.

The GA cell continues to close the gap until they've both dropped to 3.3V at 27 minutes, at which point the GA starts to pull ahead. If you have a light that can draw 5A from a cell right down to 2.8V - very unlikely - the 30Q will only give you 34 minutes, but the GA will hold out for 38 minutes, turning the tables for the GA to buy you an extra 4 minutes.

Note that you can only draw these conclusions from this graph for as long as the light stays in 5A constant-current regulation. The moment the voltage gets too low, the light will drop out of regulation. As Chris says, the driver may shut down completely, or step down to a lower-current mode. It may also drop into direct drive. Whatever it does, it won't be drawing 5A anymore, it'll be drawing less, which means output will be lower, but total runtime will be longer than the graph might otherwise seem to imply.

Don't get me started on boost or buck drivers with switch-mode regulators As with direct drive, this type of graph can't say much about performance with those.

#### seery

##### Flashlight Enthusiast
Thank you Chris.

So the +500mAh of the NCR doesn't come into play until the cells are discharged down to or below the 3.3V?

If both cells are only discharged to ~3.3V, then the 30Q would hold higher voltage from start to finish? Wether a particular light benefits from the higher voltage would be the question.

I usually run the K60 for extended times on high and turbo and after trying a few different cells I'm most impressed with the Sanyo/Panasonic NCR18650GA 3500mAh cells.

Just trying to decide if the 30Q's would be worth the \$ try.

#### ChrisGarrett

##### Flashlight Enthusiast
Thank you Chris.

So the +500mAh of the NCR doesn't come into play until the cells are discharged down to or below the 3.3V?

If both cells are only discharged to ~3.3V, then the 30Q would hold higher voltage from start to finish? Wether a particular light benefits from the higher voltage would be the question.

I usually run the K60 for extended times on high and turbo and after trying a few different cells I'm most impressed with the Sanyo/Panasonic NCR18650GA 3500mAh cells.

Just trying to decide if the 30Q's would be worth the \$ try.

Like Phlogiston states, it will ultimately become a question of runtime for most multi-mode regulated lights and those that are direct drive. If you find yourself stuck in a cave for an extended period of time, you'll want the NCR-GAs. If you have the ability to swap cells, or recharge them frequently, you might want the 30Qs in something like my 3Tronics Convoy S2+ 6A direct drive triple XP-L HI light, if going for maximum badassery is what you're after. Me, I have multiple lights/cells and I don't use them all that much, so runtimes generally aren't a priority for me.

You can't really go wrong with either cell and the NCR-GA is right there at the top of the hill for 10A cells. You're going first cabin with either, however.

Chris

Last edited:

#### seery

##### Flashlight Enthusiast
Thank you Phlogiston.

This helped me understand and tie together what Chris was saying. A lot to try and comprehend, but feel it's starting to sink in.

Will have to research the Acebeam K60 specs and plug them into the above and hopefully get a better understanding of how the 30Q's may or may not benefit over the NCR's in that particular light.

It'd make me feel a lot better if I could understand what and why the stop watch shows what it does.

Your guys help is very much appreciated.

#### seery

##### Flashlight Enthusiast
And thank you Chris for the follow-up as I was typing.

#### ChrisGarrett

##### Flashlight Enthusiast
Thank you Phlogiston.

This helped me understand and tie together what Chris was saying. A lot to try and comprehend, but feel it's starting to sink in.

Will have to research the Acebeam K60 specs and plug them into the above and hopefully get a better understanding of how the 30Q's may or may not benefit over the NCR's in that particular light.

It'd make me feel a lot better if I could understand what and why the stop watch shows what it does.

Your guys help is very much appreciated.

The stock K60 with one XHP70 draws about 7A on turbo/high/L6 and since the four 18650s are in series, each cell will see 7A, so you're getting close to your limit with the NCR-GAs, although 7A is well within that dangerzone. They might wear out faster than say a 30Q, with a bit more headroom, but I'm just speculating on that. Kind of like constantly running your car near, or at, its redline vs. a normal, more benign RPM curve.

I find that the battery/cell/charger facet of this hobby is more interesting and enjoyable, but it takes a lot of reading, for sure.

Good luck!

Chris

#### swan

##### Banned
With the k60 it probably would pull 7 amps at the emitter at around 7 volts , so running with 4 cells in series with a forward voltage of 14.40 to 16.80 volts each cell will be drawing less than 4 amps.
So i would just run unprotected sanyo ga,s for max run time.

#### markr6

##### Flashaholic
This is a very good example why one shouldn't fall for the "3500mah vs 3400mah vs 3000mah". In most cases. Larger isn't always better, and even when it is, it's practically negligible (a few extra minutes IF you push the cell all the way to it's low limit, which I never come close to)

#### sidecross

##### Flashlight Enthusiast
I just resubscribed to candlepower-forums after being absent, and reading this thread made that purchase a good investment. :thumbsup: