Sorry, I had these ordered, for less but that vendor changed prices
Gearbest has the for $17 (4) but IDK how shipping is.
Let me check AE account:
wut?!? https://www.aliexpress.com/item/4PCS-Original-Sanyo-18650-3-7V-3500mAh-Rechargeable-Lithium-Battery-NCR18650GA-battery10A-Discharge-Free-shipping/32693124032.html
what happened to that price?
You can compare the 5A discharge curves of the 20R's, 25R's, and 30Q's via HKJ's comparator to see. Against the 25R, the 30Q matches for the first 1/4 Ah, solid better after that. Against the 20R, the 20R is better for the first 1/8 Ah, then drops like a rock after that. I still have quite a few 20R's, lots of 25R's, and lots of HE2's as well, but the 30Q would be my choice every time over those. The VTC6 compared to the 30Q at 5A has the VTC6 stay'n ahead the whole way through a full discharge, but not by much.
No reason to pay the extra $$$ for VTC6's when the 30Q's are so close in performance and capacity. I just like seeing the best, testing the best out there, and right now that's the VTC6's, disregarding cost of course. HKJ's internal resistance rating on the VTC6 is 0.03 ohms - only cell I think he's tested that got that low - the other high performance cells made 0.04 ohms. Of course he hasn't tested the "best" cells though I know of (26650's, so not relevant for the Q8): the LK 26650 5000, and the IJOY 26650 4200. The 2 high performance EFEST 26650's come close, but are still 0.04 ohm rated/tested.
It's all here: http://lygte-info.dk/review/batteries2012/Common18650comparator.php. Great tool in looking at specific drain amp settings. For our max/turbo use, the 3A or 5A curves would be the ones to look at, for 12A to 20A total in the Q8.
It should be made clear though what's so nice about SRK lights, and the Q8 specifically, is how flexible it is in terms of batteries for the world's most popular LiIon form factor - the 18650. Think about it - you can use just one cell, or up to 4 matching cells, from recycled power packs, or the latest greatest. Only limiting factor is having a protruding Batt+ top.
Again go to the the ultimate source: http://lygte-info.dk/review/batteries2012/Common18650comparator.php
Pick the HG2 and the 30Q, and only turn on the 5A curve. I think this is what I've seen in my own tests - slight favor goes to the 30Q, but again, they are very close. If you can find a better/cheaper source for HG2's or already have HG2's, they are a good cell and will perform well.
The differences shown in HKJ's tests could be negated by normal batch to batch, or cell to cell variations.
To add confusion, some lights perform better with a cell that tests lower. I’ve seen a cell give more amps, but less lumens. The thing is that it’s not cold hard facts, there are always variables.
Any of the top cells will work well. The ultimate high performance is usually, or typically, only in the first seconds of use anyway and even then the differences are subtle and undetectable to our eyes.
Okay, a question I should know the answer to, but just for assurance, how “matched” do the four matched cells need to be? Should we check capacity before using them as a set?
As long as they are within .1-.2V they should be fine, the closer the better naturally.
I charge all my cells on a parallel charging bank and My rule of thumb is .5-.6V between cells before connecting them together. I have done as much as 1V without any effects though.
Well, others are more qualified to answer this, but for me - reasonably close. Certainly they should be rated the same and be the same make/model, and hopefully a top 5 brand (LG, Sam, SANYO, PANA, SONY). Better still from the same batch, typically if bought at the same time. If you met those conditions, I'd feel safe enough. If fairly new and not much charge cycles, I wouldn't check, but if they are a bit older and racked up some usage, good idea to check they are charging up to the same level, taken into account charger slot differences, etc.
That’s what I thought. Thanks guys.
I don’t think the matching is as crucial with 4 cells in parallel, it’s in series where it really starts making a big difference.
That said, I still use the same batch, charge together, and watch for one to get off from the others both coming out of the light and going back in.
yeah I have sets, check IR and mAh
use them as sets (both series or P)
when I put them in the charger check regularly to see if they all charge about the same.
I’ve got to say that I’m absolutely giddy with delight over the idea that if TF gets this right (*crosses fingers AND toes), I’ll be wielding a pair of bada$$ lights at a superbly economical price. Thank you to everyone involved with getting this thing happening!
I wonder what the TF price will be after this initial group buy if it’s made per the specs on the first page. It seems like it will stack up against lights costing easily three or four times as much.
Seems so simple doesn’t it? All they have to do is build the driver exactly as it was designed and tested and they have a serious top shelf winner on their hands. Don’t follow the recipe, they just have a mess.
We pinged them again - no reply yet. Tonight will have the protected GA's in, so will add those and solder topped HG2's to the list.
I work with many Chinese shop owners and the one thing I notice for 110% certainty is they are extremely tight as all mighty penny-wise. An example is asking for credit for literally 1/2 a bottle of fizzy drink - because the boxes come in 15’s, and 7 were leaking. I said I would credit half a box…… meaning the 7 damaged…. but he took me literally and demanded 7.5 were credited :person_facepalming:
We are all hoping for that driver to be to 100% spec but I’ll be honest I’m getting a little apprehensive now. Obviously I know not of what’s going on behind the scenes. But I think 8 months is pushing the boat out……
Adding latest battery tests, including the protected GA's just received last night, the LG2's with solder tops, and 30Q's solder tops to make the comparison with the LG2's
Measurement of the Protected GA's:
Measurement of the HG2's:
Measurement of the 30Q's (new cells):
Notes
- Those Protected GA's seem to do well for protected cells, least better than I thought. They even out performed the solder top GA's. Could be being brand new, charged slightly higher, and the new Seiko high performing protection circuit, or combination thereof.
- The last test of the solder top 30Q's produced the highest amps, even better than the VTC6's. These are also very new 30Q's recently bought from GearBest when they had them in the US warehouse. This may have been the first time used as well.
- For the most part, lumens output and amps track pretty well.
Edit: Ooops, typo: lumens taken at 30 secs, not 20 secs
Nice Tom
Good tests!
Thanks
Nice info Tom. One question. Do you sleep? 
Re: matched cells,
I don’t think the capacities or discharge rates of the cells need to match, only the voltage close enough so they don’t charge each other so fast they heat up. I am sort of planning to use whatever cells I am not using elsewhere, some of the time. It will run off the high discharge ones when in turbo mode and the high internal resistance ones will charge them back up when it is switched off. That won’t give 6,000 lumens, but it beats running on less than four cells.
Is that safe?
Some of the better protected GA’s (e.g. Seiko chip and 3 MOSFET) have been measured to have minimal difference to unprotected at around 5A draw.
Thanks for all the tests. I’d be interested to see how well unprotected GA without solder blob works.
Is your local hospital burns unit expecting you?