$100 Budget MONSTER-Lights w/ 15000+ lumens: Wurkkos TS32-Nichia 519a. Comparison with output HEAVYWEIGHTS: Haikelite HK05, JKK76, Nightwatch NS59v2 Chaos, etc. (Summary & measurements on P. 1)

I ordered JP40 from https://18650-21700.com as the site indicated “available.” 3 weeks later they still don’t have an aswer for availability. Austin at the site offers to send me instead their AM04 battery, which I plan to turn down primarily because it hasn’t been reviewed by @Mooch, like the JP04. Do you (or anyone) know what this AM04 is? Thanks.

AM04 VS JP40:
AM04 offers cooler operating temperatures under high loads.
AM04 has higher discharge rates with equivalent energy delivery.
AM04 extended operating range (down to 2V per cell).

1 Thank

I’m testing the AM04 now and doing a lot of poking around (taking quite a while).

Extraordinary claims require extraordinary evidence to back them up. Ask Austin what he means (for any statement you’re wondering about) and for the evidence to support it.

3 Thanks

Good to hear you’re involved with this AM04. For all my hobbies (cars, photography, audio, etc.), I’ve learned to trust only third party’s evaluation. Or rather, good third party evaluation.

Using data from your review of JP40 am I doing this correctly for calculating how long the JP40 lasts at 60A to 81.8°C?
(2.9 AmpHrs * 60 min/Hr)/60 Amp = 2.9 minutes?

Other “advertised” specs regarding AM04 that woud be interesting to compare to a third-party’s test:
160 Amp Peak for 2 Seconds
45A Continuous, 60A for 3 Min to 80°C
1500 Cycles to 80% Capacity at 12A Discharge & 4A Charge
<30 Min to 85% Capacity, Fast Charging Capabilities

1 Thank

Correct, for a new single cell in open air.
For a pack it will happen much sooner. And as a cell ages it will happen sooner and sooner.

Not me. :joy:
That is a claimed one time “capability” and not a formal spec. IMO it is extrapolated fantasy based on Ampace’s JP40 capability claim of 140A for 5S

The time between pulses is not specified by AM (can I pulse it like that every 3 seconds? Heck no) and neither is its effect on cycle life…which might be extreme.

I don’t understand what the test is here…60A pulses with 45A between?

Takes several months to do.
Needs other cells to compare against or not much use, a huge effort. AM would be the only one who might do this…see what he says!

Any cell can be charged incredibly fast. It’s the damage it experiences that is the important thing and that takes cycle life testing to work out and months to do. Fast charging plates out metallic lithium too (can cause short circuits) so safety is an issue here.

I’d actually love to see the results of all the above testing but by the time it’s done we’ll probably have samples of the next couple of wonder cells in our hands. Worth asking AM to do it though. He makes the $$$ if the results are good :slightly_smiling_face:

3 Thanks

Thanks. It’s often “illuminating” to read a good reviewer’s comments on advertisement claims. Regarding fast charging, am I correct in thinking that you don’t recommend using “fast charging” capability of any flashlight or battery? I have an Opus BT-C3100 charger and it always seem to start charging at 0.5A and gets lower towards the end of the charge. Is that considered the “best” for prolonging battery’s life and capability?

I was thinking the underlined part is to imply it has the same high current discharge capability as JP40, which in your test was 60A for 2.9 minutes? That’s why I was asking that question on how to calculate the duration (2.9 AmpHrs * 60 min/Hr)/60 Amp = 2.9 minutes). Really I was just putting out there to see if you have any comment. Curve in Mooch’s JP40 test, for anyone following this discussion:

Imgur

2 Thanks

It depends on the cell’s charging specs and the charging current from the light/charger. 3A could be very fast for some cells but a P50B would take that for a couple thousand cycles.

For any cells though, yes, charging at above the “standard” charge rate will speed up aging. Slower than the standard rate will help slow aging a bit. We’re not talking fundamentally noticeable differences in life here, just little things we can do to help a bit.

Ahh……thanks, yes, about 2.9minutes for LT22710A. I don’t call it a JP40 because of the Lt22710A on the wrap and there are Ampace-wrapped JP40’s available now (in testing) and those I’ll call JP40’s.

Test report for the AM04 this week I hope. Soooooo much to cover.

3 Thanks

Niiiice. Stay of execution for my unfilled order of JP40 from Austin. If AM04 is indeed as advertised I might switch to it.

Will you be posting AM04 review here on BLF?

A link at least. My Patreon supporters will see it first and then some time later it will go public.

4 Thanks

I ordered JP40 from https://18650-21700.com/ but after 3 weeks they still are not available. I asked for refund but Austin sent me AM04. So little money is involved and he worked hard to answer my questions so I am not mad at all, actually looking forward to testing it.

Because of my questions, Austin did test to see how long he could run the AM04 at 80A. The result was that the battery delivered 1.25ah in 55-60 sec on average from the AM04 cell. There was a graph that seems to show internal Tmax around 76 C, but the graph seems to look clipped above 76 so I am not sure. Austin says the AM04 runs cooler especially at the highest discharge rates and will match the JP40 performance within a few percent.

So whether I want to or not, I am going to be one of the first crazy hobbyists to have hands on AM04. Austin is so high on it, and he used to sell JP40 so it’s hard to think it’s going to be a dud. It should be fun to find out for myself.

2 Thanks

Thanks for the update and good luck!

2 Thanks

I think because of the questions (where’s the proof please) that I asked per your suggestion :+1: :slightly_smiling_face: , Austin is now in full testing mode of AM04 and has loaded test info onto the web site. I brought up your test of JP40, specificall the 60A @ 81.8 C curve, so he’s put internal temp probe as well.

The latest info Austin shared with me is AM04 appears to be more stable and cells are more consistent, and also goes to about 3 minutes at 60A (looks like 2 min from web site?) or 30 sec at 100A (!). Meanwhile below is my own strictly amateur-level test of the batteries. The light is Nightwatch NS14R v2 with 14 x SFQ55.4 LED. For anyone following this and not familiar it is likely the most powerful single 21700 flashlight on the market with a big plate (no spring) on tail cap.

NS14R v2, top 3 readings of multiple, batteries fully charged and lights cooled before each run.
AM04 from 18650-21700.com: 21300 (top 3: 21000, 21000, 21300)
Nightwatch 45XP “80A”: 19300 (18890, 19200, 19300)
Molicel P45B: 18620 (18240, 18520, 18620)
Lishen LR2170LH: 16980 (16680, 16810, 16980 - old batteries)
Samsung 30T: 16840 (16090, 16340, 16840)
Wurkkos 5000 mAh No name: 10800 then the light died

Imgur

2 Thanks

In regards to my post in in this topic, I promised are more detailed/extensive reaction, as not wanting to spam the other topic.

Those numbers are not exactly accurate and in my experience with my Opus BT-C3100 it is often also not very consistent between several measurements and/or slots.

I did a quick/short test with my Opus BT-C3100 tonight, for my JP40. I cleaned all contacts of the charger and battery with Iso, so that it not would be an influence. I did 5 “Quick Tests” with the cell, for each of the 4 slots. Every time disconnecting the battery for several seconds, before doing a next test.

Here are the results in mOhms:

Slot 1 Slot 2 Slot 3 Slot 4
39 32 34 30
39 32 34 36
41 32 33 35
40 32 35 29
36 32 35 33

Since I was so surprised by the consistency of ‘slot 2’ (I literally never had 5 identical results in a row before!), I decided to test a 40T5 in slot 2, which was (almost) charged to the same voltage, to see if that slot really test so consistenly! (Also cleaned the contact points of the 40T with Iso.)

This gave the following results:

Slot 2 with 40T5
63
49
78
67
59

So it was wildly inconsistent with the 40T5. I have no clue what caused that, other than that the Opus may favor the (way) lower IR of the JP40 over that of the 40T.

I also did 3 quick tests on my Fnirsi HRM-10 Battery Internal Resistence tester, just to check the consistency and the ‘actual’ difference between the JP40 and the 40T5. There are some fluctuations between the measurements, but remember that we are talking about way more accurate and lower numbers, going down to sub-mOhms, and therefore minute differences.

Results:
JP40: 2,93 mOhm, 2,90 mOhm, 2,93 mOhm
40T5: 11,05 mOhm, 11,03 mOhm, 10,95 mOhm

Quite consistent results, with minor differences, which could even be down to placement into the somewhat finicky battery holder tester.
The difference/delta between the JP40 and 40T5 is to be expected and also consistent, in contrary to the readings from the Opus.

Most “cheap” dedicated battery testers, including my Fnirsi and also the one that Austin used in that photo, measure ACIR/AC IR, which means there is no load put onto the battery and it just measures it’s (‘at rest’) impedance. Most of the affordable testers are quite accurate ‘out of the box’ now a days. These testers can measure way lower/smaller resistances than for example most digital multi meters. The other advantage is that they use a 4-wire measurement (so 2 different contact points and individual measurements), which gives a more consistent result and less chance on ‘faulty readings’, than a 2 wire measurement, like what the Opus uses.

I do assume that the Opus also does an ACIR measurement, given how quick it does the test. The other test method, I will explain more about that later, is the DCIR test measurement, gives a bit higher value than the ACIR measurements, but nowhere near the amount to explain the high numbers of the Opus.

What I think is that the Opus, has higher internal resistance, influencing the results and that the fluctuations between readings could (maybe) partially be explained by the ‘2-wire-method’ which has a higher chance of faulty readings.
It would/could still be usefull if it gave consistent readings (almost) every time, with every differnt cell. That way you can at least extrapolate which cell has lower IR than others, between different types or between batches of the same cells, to see which are better than others.

If you have a new battery and you want to know if your test results match the factory spec (and so if they are accurate), you can always compare them to the (official) data sheet.

As an example below a picture of the (official?) data sheet of the Liitokala 4695E I recently purchased:

As you can see it mentions the AC(I)R value to be equal or below 1.8 mOhm at “BOL status” (= Beginning Of Life) and the DC(I)R equal or below 4.0mOhm.

I measured my 4695E ACIR and with 1,72 mOhm it was indeed within spec!

Now onto DCIR measurements:

Since this is a bit more complicated than ACIR measurements, and my knowledge is still very limited on this subject, I will give you some general explanation, to prevent me from getting certain details wrong. I will link you afterwards to a really interesting topic I found here on BLF a couple of days ago, with a lot of information about DCIR measurements. (I believe I also found an interesting topic on EEVblog(.)com, but I will try to find that one some other time.)

DCIR measurements are internal resistance tests when the batteries are under load, which means it gives a way better indication of how the battery will perform when in (high drain) use. Equipment for properly measuring DCIR is quite a bit more expensive (and extensive) than the “cheap” ACIR tester. It also requires a lot more indept knowledge and skill to perform (and interpret) these measurements.

There is an IEC standard for how to measure DCIR, although a lot of people testing battery prefer their own process/standard, for several reasons.

In this topic (Measuring impedance) you can find tons of information about measuring impedance, including a LOT of information on DCIR testing, including some very interesting posts by @docware, @Pajda_cz and also @HKJ.

To the experts on this subject who maybe read this post: I hope I didn’t make too much mistakes. My knowledge is still pretty fresh and limited, so therefore I kept it more of a ‘general explanation’, to avoid incorrect detailed information. :sweat_smile:

2 Thanks

Thanks for taking time to write detailed and helpful summary. Thank goodness I have not developed an interest in testing batteries yet so my Opus will have to do for now.

What about capacity measurement by Opus 3100? Please don’t tell me it’s sh*tty also lol.

My AM04 batteries are more or less “destroying” every battery in its path, albeit against older batteries. The Nightwatch Mystery 45XP and excellent Molicel P45B are a close second.

2 Thanks

Austin did the test below in response to my quoting of your chart above. It seems like the AM04 gets to Tmax 80 a little earlier @ 2 minutes. But as an amateur I don’t know if this means anything, or anecdotal, just something routintely seen due to battery to battery variation.

As for my simple test, output at Turbo turn of extremely current-hungry lights, the AM04 battery has continued to exceed all other high-current batteries that I have which include Lishen LR2170LH, Molicel P45B, new Nightwatch MX45 “80A”, and Samsung 30T.

Source: AM04 21700 - 160 Amp Capable - Racing Performance & Long Life – 18650-21700.com

Imgur

I think you’re looking at the internal temp curve and not the external temp (which I measure)?
Running the AM04 at 60A until 2Ah have been discharged doesn’t bring the cell up to 80°C external.

2 Thanks

Thanks you’re right, that Tmax 80C is the internal temp curve. Austin has pictures of the process of internal temp measurement, complete with “Do Not Attemp” warning. It looks so dangerous.

Latest measurements of Turbo output, this time with Nightwatch Super Valkyrie that was reviewed here. With SFP55.2L LED, this is likely the most powerful single LED, single 21700 battery flashlight on market. The AM04 continues its good show, bringing the Super Valkyrie to highest ever level (13340) in my many different attempts with different batteries. It seems at least with my amateur testing, AM04 is the most “powerful” battery I’ve had.

AM04 from 18650-21700: 13340 lm (top 3 readings13020, 13300, 13340)
Molicel P45B: 12960 (12670, 12940, 12960)
Nightwatch 45XP “80A”: 12340 (12270, 12280, 12340)
Lishen LR2170LH: 12330 (12110, 12190, 12330)
Samsung 30T: 11300 (11270, 11270, 11300)

3 Thanks

AM04 and JP40 both tout their “tabless” design. I have no idea how unique this is (are they the only ones who have it?), but curious enough to find this article:

The enhanced efficiency of the ProCore18V+ battery is achieved thanks to the tabless design of the integrated 21700-series battery cells. In conventional lithium-ion batteries (left), each individual battery cell inside has a narrow connection band to the anode and cathode on both sides – the so called “tab”. These tabs create a bottleneck for the power delivered by the battery and contribute to the cell’s electrical inner resistance, causing heat to be generated. To ensure that the battery doesn’t overheat, particularly during high-performance applications, the tool shuts itself off, leaving the remaining energy in the battery unused.

In tabless cells (right), the electricity can flow through countless paths across the entire length of the anode and cathode rather than being restricted to one or a couple of tabs. This new design reduces the inner resistance of each individual cell by around 50 percent and therefore substantially for the battery pack as a whole. Significantly less heat is generated as a result, which is a limiting factor in demanding applications. All in all, this translates to up to 71 percent longer runtimes compared to conventional batteries thanks to the combination of tabless design and precision-tailored battery management.

1 Thank

The AM04 (identical to the JP40), JP40, 40PL, VX40, and 45D are all “tabless” designs.

Interestingly they all have a single tab to the top contact. :joy: The only true tabless cell so far is the Tesla 4680 and maybe other 4680’s. All are “continuous tab” or “full tab” designs though that does indeed significantly reduce the IR of the cell.

2 Thanks

Do these “ture” tabless cells have any significant advantages over cells like the JP40 and other “half-true” (?) tabless cells?

1 Thank

IMO we can”t say yet whether there is a practical, noticeable difference when using fully tabless cells. The individual cell chemistries and construction might make a bigger difference than the tab.

We’ll have to wait and see if fully tabless construction is ever used on the smaller (18650/21700) round cells and then start comparing performance. Could be months or even a year or two before that happens though.

Though I’m a big fan of diving into the tech details IMO it’s best to just compare specs and actual performance vs what construction or chemistry is used when choosing a cell. Tabless or not doesn’t matter once we put the cells in our devices. :slightly_smiling_face:

2 Thanks