Unless otherwise mentioned, capacity results are in ascending order, not in slot order. The ranking order between particular tests may vary.

Sanyo Eneloop 2000

Bought from Thomas Distributing (8 cells), manufacture date May 2006; bought from Online Shop Singapore (4 cells), manufacture date November 2006. Packaging for Online Shop cells labeled RWPrince,... Singapore, although the cells shipped from Hong Kong. As delivered they were partly charged, with an average cell voltage 1.308V (from TD), while the newer cells (from OSS) were all 1.310V.

• Off shelf discharge at 500mA (0.25C): 1420, 1420, 1423, 1441 mAh – average 1426 mAh (variation 1.5%).
• Off shelf discharge at 400mA (0.2C): 1413, 1431, 1452, 1455 mAh – average 1438 mAh (variation 3.0%).

• 1st break in (400mA, 0.2C): 1991, 1996, 1999, 2005, 2010, 2012, 2018, 2036 mAh – average 2008 mAh.
• 2nd break in (400mA, 0.2C): 1977, 2000, 2002, 2005, 2007, 2017, 2020, 2032 mAh – average 2007.5 mAh.
• 1 amp (0.5C) discharge: 1980, 1983, 1987, 1992, 1992, 2000, 2016, 2025 mAh – average 1997 mAh.

Testing the charger's effectiveness at fast charging these cells, I also ran some other tests on the 8 TD cells.

• 1 amp (0.5C) discharge after 1A (0.5C) fast charge: 1923, 1930, 1933, 1934, 1936, 1939, 1962, 1962 mAh - average 1940 mAh.
• 1 amp (0.5C) discharge after 1A (0.5C) fast charge and leave overnight: 1934, 1937, 1973, 1974 mAh - average 1954.5 mAh (first set of four only).
• 1 amp (0.5C) discharge after 2A (1C) fast charge: 1889, 1910, 1912, 1930 mAh – average 1910 mAh (first set of four only).

The first break in capacities seemed pretty good but to see if they would all go over 2000 mAh I restarted the break in cycle after the discharge rest, and sure enough, the 1999 mAh cell was up to 2005 mAh. The second break in with the second set of 4 left just one cell below 2000 mAh, at just 1977 mAh, but strangely the same cell gave 1987mAh at a 1 amp discharge (both in outside slots on the charger). With the best two cells from the second set, the best cell was always the one positioned in slot 2 or 3, while the other was in slot 1 or 4.

Update 20 June 2007: I've just put one set of Eneloops through another break in and was pleasantly surprised to see the results, both for their high capacity and for their consistency, within 1.1% of each other.

• Break in, about half a year on: 2092, 2100, 2115, 2116 mAh.

Update: Another six Eneloops have been tested (exact date unknown). Variation figures not shown for these because they are not from a single packet, and thus may or may not be from the same batch or manufacture date.

• Off the shelf voltage: 1.320 V average.
• Off the shelf capacities: 1341, 1348, 1374, 1396, 1387, 1392 mAh – average 1373 mAh.
• Break in: 2061, 2072, 2067, 2059, 2086, 2067 mAh (same order) – average 2069 mAh.
• 1 amp (0.5C) discharge: 2045, 2050, 2069, 2066, 2100, 2078 mAh (same order – some have gone up!) – average 2068 mAh.

Update April 2008: Another four Eneloops are being tested, with the date stamp June 2006. These results have not been included in the results shown in the table because they were made using an analyser from a different batch. These capacities all in slot order.

• Off the shelf voltage: 1.295 V average.
• Off the shelf capacities: 1316, 1305, 1338, 1318 mAh – average 1319 mAh (variation 2.5%).
• Break in 1: 1946, 1936, 1940, 1930 mAh – average 1938 mAh (variation 0.8%).
• Break in 2: 1943, 1934, 1939, 1929 mAh – average 1836 mAh (variation 0.7%).
• Break in 3: 1944, 1933, 1935, 1928 mAh – average 1935 mAh (variation 0.6%).
• 1 amp (0.5C) discharge: 1929, 1928, 1930, 1919 mAh – average 1927 mAh (variation 0.6%).

Update October 2017: A bunch of four year old 3rd gen Eneloops (manuf. code 13-09SA), which have just been sitting around unopened.

• 1st pack of Eneloops.
• Out of packet (500 mA): 1246, 1253, 1239, 1243 mAh. Average 1245 mAh. Variation 1.1%.
• Break in (400 mA): 1916, 1913, 1928, 1889 mAh. Average 1912 mAh. Variation 2.1%.
• Post-break in (500 mA): 1917, 1913, 1928, 1892 mAh. Average 1913 mAh. Variation 1.9%.

• 2nd pack of Eneloops.
• Out of packet (500 mA): 1246, 1255, 1236, 1238 mAh. Average 1244 mAh. Variation 1.5%.
• Break in (400 mA): 1922, 1917, 1921, 1898 mAh. Average 1915 mAh. Variation 1.3%.
• Post-break in (500 mA): 1925, 1920, 1925, 1900 mAh. Average 1918 mAh. Variation 1.3%.

• 3rd pack of Eneloops.
• Out of packet (500 mA): 1255, 1260, 1253, 1237 mAh. Average 1251 mAh. Variation 1.9%.
• Break in: (400 mA): 1908, 1915, 1907, 1911 mAh. Average 1910 mAh. Variation 0.4%!
• Post-break in (500 mA): 1909, 1916, 1908, 1911 mAh. Average 1911 mAh. Variation 0.4%!

• 4th pack of Eneloops.
• Out of packet (500 mA): 1255, 1260, 1249, 1256 mAh. Average 1255 mAh. Variation 0.9%!
• Break in: (400 mA): 1930, 1921, 1892, 1906 mAh. Average 1912 mAh. Variation 2.0%.
• Post-break in (500 mA): 1911, 1909, 1882, 1892 mAh. Average 1899 mAh. Variation 1.5%.

• 5th pack of Eneloops.
• Out of packet (500 mA): 1239, 1245, 1237, 1236 mAh. Average 1239 mAh. Variation 0.7%!
• Break in: (400 mA): 1923, 1918, 1907, 1920 mAh. Average 1917 mAh. Variation 0.8%!
• Post-break in (500 mA): 1930, 1925, 1913, 1927 mAh. Average 1924 mAh. Variation 0.9%!

I've posted some notes on the slot consistency for these Eneloops on the MH-C9000 page.

Eneloops get a ¾ charge when they are made, so those capacities represent 83% of their original charge! Not bad for four year old cells, and I fully expect those to increase at least 100 mAh after a few cycles. Also, each pack got really good consistency; 3rd gen Eneloops are very well made.

Eneloop XX 2500

Bought online from Intl Outdoor Shop, which no longer sells them. The XX was rebranded the Eneloop Pro at the end of 2011 but is still sold in some markets as the XX (powered by Eneloop Technology), normally referred to as Eneloop XX. Battery model HR-3UWX. Date code "11-10 HM", October 2011; 5 months old at time of test in March 2012.

Four Eneloop XX cells have been tested.

• Off the shelf voltage: 1.3075 V average.
• Off the shelf capacities: 1230, 1272, 1254, 1284 mAh (slot order) – average 1260 mAh (variation 4.4%).
• Break in capacities: 2444, 2467, 2435, 2496 mAh (slot order) – average 2461 mAh (variation 2.5%).
• 1 amp (0.4C) discharge: 2438, 2477, 2449, 2506 mAh (slot order) – average 2468 mAh (variation 2.8%).

GooP 3000 mAh

With such a high claimed capacity I don't hold high hopes for these. (A claimed capacity which is simply a lie indicates a poor quality manufacturer.) Sold by DealExtreme here (sold out), although I didn't buy them.

• Off shelf voltage: 1.302, 1.305, 1.295, 1.303 V ("slot" order, or rather how they came out of the packet) – average 1.301 V (variation 0.7%).
• Off shelf discharge at 500mA (0.167C): 1539, 1529, 1460, 1516 mAh (slot order) – average 1511 mAh (variation 5.4%).

These voltages (and later, the off the shelf capacities) are high enough for me to think they could have some LSD ability, which doesn't fit with the claimed capacity (the higher the capacity, the less low self discharge ability cells have). The variation of the off shelf capacities is getting quite high.

• Break in: 2123, 2100, 2097, 2075 mAh (slot order) – average 2099 mAh (variation 2.3%).
• Discharge at 1 A: 2122, 2099, 2102, 2069 mAh (slot order) – average 2098 mAh (variation 2.6%).

That consistency is actually quite reasonable; the capacities are not all over the place. The voltage on discharge was a little low, though, down to about 1.17 V after the first 500 mAh. My present thought is that they are clearly not as good as Eneloops, and should have been marketed as 2100mAh or 2200mAh, certainly no more than that. As they are, especially with the claimed 3000 mAh, they rightfully deserve the name "GooP".

One hour after the break in had finished I started the 1 A discharge. At the end of that discharge the cells all rebounded to 1.14V. That's a little low.

Next tests, both after a 0.5C charge.

• 1 A discharge: 2064, 2048, 2026, 1992 mAh (slot order) – average 2032.5 mAh (variation 3.6%).
• 500 mA discharge: 2061, 2037, 2015, 1992 mAh (slot order) – average 2026 mAh (variation 3.5%).

Solar Magic

Bought (not by me!) from Bunnings Warehouse in November 2012. (Also tested that month.) On the cell is "Cat no. SMB1000". Apart from that there's just instructions, and no other cell markings on these things, almost like they aren't a serious cell. On the packet is "Distributed by Merlin Distribution Pty Limited" and "Made in China". Also on the packet is "Batch Code: BATCH 1X" – again, not a serious batch number. They are marketed specifically for solar garden lights.

The basic advice for these is simply do not buy! These are the most pathetic brand new cells I have ever tested. I didn't have most of them for long enough to try a full Break In, but did do a bit of 300mA testing, which showed they're all over the place. The thought occurs to me that Eneloops cost less than twice as much, have more than twice the capacity, and have consistency a couple of orders of magnitude better.

The single cell I did do a break in on definitely improved, but very few people have a charger capable of doing that, or of testing cells to get a matching set. So most people using these things in their 3 cell solar light would have been picking three cells from the four at random; there would have been a 75% chance the worst cell would have been selected.

Test, Discharge Current	Cell Number				Variation
	1	2	3	4
1st test, 300 mA	160 mAh	636 mAh	393 mAh	713 mAh	345.6%
2nd test, 300 mA	278 mAh	789 mAh	794 mAh	955 mAh	243.5%
3rd test (break in), 200 mA	511 mAh	In service in a solar security light.
4th test, 1 A	497 mAh
5th test, 500 mA	479 mAh

Varta Ready to Use 2100

Straight out of the packet, loaned for test.

• Off shelf voltage: 1.309, 1.309, 1.310, 1.311 V – average 1.31 V (variation 0.15%).
• Off shelf 500 mA discharge: 1650, 1645, 1672, 1694 mAh – average 1665 mAh (variation 3.0%).
• Break in: 1957, 1950, 1978, 2015 mAh – average 1975 mAh (variation 3.3%).
• Post-break in (500 mA, after 1 day pause): 1939, 1930, 1959, 1997 mAh – average 1956 mAh (variation 3.5%).

These don't have the consistency of Eneloops (even four year old Eneloops), but they don't look too bad. The front of the packet equates 1 of these cells with 175 alkaline cells implying they'll last that many cycles, but the back of the packet claims 300 cycles.

Another couple of packets – look very slightly tatty (for example, the cardboard is fraying at the corners).

• Off shelf voltage average 1.305 V.
• Off shelf 500 mA discharge: 1677, 1692, 1677, 1689, 1649, 1677, 1640, 1679, 1654, 1660, 1638, 1654 mAh – average 1666 mAh. They have a slightly lower voltage while discharging than Eneloops but it jumps up to match the Eneloops' voltage if the cells are rotated in the slots. Dirty terminals?

Unless otherwise mentioned, capacity results are in ascending order, not in slot order. The ranking order between particular tests may vary.

Camelion AlwaysReady 2100

Loaned from a caver to see just how well they perform. More details are on the LSD Shootout page.

• Break in (discharge 420mA, 0.2C): 1855, 1944 mAh – average 1899.5 mAh (variation 4.8%).
• Slow charge (210mA, 0.1C); 2 day rest; 500mA discharge (0.238C): 1771, 1859 mAh – average 1815 mAh (variation 5.0%).

Eneloop 2000, 3rd gen

Tested when new in October 2017, these are back for a retest.

• Break in 2126, 2107, 2124, 2099, 2115, 2091 mAh, average 2110 mAh (25 & 27 April 2021).

The capacity of Eneloops increasing after a few cycles is well noted, so this increase is expected.

Eneloop XX 2500

Retested after intermittent use, and possibly some abuse. Three of the cells have a small amount of surface rust on the exposed metal surfaces. I gave them a break in, followed by a 500 mA discharge.

• Break in (set to 2400 mAh): 951, 1078, 1912, 978 mAh – average 1230 mAh, or 1002 mAh for the three "bad" cells. (18 April 2021.)
• 500mA discharge: 1226, 1330, 2083, 1245 mAh – average 1471 mAh, or 1267 for the three "bad" cells. (18 April 2021.)

Seeing that the capacities had increased, I attempted to cycle them three times at 1 amp charge, 500 mA discharge. I got back to the charger later to find the backlight was off and cell 3 was hot – very hot. My thermal camera saw 84 °C. It has resulted in the label shrinking and probably internal damage, although it measures 1.350 V.

Did the charger miss charging termination and also miss over-temperature termination, or did it only respond to the temperature when the cell got extremely hot? Did the charger keep charging after thermal shutdown?

As soon as I took the overheated cell out of the charger, the charger rebooted, recognised the other three cells, and proceeded to charge them without any further input from me (as it does if just left with newly inserted cells in it). The charger reported charging values ~2000 mAh, so they've taken a fairly full charge.

I set the three surviving cells to discharge at 500 mA (19 April 2021), then figured I'd try discharging the overheated cell as well. As soon as I inserted it the charger turned off. I pulled it out again and the charger returned to normal. There's definitely something wrong with that cell, but not completely short circuit, since it measured 1.350 V.

There appears to be something wrong with the charger too. Instead of 500 mA the cells are discharging at 552 to 557 mA. This is much more than the normal variation-from-nominal of the MH-C9000. The discharge went well, though, and it appears that the three cells have been completely revived. Even with a minute or two of discharging that was "lost" when the charger reset as the bad cell was inserted, they have come out very well.

• 500mA discharge: 2448, 2403, 2419 mAh – average 2423 mAh. (19 April 2021.)
• Break in (2500 mAh): 2614, 2546, 2600 mAh – average 2587 mAh (22 April 2021.)
• 500mA discharge: 2624, 2545, 2614 mAh – average 2594 mAh (22 April 2021.)

Update 23 April 2021: It's not the cell causing the MH-C9000 to turn off. It's the channel itself; the charger turns off when any cell is inserted in that channel. My MH-C9000 has lost channel 3! The toasted cell apparently charges fine when inserted in different charger.

Goop 3000

These appear to have some LSD ability but are not used much. Their typical voltage under load is only 1.15V.

• Break in (420mA, 0.2C of original 2100mAh test): 2010, 2035, 2043, 2071 mAh – average 2040 mAh.
• 500mA discharge: 2001, 2037, 2040, 2074 mAh – average 2038 mAh.

GP 1800

Sold by Dick Smith Electronics many years ago as a set of four with a fast charger.

First set of four had one cell that died a few months after purchase, leaving only three in that set. The remaining three have performed quite well in a shaver over the years, and have been observed to have a low self discharge rate – not bad at all for cells predating Eneloops by years, showing how bad the capacity race was for reliability. The three were tested in about 2007 or 2008 then again in 2017. They've lost a lot of capacity in that nine or ten years.

• Break in (360 mA, 0.2C): 1648, 1745, 1764 mAh – average 1719 mAh (variation 7.0%).
• 500 mA discharge: 1615, 1722, 1747 mAh – average 1695 mAh (variation 8.2%).
• 1000 mA charge then 500 mA discharge (in 2017): 904, 969, 896 mAh – average 923 mAh (variation 8.1%).
  That 3rd cell is under 50% of the label capacity of 1800 mAh. Time for some break in cycles.
• Break in (set to 1200 mAh): 1058, 1056, 1142 mAh. Average 1085 mAh (variation 8.1%).
  Looks like 1200 mAh wasn't a bad figure to set the break in to.
• 500 mA discharge: 1044, 1024, 1130 mAh. Average 1066 mAh (variation 10.4%).

The GP battery charger rejected the faulty cell as defective so it hadn't been used since, but for reasons best understood by squirrels it was kept. It showed 2.4 ohms and 0 volts before the semi-successful attempt was made to resurrect it. That the 500mA test capacity was larger than the 0.2C rate (360mA) indicates that the cell still needs a bit of cycling, but since its self discharge is so high it's due for recycling.

• Break in: 1048 mAh (variation with best cell of the set 68.3%).
• 500mA discharge: 1091 mAh (variation with best cell of the set 68.3%).

Four more cells from an identical batteries/charger set were tested in March 2011. They were given two full break in runs (each with two slow charges) plus a 500 mA discharge between.

• 1st break in (set to 1800 mAh): 1177, 1190, 1325, 1514 mAh – average 1301.5 mAh (average 2 worst cells 1183.5 mAh).
• Discharge 500 mA: 1157, 1165, 1295, 1509 mAh – average 1281.5 mAh.
• 2nd break in: 1162, 1171, 1322, 1521 mAh – average 1294 mAh (average 2 worst cells 1166.5 mAh).

Powerhaus 2500

Supposedly high capacity cells that are no longer driving a digital camera, but otherwise still seem to work.

Testing in March 2011 showed the inability to power the digital camera was due to one cell that was dying well before the others, combined with serious voltage sag on all cells. In particular, the voltage sag on the worst cell was so severe that it was dropping straight to 0.9 V, at which point the MH-C9000 terminates its discharge. I put the cell into a single cell LED torch and ran it for a while, then tried another discharge. The cell performed better, and didn't terminate immediately. Instead, its voltage slowly recovered then stayed steady at 0.94 V, even after 75 minutes. Voltage recovery on the other cells was about 0.2 V. A fair bit of cycling of the four cells ensued, followed by two break in runs with a 500 mA discharge in between.

Test, Discharge Current	Cell Number				Average	Variation	3 best cells
	1	2	3	4			Average	Variation
1st break in, set to 2200 mAh	1485	1969	1996	2033	1871	36.9%	1999	3.3%
Discharge 500 mA	1347	1954	1973	2035	1827	51.1%	1987	4.1%
2nd break in	1294	1961	1991	2006	1813	55.0%	1986	2.3%

That first cell is heading downhill.

These results highlight the need for testing of under performing batteries. The best 3 cells make quite a good matched set, at least for capacity. Having matched capacities doesn't do anything about fixing the seriously bad voltage sag.

Unless otherwise mentioned, capacity results are in ascending order, not in slot order. The ranking order between particular tests may vary.

Eneloop 800

Premium brand low self discharge AAA made by Sanyo.

First set of 4, manufacture date April 2006, so they're one year old.

• Voltage straight off shelf 1.30 V, average shelf capacity 568 mAh.
• Average break-in capacity 844 mAh.

8 more, manufacture date October 2006, tested September 2007; about 11 months old.

• Straight off the shelf, average voltage 1.301 V.
• Capacities straight from packet: 589, 600, 590, 602, 589, 590, 598, 607 mAh (slot order) – average 596 mAh (variation 3.1%).
• Break in: 836, 835, 844, 856, 842, 849, 839, 849 mAh (slot order) – average 844 mAh (variation 2.5%).

8 more, manufacture code 06-07 00 (July 2006), tested January 2008; 18 months old at time of testing.

• Straight off the shelf, average voltage 1.294 V – the lowest for Eneloops I've seen yet.
• Capacities straight from packet: ?, ?, ?, ?, 555, 564, 570, 570 mAh – average 564.75 mAh (variation 2.7%).
• Break in: 845, 855, 858, 860 mAh (first set), and 853, 860, 839, 857 mAh (second set, slot order) – average 853 mAh (variation 2.5%).
• Capacities @ 0.5C: ?, ?, 841, 853, 831, 850 mAh – average 844 mAh (variation 2.6%).

4 more, manufacture code 09-10 S0 (September 2010), tested March 2011; 6 months old at time of testing.

• Straight out of the packet: 557, 555, 539, 558 mAh – average 552.25 mAh (variation 3.5%).
• Break in: 796, 794, 775, 792 mAh – average 789.25 mAh (variation 2.2%).

In April 2011 I managed to retrieve 3 AAA Eneloops that were in a headlamp I'd loaned to someone for a caving trip 20 months earlier. After the caving trip they sat around doing nothing. So I discharged them at 200 mA (0.25C) to see how they fared. While their discharge (under load) voltages seemed a little low, they did very well, especially with being used for about an hour on the caving trip.

• Voltages as retrieved: 1.307, 1.308, 1.309 V (variation 0.15%).
• Capacities @ 0.25C: 560, 567, 588 mAh - average 572 mAh (variation 5.0%).

Energizer 900

The first result in a Google search for "energizer aaa 900 pathetic" (without quotes) was a page on Energizer's own web site. Pretty much sums it up. (The 6th result was a page on this site, Aqualab.)

• Break in: 258, 211 mAh (variation 22.2%).

Did some cycling to see if I could bring that up.

• Best result at 500 mA charge, 200 mA discharge was 455, 422 mAh – average 438.5 mAh (variation 7.8%). Yep, pathetic.

EniTime 800

Inexpensive brand supposedly made by GS Yuasa (a reasonably reputable Japanese brand) but actually made by Yuasa (Tianjin) Technology Ltd, China. Sold by DealExtreme here. No date code on cells.

Tested January 2013.

• Off the shelf voltage (in test order): 1.296, 1.287, 1.293, 1.300 V – average 1.294 V (variation 1.0%).
• Off the shelf capacity, 200 mA discharge: 247, 235, 243, 241 mAh – average 241.5 mAh (variation 5.1%).
• Break in capacity: 895, 911, 849, 897 mAh – average 888 mAh (variation 7.3%).
• 400mA discharge: 875, 889, 838, 878 mAh – average 870 mAh (variation 6.1%).

They have a good capacity but the variation between cells is quite high, indicating manufacturing tolerances are a bit loose. These are not Eneloops! I do not expect them to last particularly well.

The packet is marked "Ready to Use" but with such low capacities out of the packet it's a somewhat questionable claim (assuming a reasonable charge to start with). Further testing turned up a capacity for cell 2 of 920 mAh – a great capacity for an AAA cell, but at which point the LSD ability and/or the cycle life of the EniTime must surely be suffering.

Retested May 2015.

• 200mA discharge: 849, 874, 881, 901 mAh – average 876 mAh (variation 6.1%).

The four cells had been in a clock, used to power RGB LEDs on demand. Their initial recharge times varied from 75 minutes to 155 minutes. How could discharged capacities be that uneven? My guess is that their LSD ability greatly varies and some of the cells had gone much flatter than others.

Eneloop 800

These were in a headlamp loaned to a friend who didn't return the headlamp to me for 20 months. They had an unknown amount of use at the beginning of that 20 months, so the "off shelf" capacity isn't of much value, other than showing they still had a good charge. The high break in capacities show they hadn't suffered for their lack of use.

Update 8 October 2018: This next batch of three AAA Eneloop (HR-4UTG, first generation) had apparently been sitting around doing nothing for years and were resisting the idea of being used again. Basically, they would only provide only about 3 mAh thanks to some sort of high resistance problem causing bad voltage sag and very quick voltage rise on charging. They got woken up not too badly by a break in cycle.

• Break in, 0.2C (160 mA) discharge: 675, 681, 682 mAh – average 679 mAh.

Update 30 December 2018: This batch of four AAA Eneloop (HR-4UTGA, second generation) had (also) been sitting in a torch doing nothing (or not much?) for years. They gave some strange figures upon charging – much longer charge times than the amount of charge put into them, and not a high enough voltage when fully charged; it seems the MH-C9000 had reduced the charging current at some point – so I put them through a break in cycle. Three of these cells are not performing well. Time for some slow cycling.

• Break in, 0.2C (160 mA) discharge: 479, 519, 714, 473 mAh – average 546 mAh.
• Voltages at end of break in: 1.38, 1.38, 1.45, 1.38 V. There's something definitely not right here.
• Discharge at 100 mA: 585, 603, 772, 588 – average 637 mAh.
• 3rd cell at end of 400 mA charge: 790 mAh charging, 222 minutes, 1.48 V.
• The other three cells were still charging after 290 minutes: 479, 553, 471 mAh, at just 80-110 mA. The cells are sitting at 1.37-1.39 V. The MH-C9000 is winding back its charging current. Could it be because I was using a 12 V, 1 A power supply? Charge terminated.
• Discharge at 200 mA a few days later: 374, 481, 727, 402 mAh – average 496 mAh.
• Another break in, 0.2C (160 mA) discharge, using a 12 V, 4 A power supply: 800, 799, 802, 799 mAh – average 800 mAh; these cells are fine! Indeed, that's a very consistent result.

Update 21 September 2023.

• Three cells AAA Eneloop first gen break in: 697, 726, 744 mAh, average 722 mAh.

Powertech 800

Sold by Jaycar Electronics. Used regularly, seems to have quite a high self discharge. Many also have high impedance - to the point that the MH-C9000 refuses to charge 4 of the 12, and had issues with several more before eventually deciding they were OK. They haven't had several hundred charges, so at this point I cannot recommend these cells.

• Break in, 0.2C (160 mA) discharge: 603, 615, 616, 619, 651, 687 mAh – average 631.833 mAh.
• 0.5C (400mA) discharge from 16 hour charge: 563, 575, 587, 594 mAh – average 579.75 mAh. Also 615, 619 – they've gone up! Time for some cycling perhaps? These will be retested before being included in overall results.

Other testing, to see how well charger works with these cells:

• 0.5C discharge from 0.5C charge: 512, 518, 528, 544 mAh – average 525.5 mAh.
• 0.5C discharge from 1C charge: 424, 483, 497, 513 mAh – average 479.25 mAh.
• 0.5C discharge from 1C charge, left overnight: 489, 535, 541, 561 mAh – average 531.5 mAh.
• Same again after a bit of cycling: 515, 581, 608, 613 mAh – average 579.25 mAh.

Powertech 900 (untested)

Solder tag cells – sold by Jaycar Electronics. These were installed in the battery of a cellphone in May 2005, and were basically abandoned in February 2007 as being too unreliable to power the phone. They were recharged every 2-3 days on average so probably had only about 250-300 cycles, and far fewer than 200 cycles before their capacity noticeably decreased. Not good.

The two previous sets of cells each lasted over three years (a respectable 500+ cycles for each set), and while the earlier cells had a lower claimed capacity (eg, 650 mAh for the initial set), the runtime of the phone with the supposedly 900 mAh set of cells did not seem appropriately longer. So it seems as though the quality of the Powertech cells is decreasing – this is probably a direct result of the higher capacity – so once again, I cannot recommend them.