Updated: Recycle time of the FL36R with different batteries

Update: Phase 2 concluded. During testing for phase 2 it became clear that the original tests for the eneloops was faulty, and that test has been redone, yielding radically different results.

The FL36R is Olympus's middle flash, set between the low-power FL16 and FL20 and the more proffesional FL50. Compared to its bigger brother it outputs less power (GN 36 compared to GN 50) and uses half as many batteries. The latter results in slower recycle times and fewer uses per battery pack. This article looks at 5 different batteries and their effect on recycle time and amount of fires. We will see that there are in fact huge differences between the different batteries.

The contenders

There's quite a bit of variation here. Two of the batteries are slow-discharge, a relatively new technology that allows you to store rechargeable batteries without them being empty by the time you need them. One is a standard high-capacity NiMH rechargeable battery. One is a special high-power ultra-awesome lithium battery, at least according to the manufacturer. This is a battery type optimized for devices that need a lot of power. It's also the most expensive battery here, at nearly 4 EUR per battery (so 8 EUR for a set for the FL36R). The final battery is just a cheap traditional alkaline battery costing about 30 eurocents per piece. The rechargeable batteries are all rated 1.2V while the nonrechargeable are the more traditional 1.5V.

The rechargeable batteries were all a couple of months old in which time they have had about 2-5 recharge/discharge cycles. They were recharged to full using a quick charger and then left to trickle charge for another half an hour. The test was then done immediately. For each of the rechargeables I charged another set simultaneously which I'll leave lying around for a couple of weeks and then test again. This to test the slow discharge claims of the Uniross and Sanyo batteries compared to a normal NiMH. Once I've done this second phase, I will extend this article.

The test

For the purpose of this test I wrote a little program that registered a timestamp every time I pressed a button. I then put newly charged batteries in my FL36R and turned it on with default settings (which were of course the same for all tests). After that every time the "charged" led lit up I simultaneously fired the flash and pressed the button in my program¹.  This was done until the flash shut itself off for all batteries except for the cheap alkalines. For the latter I stopped the test once the recharge time went over 30 seconds consistantly, because at this point the flash unit was heating up considerably and I had been going for almost 50 minutes on that single test.

The results

Both these graphs show the same data but using a different scale. While the first one goes from 0 to 15 seconds, the bottom goes to 60 seconds. This was needed to accomodate the very different results. The Y axis shows the recycle time in milliseconds while the number on the X axis is the number of times the flash has fired at this point. For example on the top graph you can read that the 19th time they were fired the eneloops and uniross batteries took bout 8 seconds to recycle.

Phase 2

For phase two I redid the tests of the rechargeable batteries, but let them lie on the shelf for 2.5 weeks between charging and testing. This to test the self-discharge of the different battery types. The results were very interesting:

• Uniross Hybrio: The LCD panel lights up, but the flash doesn't even manage to build up a single charge. Dissapointing as this was one of the slow-discharge types.
• Energizer NiMH: These didn't even manage to light up the LCD!
• Sanyo eneloops: The eneloops performed excellently. Even after 2.5 weeks of shelving they still managed more flashes than any of the others did freshly charged. Still, they were about a second to half a second slower than they were fresh.

Observations

Well, that's a lot of interesting data and there's a number of things I want to point out.

• Both the uniross hybrio and the sanyo eneloops show similar results in recycle times, but the eneloops last a lot longer.
• The fastest rechargers are the Lithium batteries, at least at the start. Curiously after firing the flash about 95 times the recharge time shot up to around 20 seconds where it stabilized once again for a considerable time.
• A close second on the recharge time are the freshly charged NiMH batteries which kept up with the expensive single-use lithiums very well and stayed low when the latter shot up. However there's a bit of a strange bulge between 25 and 50 flashes where they went to 8-6 seconds instead of the 5-6 they held on before and after. This indicates the power output is somewhat less consistant.
• The cheap alkalines were the slowest to recycle, but also lasted the longest by far, making me end their test prematurely because it was just taking too long (and overheating the flash)
• Speaking of overheating: both the lithiums and the alkalines were scalding hot when I removed them from the flash. The rechargeable batteries were warm, but not so bad you might burn yourself.
• The rechargeable batteries kept their recycle time more consistant then their single-use relatives, stopping quite suddenly with a big spike. The alkalines were worst here, with the recycle time increasing gradually over the entire length of the tests
• The sanyo eneloops were the only ones that managed phase 2 of the tests, and did so with flying colours. Sure, they were a bit slower and lasted less long, but even so they easily beat the cheap alkalines for speed and beat everything else for duration.

Conclusions

Conclusions aren't all that clear cut and really each battery tested can be useful in its own way, depending on your needs. If you want a fast and consistant recycle time, freshly charged NiMH's seem the way to go. However you pretty much have to charge them right before you plan to use them because as we'll see in phase two (once I did that test) their performance degrades the longer they stay on the shelf.

The two slow-discharge batteries are supposed to solve that problem, but they are both slower: recycle times are roughly twice that of the two faster batteries for much of their run. Further more, the Uniross Hybrio batteries failed to fire the flash after being left on the shelf for 2.5 weeks after charging.

The eneloops kept a useful recycle time for more cycles than any other battery (even though their recycle times were slower than the best performers in the group). After 2.5 weeks, they were a bit slower and not quite as full, but still more than capable. The eneloops are probably the best combination of recycle speed, number of flashes and shelf life.

The Lithium battery seems to be a good backup for when you don't have any charged NiMH's handy. For the first 90 shots they remain slightly faster than even the NiMH, and they are more consistant. After that however their performance drops to being worse than even the cheap alkalines. While you can still push a few flashes out of them, you're probably better of replacing them. The NiMH keep going for another 30 shots about before they too drop off (and they do so quite suddenly). Still, if you really need the fastest recyle times and can afford the rather steep price of 4 EUR per battery, the lithiums are the way to go.

Finally the cheap alkalines: they may be the slowest in the test by a considerable margin, but they are also the longest lasting by far. If you need to use a flash for a long time but don't worry too greatly about recycle times, this cheap option is for you.

In short:

• For fast recycle times use freshly charged NiMH or Lithium batteries
• For long shelf life: use lithium or eneloops
• To get most usable flashes out of your set, get freshly charged eneloops
• If you really don't care about recycle time and just want cheap and long-lasting: get cheap alkalines

^1. Observant people will notice that this introduces a systematic error due to my reaction time being slightly lagged. However I'm confident that in most cases my reaction was below 0.2 seconds, and even if not they should be consistant for all tests.

^2. I am ignoring the final couple of recycles as for most batteries these are not representative