I agree with your point--in my opinion, AA-compatible flashlight manufacturers need to focus much more on optimizing performance on NiMH as opposed to 14500. Take the recent T3, for example: compared to its predecessor T2, it is 20% more efficient on 14500, but 40% LESS efficient on NiMH! To me, this is a horrible tradeoff because the main reason to get the T3 over S2+ is for NiMH use, since with lithium batteries, the S2+ outclasses T3 in both output and capacity, while not being much bigger.
EDIT: It is good to have the lithium compatibility as an option, but it should NEVER sacrifice NiMH performance!
The T4 was designed for multi-chemistry capability. The point is you have an option. Not sure it’s any more complicated than that. Use what you’re comfortable with.
These days we have some very capable 14500 to choose from such that they compete or surpass what is available in 16340 cells and aren’t that far under most 18350 cells. So you get a slimmer body with more length, which many people find more convenient compared to the short squat bodies of those other lights, typically. S2+ feels a whole lot larger in both pocket and in hand, plus it’s a lot heavier. So 14500, I like it a lot. More light. For me, NiMH doesn’t offer the lumens I prefer although the extended run times are great. The fact that a 14500 light takes multi-chemistry cells is just a bonus for me, in case I need to run those sometime.
But it’s like any light…there’s a point or there is no point, depending on your needs or preferences, and just buy what you like best.
A lot of people liked the T3 because of the different look and the color options…plus they weren’t aware that the driver for the T2 had been improved since the first version of it, and the tube bore was widened a little (T2 was awfully tight for many cells, sometimes requiring you to remove head and tail to poke it out for cell changes). If he offered the T3 in black I would have gotten that because I do like the new design pattern, but the blue is really nice. I have a couple of both T2 and T3 and there’s not a lot of difference other than hand feel. In normal use I don’t think you’ll notice a huge difference in run times (efficiency) between the two drivers but I guess it depends which one is in your T2. When you just turn a light on, leave it alone and graph the output, that never replicates real use for the majority of people (but it’s a helpful comparison tool).
I agree with your point, if the efficiency difference is on the order of 10-20%. Here, however, the T3 driver takes 60% more power to give the same output as T2, and I do notice the difference: there is noticeably more heat, and some of my NiMH cells could not provide enough amps to power high mode in full brightness! The same cells had no issues with T2.
Especially unacceptable to me is the fact that the new driver is supposed to be an upgrade to the old, and marketed as being MORE efficient--I should have noticed that the battery type was not specified in that statement! I think there is some miscommunication between Convoy and the factory designing and making the drivers.
An unrelated design flaw that I have not seen mentioned anywhere: the T2 and T3 heads are not waterproof! The head O-ring is way too thin to keep the glass lens tight and secure the LED board to the shelf. Without modifications, the thermal contact is terrible because there is no pressure, and the board just floats on a layer of thermal paste. I was once rubbing the lens with a cloth in a rotary motion to clean it, and the cloth made contact with the O-ring and pulled it clean out with just friction! I fortunately had a thicker O-ring with the same diameter, which fixed the issue.
These efficiency discrepancies would have been noticeable in the runtime test I did on the T3. I suspect that number you’re referencing comes from a QA failure.
Also, my t3 has been fully submerged. It’s the ti version, if that matters.
For a bit more context regarding T3 versus T2 efficiency, check out this thread. Initially I saw ZeroAir's runtime charts and current measurements, which suggests that the T3 driver is quite a bit less efficient than T2--it takes about 68% more current for the same output on high. This difference is too much to be explained away by efficacy differences between 519A and 351D, or the Ti vs Al hosts.
I couldn't find it right now, but user g_damian has a thread comparing T2 LH351D and T3 219b, both 3500K, where the T3 on high takes more than 2x the current to make less output than the T2.
I also measured my own T3 (in aluminum) and got comparable numbers. Given this sample size of 3 lights, I think it is reasonable to conclude the issue is not an occasional QC mishap--there is a problem with the driver design.
Surviving a single short-period submersion tells little about waterproofing. In real use, thermal expansion and contraction of the air inside the light could easily suck water in--this could happen if the light is run hot before encountering water. There is also the possibility of encountering pressurized water jets, or submersion to a depth where pressure becomes significant. If you remove the MCPCB, you will see that the thermal paste is still a relatively nice, thick blob rather than completely smeared out by the pressure from a tight, proper fitment.
I have Convoy T2-4 mode with samsung LH351D 4000k led and I am very satisfied with it. My friend has Convoy T3 with same led and with 13 mode driver.
On nimh T3 give slightly more light on 100% mode than T2 and use higher current from battery - T3 use 1.3 amps and T2 about 1 amps.
When use 14500 battery than I have a case that T2 use slightly higher current. This results are not precise because I use old multimeter with high burden voltage but it is enough to see diference.
My Convoy T2 give more light on 1% mode when use NIMH battery in relation to the case when I use 14500 battery on 1% and it is also one of the illogicalities.
My friend have T2 with XPG2 led and it have better battery life than version with LH351D and it has more throw but light tint is worse.
Unfortunately no, not only the high mode is involved.
Even at medium mode, the difference is drastic.
I hope they get a fixed driver out soon.
It would have been enough to write a new firmware on the T2 driver, I don’t understand why it was good to design a completely new circuit.
The above graphs are where I think a re-test would be called for. They don’t match my numbers on the Ti T3. It’s plausible there are driver versions — although that would be annoying in its own way. There are dimensional differences on the Ti also which would explain why it is fully waterproof while your Al T3 is not.
First, need to be referencing specific drivers, not light models (this, because although the T3 was introduced with just one, the T2 has had two different drivers and most of the reviews out there are the original release). Second, no offense intended at all, but Zeroair’s graphs don’t tell us much really. Better to do an electrical test or at least use the same exact battery and conditions. Runtime and lumens is a good reference but it definitely does not tell all. That said, the little I read some years back about how he tests lights may be outdated, who knows…if he’s got more going on in testing methods it doesn’t show in reviews and he still has that old disclaimer in each one (and all of that is fine…he gives great reviews and spends a ton of time doing that for us!).
My T2s and T3s have that thinner o-ring and there’s no rattling of the glass on any of them…haven’t tested for water resistance. On that note, though, Simon has never gone for high water resistance and I’m not sure he even tests them anymore……and on that note you know that all of the IP “ratings” are done with voluntary tests and I’d bet that only a couple-few of light manufacturers actually send their models off for testing or have bothered to try to make testing rigs that meet the suggested criteria. Lots of IP68 lights have leaked when dunked…lots of untruths and copy/paste or “assumptions” in flashlight marketing. Aluminum alloy, drop resistance, etc. etc. Much of it is hooey.
Those thinner o-rings are kind of a pain, though. I have some other lights with similarly skinny ones and they tear easily and can be difficult to get seated so they don’t shift under tightening pressure. If a thicker ring fits, great, otherwise if not all the lights do that then I’d suspect slight defects in machining the threading or length of the bezel/head/tailcap/whatever (or anything in the stack, such as a thinner mcpcb or gasket, etc).
Would be nice to see a good standardized test of the drivers from someone who has the equipment and time. Same current or normalization for noted differences there, same emitter, same battery, same host, all that jazz. My anecdotal input is that in real use I just don’t see any drastic changes in run times (my use tends to be varied, using all modes for whatever length of time and generally with these for brief periods where voltage can rebound…do use high modes a lot, though).
I agree with your assessment, but I do have more faith in current measurements. It actually surprises me that your measured only 1.3A; I've measured more than 2 amps.
I am starting to think there might be an alternate explanation, which is that the efficiency of the driver depends on the electrical characteristics of the LED, e.g., forward voltage. I've swapped the 2700K 519A for a 5700K 519A, which dropped current from 2+ amps to around 1.6. My guess is that the current is even less with LH351D, and perhaps the driver is just more efficient for high Vf LEDs. I might test this conjecture one day by swapping in an LH351D into my T3 and see what happens.
There is quite a bit of confounding between the old and new T2 versions, but the reviews do make the precise version abundantly clear.
My concern with the thin O-ring is more than water ingress. When the head is assembled, there is not enough pressure to secure the LED board to the shelf, resulting in terrible thermal contact as the board is practically floating on a blob of thermal paste. I worry that the LED may age prematurely if one does extended 14500 runs on the highest mode.
I would like your help in choosing a flashlight ConvoyT4?
I had the opportunity to try that flashlight with diodes 519A 4000 k and with sst-20 3000k which have my friend and another friend have T4 with a samsung LH351D 4000k that he bought because he liked the color of the light when he tried my Convoy T2 with that diode. The only flashlight I have with 2 AA batteries is the Zanflare F2 with Cree XP-G2 4500k and I am very satisfied with it, with the autonomy, color and range of light, but I would like something of the same format but with more lumens and with at least an approximate or greater range of light. Samsung LH351D has a good light color but it is floody and I am looking for flashlight with better throw. It seems to me during the test that the sst-20 has a better range in throw than the 519A, but the 3000K is too yellow for me and a friend told me that the 519A heats up more than the other two diodes and that it consumes the batteries faster. My question is which diode to take then, should I go for sst-20 4000K or go to loew cri 5000K or 6500K in order to get a longer range and a higher intensity of light. I have no experience with those diodes and the color of the light they give and if the low cri sst 20 have a light that pulls on green? It is not important to me that the diode is high cri, but it is important to me that it has a solid autonomy and that it is efficient.
The 519A draws the same current as the SST20 on the T4 flashlight.
[/quote]
You’re right! My friend made a mistake and I also made a mistake in taking his impressions for granted and not testing it myself. He is a banker and does not understand multimeters and he tested the lamps by charging the batteries to the end and then turning it on to work until the battery is exhausted and measuring the operating time. He compared the light intensity using a phone application that measures the reflected light from the wall. The mistake was that in one lamp there were batteries from Simon’s store that came with the lamp and in the other there were Xtar 800mah batteries with protection. At 100% mode with full batteries both lamps have a tailcap current of about 1.32 - 1.35 amps which I measured with an Aneng 8009 multimeter. However, the lamp with 519A and Xtar batteries works for 24 minutes until switching off, and the lamp with SST-20 and Simon batteries works for 36 minutes until switching off. I put the batteries in the liitokala lii 500 charger and started the normal test with a discharge current of 250 ma and determined that the Simon batteries have a capacity of 809 and 817 mah and the Xtar batteries have 805 and 510 mah. so one battery was damaged and that was the reason for the shorter working time and the wrong conclusion. When the lamp went out, I immediately took out the batteries and measured the voltages and got the following values: Simon’s batteries have about 3 volts when measured, and Xtar one has 2.5 volts and the other Xtar has 3.6 volts, which means that in the worse battery the protection reacted and turned off the circuit. I apologize to the members of the forum if I gave unverified information in the previous post and misled someone.
This is amateur testing, so the results can sometimes be unexpected compared to laboratory conditions.
Anyway, it isn’t that simon made the T4 work with AA alkaline or Eneloop, it always did. It didn’t work with Primary Lithium batteries, and it now does.
