I agree that the Convoy numbers don't make sense, but not because of specs from Simon.
According to the numbers above, the Convoy drain is above 2mA at all times, compared to Anduril bright button at 1.5mA. Despite the higher drain, the drop in voltage is almost 3x lower.
Another reason the numbers don't make sense to me: let's assume the best case, that the Convoy drains only 2.74mA constantly. Then over a three-week period, the amount of capacity drained is at least 2.74*24*7*3=1380mAh. This is between 1/3 and 1/4 of the total capacity of a 21700, which should drop the voltage much more than 0.067V.
I feel that I might be misunderstanding some parts of your post, but am unsure--any clarification would be appreciated!
BTW, my S2+ with blue rubber button always drains less than 2mA; below around 3.9V, the drain is low enough to be undetectable by my multimeter with 1mA resolution. My orange button drains less than 4mA.
Thanks for the comments. Another set of “raw” numbers - from 8/27 to 9/21:
Sofirn Bright drops 1.5*24*7*3=756mAh
Voltage drop 4.192 - 4.028=0.164v
Does this one make sense?
A couple things I’m thinking:
first is that the voltage drop is not linear - very flat at top,
second, the blue light draw in my S21D is neither of Simon’s spec, and less than 2.74mA. Not as easy to measure but I’ll try later.
third, thanks to your number 2mA, does it mean Simon’s specs below on Aliexpress are just wrong? I maybe should ask him.
fourth and probably most important, I possibly made a mistake and need to run test again. In case you have any interest, maybe run the test yourself for 1 week.
(Metal button + adapter + 20mm lighted switch) kit for for M21A M21B M21C M26C L21A S12 S16 etc.
When the battery is fully charged and the voltage is 4.2V
Blue light consumption current 9.4mA,
Orange light consumption current 14mA,
When the battery is low and the voltage is 3.0V
Blue light consumption current 2.74mA,
Orange light consumption current 6.8mA,
The new Sofirn numbers you just posted seem very reasonable. It is helpful to reference this mAh loss versus voltage drop plot of a Sofirn 21700, taken from HKJ's review.
I do agree that the voltage drop is highly nonlinear, but even with the flat top of the curve, losing 1380mAh would easily sink 0.2V, which is way off from 0.067V!
Ok, I think I understand my confusion now--the current draw from post #100 are Simon's numbers, not yours! I would love to see what kind of numbers you get with your test equipment. My meter only has 1mA precision, which is as good as none at this scale!
My numbers certainly disagree greatly from Simon's, but I couldn't say for sure that his are wrong for all lights. My setup is S2+ running 4-mode 5A driver, and it is known that driver choice greatly influences current draw. With a 17mm ramping driver instead, the drain is so low that my meter cannot register it at all, and the tailcap becomes much, much dimmer. Maybe there is some driver that causes faster drain, but I don't know. I feel that Simon is really erring on the side of caution with his numbers, which I do appreciate. It's better to expect hazardously high levels of current drain and get low drain in reality, than the other way around.
Finally did the measurement with this end switch arrangement and didn’t blow up anything (the end switch thing is the only reason why I haven’t measured it). 4 volt Sofirn battery with both S21D and M21B, blue metal switch. I have one lead on spring, other lead on battery rear, then press really (really) hard of end of cap to body of light for the blue light to go on. I hope that’s correct.
I think I found an explanation for why my empirical voltage drop is so low: the Convoy blue metal switch’s current draw is very low. So low that I have to have my wife take a picture for proof . Anyone with similar light if you’re interested please check to confirm my results.
Assuming I have done test correctly (I think so), the blue light metal switch in M21B and S21D is a miser, running about 400-500 µA, still higher than Anduril Dim (~100 µA), but nowhere near specs posted on Aliexpress of 9.4 mA. 9.4 mA is very bad and this error puzzles me - has anyone discussed this?
Thank you so much cannga for the data! Your testing method looks right, and the numbers are believable. 395 microamps over 3 weeks is 100mAh, which should result in less than 0.05V drop. The fact that you measured more voltage drop suggests that the loss from lighted tail is on the same order of magnitude as loss from self-discharge, which is really good news!
As far as I know nobody has discussed the discrepancy between Convoy's numbers and the actual current draw. Maybe we could make a post in the Convoy thread about this discovery?
Thanks. At the least Simon’s 9.4 mA is way off no? 9.4mA x 7 x 24 = 1579 mAh per week - is my math correct?
Most definitely my Convoys don’t go into LVP mode for at least a couple of months, let alone 3-4 weeks.
That 9.4 mA spec is actually the reason I included Convoy in this comparsion - it has bothered me.
Hello Argo, I was reading one of your reviews the other day and could EASILY tell you are a seasoned shutterbug and pixel peeper (like me ). That was the first time I noticed that with any flashlight reviewers.
Thanks for the very nice photo comparison.
Your math is correct, and I'm inclined to believe that Simon's numbers are way off. The only conceivable reason the numbers might possibly be right is that there might be some driver/switch combination that drives the tailcap really hard, perhaps to counter light loss from a metal switch (as opposed to silicone). But I don't see any evidence that such a combination exists.
I think it is a good idea to ask for a clarification. 9.4mA looks really bad, it's more current than the 0.1% mode on his 5A driver, which consumes 8mA at full battery.
I vaguely remember trying to measure the lighted tailcap current years ago by hooking it up directly to the battery, without the rest of the light, because it is awkward to perform the measurement with the battery still in the light. I noticed that the tailcap was way brighter than it was while on the flashlight, and immediately stopped the test for fear of burning it out. I think I did get a number around 10mA. I think the driver does perform a role in reducing the current that goes to the tailcap LEDs.
Also, I have an S2+ tailcap from 2018 (not bought from Simon), and upon disassembling it I noticed that the epoxy casing on the SMD blue LEDs has turned a bit yellow, presumably from 6 years of having blue light continuously shined on it. It is also noticeably dimmer than a brand-new lighted switch on the same light/driver combo. I couldn't tell if the loss of brightness is from absorption by the yellowed casing, or if the LEDs themselves have degraded. I mean, 6 years is 52000+ hours, which surpasses the rated lifetime of some LEDs.
EDIT: I mean 4 years (35000hrs) from 2018 to 2022, not 6...maybe I should consider dropping my math major...
I just received my Wurkkos TS21 today. So far, I’m loving the light. The only issue I have is that when set to turbo the light seems to step down rather quickly. And by quickly, I mean it steps down after about 30 seconds or so. The thermal limit seems to be too low. I tried to raise the thermal step down temp. But, It is still stepping down way too quickly. Either, I’m doing it wrong, or something else is not correct. Can anyone advise me on how exactly to set the thermal step down higher? This is how I tried to set the thermal step down. I clicked 3 times from off, to give me a battery check. 2C for a temp check. 7H release after the first flash. I set the current temp to 1. After that, 7H released after 2 flashes. And then I clicked 40 times. So, 40 celsius plus 30 celsius equals 158F. But, the flashlight still steps down after about 30 seconds or so. I try setting it to 25 clicks. But it is still stepping down way too early for my liking. I measured the temps on the head and body. And it is stepping down at around 106 degrees on the head, and about 96/98 on the body. It feels slightly warm to the touch. Any ideas about what am I doing wrong?
Sounds like you did the setting correctly. And yes even at 40 click (70° C) the Wurkkos still steps down fairly rapidly, to about half brightness after 30 seconds.
A few suggestions - let’s retry 30 clicks (Tmax = 60° C): 1. Repeat Tmax setting: after second flash, 30 clicks. (After your 30 clicks, just let go and don’t click again right?) Afterwards let light cool (don’t use it) for a few minutes, then recheck temp of light. It should still read room temp around 21 C. If it reads around 30 C, you probably have made a mistake and set temp, instead of Tmax. 2. Now the checking: Turn on light at Turbo level and leave it there for about 10 minutes. At 10 minute mark check at head of light, it should read about 50 C. 3. Now turn light off and immediately (very quickly) 3 clicks then 2 clicks to check temp. It should read about 55-58° C. It decreases rapidly so you need to check it very quickly.
In general and roughly, I’ve found a. external temp (with the IR gun) at the head reads about 10 degrees below Tmax b. internal temp check reads about 3-6 degrees below Tmax, but you have to check very quickly right after.
Another example, if you have set Tmax to 70° C correctly, external temp should read about 60° C, and internal temp should read about 65° C. Hope this helps.
I’ve edited the first post of this thread to more clearly indicates this light’s heat handling is not the best and if you like sustained brightness or like to use your light at Turbo level, perhaps TS21 is not the right light. Mainly I think of TS21 as being for hobbyists who want to “collect” the Nichia 219c LED which is outstanding, or for those who use light at dimmer level than Turbo. IMHO.
Repost from my Convoy thread, probably even more useful here.
Turbo run-down to eventual brightness: ceiling bounce percentage of max brightness that each light settles to after a Turbo start: 1. Convoy Headlamp H2 w/ Nichia 219b 4500k: 61% 2. Convoy S21D w/ Nichia 219b 4500k and 519a: 37-40% 3. Convoy M21E, F, and B w/ Getian GT-FC40: 35-37% 4. Nitecore MH12S w/ SST40: 33% 5. Fenix Headlamp HM65R-T w/ SST40: 32% 6. Emisar D4V2 with Nichia 219b 4500k w/ 70° Tmax : 30% (int temp 64° C, ext temp 59° C, cannot hold w/ bare hand) 7. Emisar D4V2 with Nichia 219b 4500k w/ 60° Tmax : 23% 8. Wurkkos TS21 w/ 70° Tmax: 22% 9. Wurkkos TS21 w/ 60° Tmax: 16% 10. Wurkkos TS30S w/ SBT90.2: 16% 11. Coast XPH30R Headlamp: 15% 12. Emisar D4V2 with Nichia 219b w/ 45° Tmax: 14% 13. Wurkkos TS21 w/ 45° Tmax: 8% my poor Wurkkos, at least the LED is a gem
I will your suggestion. Thank you for the advice. I should mention that I tried raising the thermal max temp, on my Sofirn SP36-Pro. And I am having the same issue. It steps down way too early. After only a few seconds. I wanted to see if it was something I was not doing correctly on my Wurkkos. So, not sure why on the bigger Sofirn light it still steps down too early. But, I’ll try your suggestion. Thank you.
You’re very welcome. Anduril temp configuration in particular that first flash is very difficult at first because it coincides with flashes from button pushes number 4, 5, 6 or so. Also keep in mind ALL flashlights will start stepping down immediately when you start at Turbo and many (except some Convoy’s for example) will start stepping down at less than Turbo also.
The only difference among lights is how fast the step-down (not whether it will happen right away) and it is true the TS21 is among the worst flashlights when it comes to stepping down. Increasing the Tmax of TS21 will slow the stepdown a little, but it will not stop it.
The graph below has runtimes of different flashlights and shows the ubiquitous (thank you spell check lol) stepdowns. TS21 would look like it’s falling down a cliff, the left most light blue one.
Just saw this post on Reddit, which suggests that the lighted tailcap pulls way more current on the 12-group 5A driver compared to the 4-mode 5A driver.
I had 2 lights both running 4-mode 5A driver, one with orange tailcap and the other blue, and the orange looked easily 3x brighter. In the Reddit post it's the other way around, so 9mA would be a reasonable estimate.
Simon's numbers are probably right for the 12-group 5A driver. Makes me very happy that I have the 4-mode version...
Hi QR, you are not talking about S21D right? My S21D w/ 12 group UI still looks very much like only 1mA (ie Simon’s number is 9x too high).
I just finished a second battery voltage-drop test to confirm my current measurement: from 10/2 to 10/22, S21D’s voltage for light w/ blue button went from 4.178 to 4.083 = 0.095v.
That drop amount is appropriately in between a Sofirn IF25a set at bright button light (drop=0.125v) and a second Sofirn IF25a set at dim (drop=0.038v). All 3 lights were tested with Samsung 40T battery fully charged. Funny thing is I’ve removed all button light from my Convoy metal button anyway because I have too many lights and don’t want to worry about any drop.
The lights in the reddit post are C8's, not S21D. They definitely use different drivers, since the driver diameters are different.
Based on the end voltage of 4.083, some rough estimates off the 40T discharge curve suggests that your S21D blue switch draws substantially lower than 1mA on average during this run, maybe closer to 0.5mA. But I cannot be sure, since the discharge curve starts at a lower voltage than your 4.178.
The important data points are the beginning and end voltages, not the difference between the two--the same voltage drop can correspond to completely different capacity loss depending on where you are on the discharge curve.
. Anyone with similar light if you’re interested please check to confirm my results.
my poor Wurkkos, at least the LED is a gem
/Samsung%20INR21700-40T%204000mAh%20(Cyan)-Capacity.png)
/Sofirn%2021700%205000mAh%20(Black)-Capacity.png){kind=link}