Zebralight and Armytek: Thermal regulation compared

There doesn't seem to be much testing comparing the thermal regulation on these two popular brands so, I figured I would run some tests and see how their thermal regulation performs.

The lights compared are the Zebralight H600Fd MKIII and the Armytek Prime Pro C2 XHP35(Warm).

Keep in mind these are very different lights. They use different emitters and their weight and output are both very different. I wish I had an SC63w as it would be a much better light to compare to the Prime Pro as it uses the same emitter and has a similar output but, I don't. This is just a quick test and I may do a more thorough one and/or add more data/tests at a later date. This test is just for fun and I am just learning how to use the equipment.

Lights were compared with an Extech HD450 data logging light meter in a 2x2x8' closet freshly painted flat white. Both lights were ran on brand new Sanyo NCR18650GA 3500mAh cells that were hot off the charger and had similar internal resistance(25-26milliohms measured by the Xtar VP4 Dragon charger). Lights were both ran on their highest modes(H1 and Turbo 2). No cooling was used for any of the tests. Ambient room temperature was 23C(+/-1C) for both tests. Disregard the first and last 5 seconds or so as that is when I opened and closed the door and started and stopped the equipment.

I just ran a couple of 15 minute runtime tests to compare for starters. You can easily tell when the thermal regulation kicks in and the light starts stepping down rapidly. You can also get an idea of how each lights regulation performs on its highest output if you look at the flat part of the line before the thermal regulation kicks in.

The vertical numbers on the left of the graph are Lux(Brightness) and the horizontal numbers at the bottom are the Time(Eastern Standard). The Lux reading at the top is just where the cursor was when I captured the image and should be disregarded.

First up is the Zebralight. These are the stock graphs.



And, the Armytek.



These are the same graphs zoomed in.

Zebralight.



Armytek.



Here is a closeup of the regulation before the thermal regulation kicks in( I tried to keep the scales close to the same).

Zebralight.



Armytek.

Test #2: 30 minute runtime test this time. Lights tested on H1(ZL) and Turbo 2(AT) again. Same ambient temperature(23C) and cells(Sanyo NCR18650GA).

Stock Graphs

Zebralight.



Armytek.



Zoomed graphs

Zebralight.



Armytek.

Graphs updated. I figured out how to perfectly match the scale.

Very nice, thanks.
I wonder what this zig-zagging regulation on the Armytek is about. I assume there has been no cooling?

Tachead thanks for that! I have 14 AT's but only 2 ZL's and have never really thought about comparing their thermal regulation. It is good to see that my expensive AT's actually have thermal regulation as advertised. I do not own the particular ZL you tested I am assuming regulates similar to my SC600 MkIII HI.....

Connor said:

Very nice, thanks.
I wonder what this zig-zagging regulation on the Armytek is about. I assume there has been no cooling?

Click to expand...

No problem:thumbsup:.

I'm guessing the zig-zagging is because the Armytek uses a more traditional separate emitter board and driver design. The boards are separated by an aluminum bulkhead machined into the head part of the 2 piece body. Zebralight mounts all their electronics on the same board as the emitter and use a 1 piece body. This should make the thermal regulation more accurate and less prone to overshooting(what appears to be happening on the AT) because the temperature sensor and all the heat sources are together on the same board. Zebralight also uses smaller steps so it is less noticeable when it is adjusting the output. The advantage of Armytek's system however, is it is much less aggressive and tries to give you the absolute max brightness that temperatures will allow. It keeps max output much longer before the thermal regulation kicks in and keeps a higher output for much longer before it equalizes. Keep in mind too that the Armytek is running a 6 or 12V emitter and is being driven much harder(higher output) as well. So, it would stay brighter even longer if it was outputting a similar output to the ZL with the same emitter. The Armytek does get much hotter though due to this more aggressive thermal regulation. The battery was close to 10C warmer after the 30 minute runtime tests. Its interesting to see the differences. Overall, ZL has the more advanced thermal regulation imo. But, it is also much more conservative so there are pros and cons to both systems. Both systems have extremely flat regulation after the thermal regulation equalizes as well.

No cooling was used during any of these tests. I may do some more tests with cooling added.

I am glad my setup worked so well considering it was my first go at this kind of testing and I am still learning. This setup appears to be very accurate and repeatable as you can see because both the 15 minute and 30 minute tests show almost exactly the same results. I still have a lot to learn though and look forward to further testing. Please don't hesitate to give me any suggestions and/or constructive criticisms.

CelticCross74 said:

Tachead thanks for that! I have 14 AT's but only 2 ZL's and have never really thought about comparing their thermal regulation. It is good to see that my expensive AT's actually have thermal regulation as advertised. I do not own the particular ZL you tested I am assuming regulates similar to my SC600 MkIII HI.....

Click to expand...

No problem. I had always wanted to see how they performed. Yes, your SC600w MKIII HI should regulate similarly but, it has a lot more heatsinking due to the heftier body so, it will regulate less aggressively and hold its output a bit longer.

Are you using the stock thermal settings on the Zebralight, or have you adjusted it up or down? Does the Armytek offer this feature? I'd like to see the Zebralight maxed out (+5° C) against the Armytek.

eraursls1984 said:

Are you using the stock thermal settings on the Zebralight, or have you adjusted it up or down? Does the Armytek offer this feature? I'd like to see the Zebralight maxed out (+5° C) against the Armytek.

Click to expand...

Yes, stock PID settings and my light is one of the early ones so is likely from the earlier production with higher stock PID temperatures then the newer lights. No, the Armytek doesn't offer user adjustable thermal regulation. Adjusting the PID settings only makes a very small difference(especially from stock to +5). Here is a comparison of the SC600 MKIII courtesy of Selfbuilt...



The smaller ZL's like the SC63, H600, etc. will likely make even less of a difference because they have so much less mass and heatsinking.

wow Tachead! I feel even better about the cash I dropped on my MkIII now...

Connor said:

Very nice, thanks.
I wonder what this zig-zagging regulation on the Armytek is about. I assume there has been no cooling?

Click to expand...

That's a classic underdamped PID control response, also known as "ringing." Not something the general public ever deals with, but a very classic case in control systems.

Tachead might be right about the driver layout being the cause, making it harder to tune the PID, but it is usually possible to tune a controller to a very low level of ringing.

I suspected it shouldn't be like this. Thanks for the explanantion! :thumbsup:

Connor said:

I suspected it shouldn't be like this. Thanks for the explanantion! :thumbsup:

Click to expand...

In this case, there's not much major consequence. The output is less consistent than ideal, but it sounds like the temperature still stays in the appropriate range, and you still get more output and better over-temperature protection than just using a timed stepdown.

Here is an example of Armytek's thermal regulation on one of their larger models. This is a 1 hour 48 minute runtime graph of the Viking Pro V3 XP-L(Warm) on its highest output mode(Turbo 2/1050 lumen).

Test Description

Ambient room temperature 22C

No cooling

Battery - Hot off the charger Panasonic/Sanyo NCR18650GA 3500mAh

Voltage at end of test 3.03V

Lux on the left(vertical) real time on the bottom(horizontal)