I see what you did thereâŠ
Yep I hear you it is good to view more as tools than toys and good to check your tools now and then.
I have too many as toys.
Using distilled water and letting it dry throughly after the test would reduce the chances of a short dramatically, if you plan on doing this in the future. Without dissolved salts in water, water is a fairly decent insulator. 
Frozen hamburgers, bite your tongue!
I do this to test the thermal step down, but under slow running water and position the light in a way that the water just hits the fins near the mcpcb shelf and not on any thread or buttons
Same effect as submerging without risking water ingress.
According to the numbers the water pressure at 1M depth for ip67 ratings is just 9.8kpa, which is probably less than the negative pressure created from air cooling down.
Back of the envelopeâŠ
If the air inside is heated to 333K (60C), and is fully equalized to the pressure outside, then is cooled to 293K (20C), then the pressure drops to 88% of atmosphere, which is about â12kpa.
So, yeah, the negative pressure might be greater than the water pressure, but thatâs assuming the hot air escaped when the light was hot. The two pressures combined would be equivalent of about 2 meters depth.
If the light can take 2 meters depth without any heating/cooling, then it should be able to take 1 meter depth when being submerged hot.
IMO, it doesnât sound like much of a torture test. The light should be able to survive a water test, hot or not. It seems like it mostly did survive, with the exception of the switch.
1 Thank
Interesting observation. So you are saying IP67 rating is cold only. So when the light is on while under water, a vacuum may be created literally sucking the water in?
Well, it shouldnât suck water in, any more than regular water pressure would do to a cold light. But, it could add the equivalent of 1 meter of additional pressure on the light. IOW, a light rated for 2 meters, might only hold up to 1 meter of water if it was hot when you dropped it in the lake.
In this case, the light was just in a glass, so essentially at 0 meters. It might be at the equivalent of 1 meter depth.
Interesting, now you both have me wondering about this and I needed a brain teaser. It has been ages since I have done analysis, but I'm guessing the science has not changed much and I am also sure someone will correct me if I am wrong, so below is my estimate which agrees with yours.
"Absolute gas pressure" (P) would change proportionally with temperature (T) changes when using Robert Boyle's ideal gas Law (PV = nRT) and disregarding any minor effects from changes in volume (V), number of moles (n) or the universal gas constant (R).
"T" is measured in degrees Kelvin (degrees Kelvin = degrees Celsius + 273.15Âș) and using an atmospheric pressure at sea level of 101.325 kPa (~14.7 PSI) with an ambient temperature of 21 degrees Celsius (~70 degrees Fahrenheit) calculates out to:
101.325 kPa / (21Âș C + 273.15Âș) = (101.325 kPa + 9.81 kPA) / (Kelvin Final Temperature)
"Kelvin Final Temperature" = 322.63Âș which is (322.63Âș - 273.15Âș) or ~49.5Âș Celsius
Temperature difference = 49.5Âș C - 21Âș C = 28.5Âș C
Celsius degrees are 9/5 larger than Fahrenheit degrees so 28.5Âș C = (28.5Âș C x 9) / 5 = 51.3Âș F
Also, the results of this calculation using a negative instead of positive 9.81 kPa pressure change are approximately the same.
In summary, if no gases leak in or out, a pressure change of 9.81 kpa (~1.42 PSI) would occur in an enclosed volume from a temperature change of ~28.5Âș C (~51.3Âș F). An increased temperature would cause a higher pressure and a decreased temperature would cause a lower pressure.
Now itâs starting to sound like quality control and not excessive pressure.
Well, itâs at the borderline, if we assume he put a hot light (with equalized pressure) into cool water. The equivalent depth of pressure (due to the air pressure drop inside the light) would be about 1 meter, which is the IPx7 rating.
So, I dunno. Personally, I donât think it should have leaked. But, itâs getting close to the limit. So, I wouldnât blame the light. Itâs a good thing to know, though.
So how many inches under water in a glass = 1 meter ?
I think some guys would be very happy with this math 
Took me a few seconds to understand the humor but I got it.
But this thread made me aware that if you accidentally drop a "hot" IPX7 rated light into a shallow puddle of water, it could suck in some water if you do not take it out quickly before it cools down.
Thanks for the deep analysis gentlemen. For added datum point, the light was submerged in roughly 8cm of water. Went in warm, came out warm after heating up the cold tap water. So there was certainly a cool down and warm up cycle submerged. I would estimate that 10cc of water ingressed (Shocked at the amount). First power up was basically uncontrollable. Modes and levels were changing without touching the button. Now it is basically controllable, but it is difficult to change any settings due to unpredictable button behavior. The button is still touch sensitive after many days in the rice bag fully opened up. Thinking it is time to reorder, and chalk this one up as spare parts. Worst case at least the battery tube doubles as a charger and cell holderâŠ
Edit: Reorder placed, $124 lesson learned.
NOCTIGON K9.3 DUAL CHANNEL 7400LM 21700 LED FLASHLIGHT
Color
Dark Grey
Switch back light color
Amber
CH1 (9 inner LEDs)
Warm White - SST20 2700K 95CRI
CH2 (3 outer LEDs)
SST-20 Deep Red, 660nm
Anytime Iâve dunked my cell phones (before I could afford properly-rated ones) I immediately remove the batteries, drain the capacitors, and put it in some kind of desiccant (dry brown rice or moisture absorbent packs) and leave it for a few days before firing it back up. It usually works, but the last one killed my Note 4 after a kayaking incident so I got a IP68 Note 8. Itâs gone swimming more than once and no issues.
For flashlights, they are only as good as the person assembling them. Grease and o-rings can be omitted, the wrong o-ring can get in by accident, or it can get damaged during assembly. Maybe a poorly machined part that doesnât mate properly? Generally, reputed manufacturers are trustworthy with this, test, and rate their lights properly, while someâŠnot so much. Like with lumens or throw, itâs easy to stick a IP rating on a box when it hasnât been properly tested or rated.
While Iâm not an expert on Ingress Protection ratings, IP67 would only be a water resistance test where jets of water or a certain size would be focused on the light and it would have to survive that for a set period of time like 1/2hr. Water and dust âresistanceâ is way different to the IPX-8 type âwaterproofâ rating where the test involves submersion into 1m+ of water for a certain amount of hoursâŠI think itâs 8 hours or something. Then after being certified the manufacturer states the exact depth say up to 2meters. I would NEVER submerge an IP-67 product into water, itâs asking for trouble, would I take it out and use it in the rain, absolutelyâŠthatâs what its rated to do.
Youâre wrong. Wiki
Test for IPX7:
Test duration: 30 minutes.
Tested with the lowest point of the enclosure 1,000 mm (39 in) below the surface of the water, or the highest point 150 mm (5.9 in) below the surface, whichever is deeper.
Exactly, and IPX8 has a kind of floating set of specifications.
Where the manufacturer puts in his/her own variables.
Wheneven you read IPX8, it should always be followed by: X duration and Y depth.
You my friend need to put a mild solvent in there to try and clean some of the dissolved ions off of the switch. You will need to take the switch cover off. Try Isopropyl Alcohol, must be really pure like 99%, it works great.
Just want to mention I dropped my green D4S in a sewage pipe one time, and had to fish it out. It was an interesting experienceâŠ
Mark, I have some 99.9% iso, switch cover is removed. You are suggesting a flush over the switch area then back to the rice bag?