Yes. They are still available. How many would you like to have?
10 of each spring and 10 pcs 3500K sst-20s
3500k LEDs aren’t available anymore, sorry. 
I see, so it has to do with at which current the spring is at its maximum allowed power dissipation. What determines what the maximum power dissipation is of a spring?
The size, number of coils, or if it is soldered or not, and of course, its resistance.
For the large spring, without being connected to a tailcap, the max power dissipation of the spring is around 1,8W, stabilizing at a max temperature of 58C, at an ambient temp of 21C.
For the small spring, I just decided to cut the rating by a factor of 4, down to around 0,4W.
The springs can easily handle more current when soldered onto a tailcap or driver, but I’m not taking any chances.
:cry: Which LEDs are still available?
SST-20 3000k.
4 of those then since I’ve runout of the ones I got from kaidomain anyway. Plus 10 of each spring. I’m kicking myself for missing the 3500K ones. Should hang out in BLF more often.
Has anyone compared the tint on the KD 3000k SST-20 with this one?
Not yet.
I’ve been told by Mouser that the tint bin available for the 3500k was GA3, and you saw how that turned out. It did respect the chromacity bin.
However, tint bin variance being what it is, some LEDs ended up right on the BBL at 1A, and other below it.
For the 3000k, Mouser told me it was the HD2, like it was said before.
1.8Watts of heat is a lot and it sounds too much for such a small metal "object" with quite small surface area. While soldering to tailcap pcb would help a little bit, center pad of tailcap pcb is not thermally connected to tailcap metal body. From numbers above thermal resistance of that spring could be calculated : Rth=(58-21)/1.8=~20C/W, which looks too low/good.
Example of 20C/W alu heatsink, just check the size and total surface area compared to spring:
https://www.ebay.ie/itm/HEAT-SINK-TO-220-TO-3P-20-C-W-Heat-Sinks-Standard-Heat-Sinks-/292097726286
I think that what a spring can handle is a bit arbritrary. I think a BLF definition would be: at what power does it by just maintain its tension and not collaps. Since a spring in practice sheds heat both by air and via tailboard and battery, how well that goes is not easily predicted and reality must prove what the real limits are.
Yeah, I was wrong in my calculations for the large nickel plated spring.
I was doing my measurements at 13A.
That would mean:
0,010 Ohms*13A = 1,7W. Now that is a bit more reasonable.
Sorry for the error.
Heck, if we go for 9mOhms(true resistance):
0,009 Ohms* 13A = 1,5W
1,8W is the absolute limit I’ve put on the large spring.
Free convection heat dissipation in ambient air is completely different from non-moving space limited air inside a flashlight.
So, I found something in my 2018 Spring documents. 
Basically, now I remember that the initial temperature measurement was done with an infrared meter, rather than a thermal couple. So the 57C measurements was completely wrong.
Sorry for the French, but I did this back in July, and completed the table in August of 2018. So that was a long time ago.
Are the measurements done with the springs relaxed or compressed like they would be in a flashlight?
Relaxed, outside of the light.
So installed in the average light the resistance will be considerably lower and the Blue springs will not overheat and turn blue.
Yes of course. The measurements are always done with a relaxed spring.
You will never have a problem with them.
And if you want the absolute highest performance, just put the small spring in the middle of the large spring.
Nothing yet.