I’m one of those who thinks the depresson is a bad idea. (but of course, I have no evidence of this.) If I were to design a star, I’d put the depressions on the other side of the board where the +ve and -ve contacts are. I’ll try to find those graphs over at the evil forum.
What graphs show lapping the IS board reduces output? Mine didn't. The graphs were virtually the same, definitely within equipment tolerances. Here's a mouseover:
Note: Both of those tests were before I reflowed the emitter to fix the bonding problem.
I haven't seen the results posted on CPF yet.
With all the circumstances mentioned, it's the "Test Results Data" that showed lower lumen output when lapped. You made it clear that is all in tolerance levels but it can't be denied that physically one can go on the review page and read the "Test Results Data" and see the "Lapped" ones are slightly lower.
What it is an interesting that a reflow issues does not affect the output much and if we are to take the below that into consideration it's only the equipment's sensitivity and tolerance that made the small differences.
Current Stock Lapped Reflowed
2A 796 781 781
3A 1080 1062 1075
4A 1304 1287 1322
I disagree, you're omitting some critical data in your table; The 6A range. A DCB star's primary benefit is when the power is turned up.
| Current | Stock | Lapped | Reflowed |
| 2A | 796 | 781 | 781 |
| 3A | 1080 | 1062 | 1075 |
| 4A | 1304 | 1287 | 1322 |
| 6A | 1582 | 1590 | 1754 |
The difference at 6A is definitely not measurement tolerances.
Another mouseover:
I think the lapping also shows a clear and definite decrease in temps……which was the point :bigsmile:
I didn't look over the 6A since it's not something I'd use efficiency wise. Consumption and heat vs lumen output I would stop at 4A.
Using copper MPCBs and with proper lapping, what’s the current needed to put the XM-L2 U2 to its highest lumens output (before it starts to decline)?
The XM-L2 option for the Shocker are T6's though 
.
Is it possible with a 5 amps tailcap measurement on a Shocker to get 4,500 lumens? 
I'm think'n it is....
Then you would get an approx. 200 lumen gain over a standard aluminum star, based on my results. That's good, but not where direct-bonded stars really start showing their wares; after aluminum stars begin to dip the other way.
If efficiency is the goal, definitely avoid hard driven emitters, and opt for more emitters at a lower current. For example, my TF AK-90 12xXM-L delivered 3500 OTF lumens with 38 Watts input (including driver losses). That's 92 OTF lumens/Watt, which I'll say is not easy to do at 3500 lumens.
Here's a graph of lumens and lumens/Watt from the Noctigon XM16 test.
| I-O Noctigon XM 16mm Star with XM-L2 U2 1C Emitter (Lapped) | |
| Lumens | lm/W |
| 96 | 174.0 |
| 184 | 161.0 |
| 274 | 156.3 |
| 357 | 149.6 |
| 436 | 143.8 |
| 510 | 138.6 |
| 585 | 134.4 |
| 657 | 130.3 |
| 725 | 126.2 |
| 793 | 122.7 |
| 857 | 119.2 |
| 918 | 115.9 |
| 981 | 113.3 |
| 1037 | 110.2 |
| 1093 | 107.5 |
| 1147 | 104.8 |
| 1202 | 102.5 |
| 1252 | 100.0 |
| 1301 | 97.6 |
| 1349 | 95.6 |
| 1395 | 93.3 |
| 1440 | 91.4 |
| 1484 | 89.3 |
| 1552 | 88.8 |
| 1590 | 86.9 |
| 1627 | 84.8 |
| 1664 | 83.1 |
| 1701 | 81.2 |
| 1731 | 79.4 |
| 1769 | 77.8 |
| 1799 | 76.1 |
One more, lm/W for XM16 and std Al star:
| XM16 lm/W | Al Lm/W |
| 174.0 | 170.3 |
| 161.0 | 162.3 |
| 156.3 | 154.4 |
| 149.6 | 147.2 |
| 143.8 | 141.0 |
| 138.6 | 136.4 |
| 134.4 | 132.0 |
| 130.3 | 126.9 |
| 126.2 | 122.5 |
| 122.7 | 118.4 |
| 119.2 | 114.0 |
| 115.9 | 110.0 |
| 113.3 | 106.2 |
| 110.2 | 102.5 |
| 107.5 | 98.7 |
| 104.8 | 94.9 |
| 102.5 | 91.1 |
| 100.0 | 87.6 |
| 97.6 | 83.6 |
| 95.6 | 79.9 |
| 93.3 | 75.9 |
| 91.4 | 72.3 |
| 89.3 | 68.3 |
| 88.8 | 64.6 |
| 86.9 | 60.4 |
| 84.8 | |
| 83.1 | |
| 81.2 | |
| 79.4 | |
| 77.8 | |
| 76.1 |
|
Even in the efficacy race, Direct-bonded wins all the way past the finish line.