The reflected image is exactly an inverted projection of the die.
It doesn’t “only focus good in the center” in terms of lateral movement.
What he means is that the die needs to be in the center of the sphere such that the inverted image is projected directly onto it.
The projection will only be sharp when it is focused on the die itself, which is why it is important to get the height right and line it up on all 3 axis.
.
This does not mean that the size of the LED or collar matters.
A larger LED produces the same effect, an inverted image on the die of the exact same size of the die.
The reflected image is of uniform brightness, just like the die.
A larger collar performed slightly better because there is a larger distance from the LED to the surface, making surface imperfections less intrusive, and possibly having a better coating due to manufacturing process.
The only possibility where having a large LED and small collar would be an issue is if the LED is so large that part of the lens can’t see a portion of the die due to the collar opening.
The red lines won’t be able to reach the lens, making the outer portions of the lens far less effective.
I don’t think this will happen with a CFT90, small collar, and A037 lens but I can double check if OP wants me to.
I understand what you are saying, but it’s definitely not what he wrote. I don’t how strongly your translator is changing the meaning.
If you take the collar and shift it on the x-y plane the image stays sharp as long as the height does not change.
A larger LED is basically this but in all directions at once ^
This is how a spherical reflector works.
It does not become unfocused at different locations.
You can see that the projected image has uniform brightness, it is not any more intense in the center than it is on the edges.
If if was more intense in the center, first of all, how the heck would SMA have measured that (does he have a 0.1mm diameter luxmeter?), and second, that would violate the laws of physics because you can’t concentrate light to a smaller point than the source.
The ideal scenario when using the collar is to have the die image just as sharp as the die (so completely in focus) and directly on top of the die.
As shown below.
