<B0101>
hmmm interesting, in a simulation, Al ions @ 30 KeV displace quite a number of Silicon atoms from the target, but none of the Al ions actually replace the displaced Si atoms...
<B0101>
so, I assume we need to get to the MeV range to get actual implantation?
<glowplug>
Megavolts?
<B0101>
yup
<azonenberg>
Hundreds of keV might be enough
<azonenberg>
In any case, diffusion is much easier to do at home
<glowplug>
We don't need that technology for ~405nm and larger devices though correct?
<azonenberg>
just coat the ion source onto the surface of the wafer and heat up to 1000C or so
<azonenberg>
glowplug: 405 is rather optimistic
<azonenberg>
i'd shoot for a micron
<azonenberg>
with 5 and 10 as stepping stones
<azonenberg>
And no, you dont need implantation for large devices
<azonenberg>
diffusion is fine
<glowplug>
I have solved quite a few problems today.
<glowplug>
One major problem is with the glass linear encoders I am curious if you have some insight.
<glowplug>
I can get 5 micron accuracy for $200 per axis.
<glowplug>
Its gets unrealistically expensive after that.
<glowplug>
If I can get 1 micron cuts in a glass strip I can handle the optics / electronics. So what I'm trying to do is get the strip outsourced and DIY everything else.
<glowplug>
The reason for the linear encoder is real-time antibacklash compensation. Without it the repeat accuracy of the machine wont be ~1 micron it would probably be far worse.
<glowplug>
Also by $200 per axis I mean $200 per linear encoder not counting any other parts. That price really need to drop to keep the machine cost reasonable.