Re: cheap-shit laser displays

BTW, since you Cced kragen-tol, I’ve forwarded your message to
kragen-discuss. I might need to fix up the kragen-discuss setup a bit
for you to be able to post there directly; let me know if you have a
problem.

Danny Ayers wrote:
> On 8 July 2010 09:37, Kragen Javier Sitaker <[email protected]> wrote:
> > The cones of cheap dynamic earphones can move hundreds of microns at,
> > I hope, tens of kilohertz.
> 
> I would imagine so - many earphones can give reasonable response at
> 10kHz+.

Well, my concern wasn’t that they were low-pass filtered below there,
but rather that they couldn’t move hundreds of microns so
fast. Clearly they can move hundreds of microns at, say, 100Hz; you
can see them doing it. Can they do it at 10kHz, or is their
displacement limited to, say, <10μm? I’m not sure.

> My guess though is that the deflection would be significantly
> non-linear.

That’s possible. At 10kHz you probably wouldn’t be able to hear any
harmonics, so even fairly strong harmonics wouldn’t affect the
earphone’s quality as an earphone.

> There should also be some advantage here in driving the scan with a
> sine wave, I believe it would reduce the 'bounce' of overtones a
> sawtooth would create. Although it may turn out to be
> straightforward to compensate for non-linearity using analog and/or
> physical components, I suspect there may be advantages in doing a
> lot of this this digitally. For example, a crude (but fastish) 1 bit
> D/A converter followed by an analog resonant filter, with extra
> damping on the moving parts might be combinable to provide an
> accurate scan curve.

That’s a really good idea.

If the speaker’s response tops out a bit above 10kHz, driving it with
something close to a sine wave might not be optional; the analog
resonant filter might be the speaker cone itself!

The downside of sine-wave scanning is that you need a higher
frequency, because the beam will waste a lot of its time dawdling
around the edges of the display, and spend less time usefully
scanning. So you get bonus horizontal pixels at the edge of the
display and bonus vertical pixels in the middle of the display.

Presumably the lower-frequency vertical scan wouldn’t suffer from this
problem if it’s built out of the same mechanism.

-- 
To unsubscribe: http://lists.canonical.org/mailman/listinfo/kragen-discuss