Laser scanning microscope from DVD pickup heads (2020) | Hacker News Reader

Laser scanning microscope from DVD pickup heads (2020)

gaudi.ch

221 pointsuramondi31 comments

If you are throwing an old DVD or blue ray player and you are sure that it will not be recycled, take a moment to disassemble it, get out the reader head and marvel at the complexity of what you are about to put to the trash.

Or a VCR.

A lot simpler but still interesting, an audio tape drive has surprising mechanical tricks, especially the coupling between physical interface and the motors/heads.

But yeah a VCR slot and head are brilliant.

How it pulls the tape out of the cassette is pretty weird, and there are several variants of the mechanism. Also crazy how flat the heads have to be or the tape will be shredded. The air pressure can't be too low, or the tape will be shredded.

Nice but very dangerous for your eyes to work with these things.

Here many people criticize C because of its "landmines". A laser pointed somewhere in the room that gets reflected until it gets to your eye can easily do permanent damage. Definitely do not try doing this at home

Alternatively, just pick up some adequate laser safety glasses for everyone who's going to be in the room while you work on it (including yourself)

It's not like this is an industrial cutting laser, you can make it perfectly safe to experiment with these smaller ones if you get the proper gear

Sure, but it doesn't hurt to warn people who have no experience with that

You can make it safer with things like this:

https://awesome.tech/adding-a-safety-switch-to-your-laser/

DVD Laser Scanner Microscope - https://news.ycombinator.com/item?id=26012652 - Feb 2021 (40 comments)

Yeah, (2020) if you care to add.

Added. Thanks!

This might make practical data recovery from fragmented media easier --- although still quite time-consuming.

So at scale, no technology is too complex to reach to consumers. I think we will see blood analysis microscope in our wristwatch, continuous diagnostic, robotics delivery etc in near future.

No tech is too complex if it brings enough profit. I'm certain about delivery bots, for they already exist. Continuous blood monitoring is more likely to remain a niche thing, only for people with specific conditions; also, having a hole in your skin to keep a bunch of sensors constantly in contact with blood is not fun.

You can monitor some things about your blood without poking a hole in your skin.

Eg oxygenation rate and blood pressure.

(I agree with most of what you said.)

And hopefully soon Blood Sugar. https://www.diamontech.de/home

Complex, no. But many other things than complexity add to the cost of making something and a lot of those do not scale to higher volumes in any way like complexity does. Indeed cost-saving innovations often increase complexity in order to reduce those other components of cost (the hoops engineers wind up jumping through to save fractions of a penny in high-volume products is crazy. For a mild example, having a full network inside a car increases the complexity drastically, but decreases the cost of making up the wiring in the car a little)

Didn’t someone already try to do blood analysis with just a drop of blood and failed miserably?

I assume you mean this in jest. I get it.

To be fair/serious if you add the "near past" with "near future" you might get enough time for things to change.

Was there something fundamentally impossible about blood analysis with a drop of blood?

One problem is that blood isn't really a homogeneous substance. When dealing with small test samples, you might see different results, perhaps widely-varying ones, from one drop to the next. (Play around with a glucose meter sometime.) Things like where the blood was drawn from, what contaminants were present on the skin at the time, and the age and health of the patient can all have an effect.

Also, many tests require a centrifuge, so that's something else you have to account for if your goal is to perform the same tests with only one drop of blood. Unless your plan is to pull some Star Trek-level technology out of your hat.

TL,DR: you didn't need to consult a team of PhDs to debunk Theranos's business plan, you could have asked any experienced med tech. The investors didn't bother doing either.

> Also, many tests require a centrifuge, so that's something else you have to account for if your goal is to perform the same tests with only one drop of blood. Unless your plan is to pull some Star Trek-level technology out of your hat.

That was basically the plan. Or failing that, to convince people that such technology existed or was close to existing. It's not too implausible that the centrifuge could be replaced with diffraction analysis or something since the goal is to test for the same things, not necessarily to "do the same tests." But that doesn't help if you're looking for something present in one in a million blood cells and only have a quarter-million cells to look at.

They wanted to believe

At the time of Theranos's debunking there were some posts to HN showing that it was mathematically implausible. Something about signal to noise and specificity at extremely small molecule counts. I don't know how to find it, but I assume somebody in the hacker news hivemind knows what I mean and can dig up the link.

Breathalyzers can't serve as a credible proxy for blood alcohol down to 100ppm without any consideration of physiological differences and yet they're accepted as divination tools.

There were some fundamental issues with some of what Theranos was claiming to do. Mainly when detecting substances in blood with very low concentrations using very small volumes, especially when the blood is taken from the finger. In those cases even if you could pick apart every atom in the sample perfectly you still would not be able to do what Theranos was claiming to do, because the substances would either not be in the sample or there would be so few molecules that the uncertainty in the real concentration would be too large. real-time monitoring is a dream which in principle does not suffer the same fundamental issues, but does require much larger jumps in technology compared to current means of running tests.

So? One failed attempt doesn't mean nobody else will ever succeed.

In theory maybe, in practice tech is very often steered by what brings money.

Just look at the current state of the internet, imagine how good it could have been

Obvious in retrospect but optical disks use pretty high NA optics (for air)

            λ       d       NA     WD
    CD      780 nm  780 nm  0.5    ~2 mm   (+1.1 mm cover)
    DVD     650 nm  500 nm  0.65   ~1.5 mm (+0.6 mm cover)
    Blu-Ray 405 nm  240 nm  0.85   ~0.5 mm (+0.1 mm cover)

I'm disappointed that there was no microscope shot of the OG: LaserDisc!

Simpler project: get the lens from a dvd/cd drive and use duct tape to fix it to your cellphone camera. Instant microscope.

i honestly have no idea what this article is supposed to be about. and i did read it.

Reply on news.ycombinator.com