Why can't they make solid state CDR drives?

[MadRat]

Seems to me that it ought to be possible to engineer a way to write and read CDRs without moving the CDR. Instead of spinning the entire CD it seems like they should be able to channel light in much the way they do on LCDs. This would alleviate one of the biggest energy wasters in a laptop. And if it could then perhaps a round, flat exposed CDRs would not be necessary so that less vulnerable designs could be used.

Anyone have any ideas how this could be done?

 

[HokieESM]

Oh, I'm sure it could be done.... it could even be a "scanner' that had the sensitivity to pick up the "dots" on the CD. Which brings up another interesting point--it could be done in parallel, making read speeds incredibly fast.

One (probably completely silly) idea is to use a CCD array to look at a "reflected" image that's been magnified. Then only a mirror would be needed to move around the light.

I think it may be a VERY expensive problem, though..... one of the BIGGEST benefits (in my opinion) of CD-RWs now is the low cost. Because if it gets close to $1/GB, I can just go buy another hard drive.

 

[Budman]

If the cds wouldnt need to move then there's no reason for them to be round or have that hole in the middle anymore would there?

Square cdrs hummm......

 

[mikable]

Yea, they'd look like cf memmory cards! Hey, how much does a 1gig cf card cost right now, and how long would the data on them last?




edit: ehh still a way to go 240 for a 1gig at newegg

 

[Topher]

Just make the media and drives smaller ala Dataplay. That would also take up less battery power.


I personally would like to see subnotebooks come with 3" CD (or DVD) drives. That would save you valuable space in making the laptop smaller. Even an external 3" CD (or DVD)-ROM would be a nice idea, although probably not very popular.

 

[ScrapSilicon]

Originally posted by: Budman
If the cds wouldnt need to move then there's no reason for them to be round or have that hole in the middle anymore would there?

Square cdrs hummm......

then they could not be called 'discs'. ..just compact squares...:frown:

 

[Idontcare]

Originally posted by: HokieESM
Oh, I'm sure it could be done.... it could even be a "scanner' that had the sensitivity to pick up the "dots" on the CD. Which brings up another interesting point--it could be done in parallel, making read speeds incredibly fast.

One (probably completely silly) idea is to use a CCD array to look at a "reflected" image that's been magnified. Then only a mirror would be needed to move around the light.

I think it may be a VERY expensive problem, though..... one of the BIGGEST benefits (in my opinion) of CD-RWs now is the low cost. Because if it gets close to $1/GB, I can just go buy another hard drive.

HokieESM Now that is a great idea. Are you familiar with optical MEMS like the micro-mirror display enabled DLPs and TVs? TI makes a lot of these 2D-mirror arrays in variable sizes (1600x1200 I think is the largest we sell). Agere/Lucent (and maybe HP?) makes them for optical network switching too. My point being that with a micro-mirror MEM it should be relatively straightforward to augment todays single laser source-single write location into a single-laser source-multiple write location device. The mirrors are about 5 micron on edge.

For CDR writing you could burn maybe 1000 tracks simultaneously. And for reading you would have a laser probe source striking the mirrors (reflected to maybe 1000 tracks simultaneously) with the reflected light striking a second row of mirrors with the light then reflected to a CCD as you suggest. The movable mirrors would allow stationary media, but you may still want to take advantage of rotational media to enhance data transfer rate (DTR).

It would be a monster for DTR, both writing and reading (and natuarlly extendable to DVD). I imagine the most expensive/performance limiting part of the system will be the DSPs to handle the flow. I'm probably out of my league (not my area), but I'll venture to posit the analogy of IDE vs. SCSI where the limitations in DTR and areal density are tradeoffs that must be optimized to compensate for finite signal processing capabilities of the DSPs employed in converting disk stored data into binary data.

Very cool idea HokieESM , when you going to apply for the patent?

 

[alkemyst]

This is ingenuity in motion...never assume anything as modern manufacturing requires so many levels of approvals and vetos it's not funny.

I have seen even hard drive data written and read as an unlimited number of cpu registers and processed at uncalculatable times.

 

[Lord Evermore]

Considering the cost of CCDs for digital cameras, which are pretty tiny chips, I think it'd be cost prohibitive to do something like that. Trying to integrate some sort of flexible/movable mirror would probably end up being more complex and expensive than it's worth. It'd be like taking one technology and trying to meld it with a completely different kind. Solid state devices are already in use, but obviously expensive compared to a CDR.

They'd also have to reinvent the way the laser system and the disc surface work. With a CDR, the laser is beamed directly at the disc, at a 90 degree angle I believe. If you tried to have a single small CCD with a mirror to send the beam to whatever area of the CD is being read, then the laser would be hitting at odd and varying angles. Setting up an array of multiple mirrors around the inside would probably start to make it expensive, and more prone to failure. Imagine the issue of dust on a CD drive lense multiplied by several dozen.

 

[Idontcare]

Originally posted by: Lord Evermore
Considering the cost of CCDs for digital cameras, which are pretty tiny chips, I think it'd be cost prohibitive to do something like that. Trying to integrate some sort of flexible/movable mirror would probably end up being more complex and expensive than it's worth. It'd be like taking one technology and trying to meld it with a completely different kind. Solid state devices are already in use, but obviously expensive compared to a CDR.

They'd also have to reinvent the way the laser system and the disc surface work. With a CDR, the laser is beamed directly at the disc, at a 90 degree angle I believe. If you tried to have a single small CCD with a mirror to send the beam to whatever area of the CD is being read, then the laser would be hitting at odd and varying angles. Setting up an array of multiple mirrors around the inside would probably start to make it expensive, and more prone to failure. Imagine the issue of dust on a CD drive lense multiplied by several dozen.

We're not talking about the 2D CCD arrays you are thinking of. They make these 1D (stripe) CCD chips for use in UV/VIS Spectroscopy and are also used in cheap flatbed scanners. There is no need to reinvent the way a laser system works with the chip and disk. It's simple geometry, no more complicated than what you described. If it wasn't, then micro-mirror enabled devices (DLP, TV, CRT) wouldn't work and be available in Best Buy. Issue of dust on the lens changes nothing, that is a problem that exists in which the probability of causing an error scales as a function of the ratio of active surface area (lens) to particle cross section (dust) regardless of where the laser is going (single data bit or to multiple). Think of particles on Si wafers and how the defectivity density effects chip yield. One chip/wafer makes bad yield, right? Can you see the data bits on your CD? Can you see the scratches on your CD that are as big as the data bits?

You are right about one thing, and that is doing anything beyond nothing will raise the price of whatever product is under discussion. I don't think that this was an unrecognized factor. How much more expensive is the question, not whether it will be.