Most of them use a series of cogs and gears etc and those cogs etc are oiled at the factory.
The oil or grease is a mix of oil and a binder so the oil stays on what it is applied to.
Over time the oil leeches out of the mixture and leaves behind just the binder and that binder is sticky and some turn hard leaving just a "glue".

Time is its enemy.

Clean off the glue, apply new grease and you should be good for another 20 years.

In some cases, the rubber belt which handles the tray opening mechanism becomes worn out from use.

Replacing that can sometimes solve the issue, but there are other causes as well.

The rubber belts don't only get worn out from use, their material slowly degrades over time and tend to disintegrate into black goo. How much time depends on the belt in question, but after a couple of decades most will fail, even if the drive is NOS and has never been used.

Kahenraz wrote on 2024-09-04, 03:42:

I can confirm that this is not a belt-related issue.

I'd like to know how you're coming to that conclusion, since 99% of the time in my experience this problem is due to the tray belt deforming and can easily be fixed by flipping it inside out.

jmarsh wrote on 2024-09-04, 05:50:

Kahenraz wrote on 2024-09-04, 03:42:

I can confirm that this is not a belt-related issue.

I'd like to know how you're coming to that conclusion, since 99% of the time in my experience this problem is due to the tray belt deforming and can easily be fixed by flipping it inside out.

I inspected the belt and it is still in good condition. Replacing the belt in drives that have had this particular issue has never made a difference. The answer to this problem lies elsewhere.

Kahenraz wrote on 2024-09-04, 07:41:

I inspected the belt and it is still in good condition. Replacing the belt in drives that have had this particular issue has never made a difference. The answer to this problem lies elsewhere.

I don't think a visual inspection is enough. Sometimes belts simply get looser with time (with no apparent damage) and can no longer adequately grip the pulleys.

Unless you've replaced the belt with a known good one, and the drive still didn't work, I wouldn't rule it out.

Just punch top of cover slightly after pressing button and tada its open! 😀

Rubber bands are the most fragile parts of just about any audio/video player. I wonder if there have been "rubber band-free" player designs.

The problem was the following in all of my cases:
There is a thin belt in front of the drive, taking an egg shaped form.
The belt will always rest with a sharp curve at the small spindle.
The belt will adapt to the sharp curve form somewhat over the years and lose some traction.
Slide the belt by 1/2 circumference out of position, so the deformed part is now on the big wheel.
Then it will work better or trouble free.

(If this is your problem, the drive will open with a disk, but not without. The force needed to pull the magnetic disk lock or sticky rubber is too strong without a disk inside unless you restore belt traction)
Leaving a disk inside is also a viable solution.

Yeah, the egg shaped rubbers are the primary cause along with caked up lubricants in the old drives. Adjusting the rubber belt position like that has helped me in past and cleaning off the old caked up lubricants also.

dormcat wrote on 2024-09-04, 09:22:

Rubber bands are the most fragile parts of just about any audio/video player. I wonder if there have been "rubber band-free" player designs.

I don't know about PC CD drivers etc. but higher end AV gear is often made without rubber bands and instead have direct drive servo motors. They'll still have caked up lubricants issue though...

This is planned obsolescence. I have a Plextor drive, a heavy duty solid thing, evidently built to last, which has a direct mechanical linkage between motor and tray.

Why aren't all drives built like this? Because then you wouldn't be forced to replace them every 3 years.

(I assume my Plextor was built for industrial customers that don't tolerate being f*cked around like this.)

Errius wrote on 2024-09-04, 10:00:

This is planned obsolescence. I have a Plextor drive, a heavy duty solid thing, evidently built to last, which has a direct mechanical linkage between motor and tray.

Why aren't all drives built like this? Because then you wouldn't be forced to replace them every 3 years.

(I assume my Plextor was built for industrial customers that don't tolerate being f*cked around like this.)

Actually, the non-opening tray is an almost exclusive Plextor problem for me. The other drives work. All my IDE Plextors got it, even the Premium.

I've also had some drives with cracked nylon cogwheels, when new they were pressed tight onto the shaft with some tension, but after years of use they can crack and then the shaft spins but not the cogwheel.

Errius wrote on 2024-09-04, 10:00:

This is planned obsolescence. I have a Plextor drive, a heavy duty solid thing, evidently built to last, which has a direct mechanical linkage between motor and tray.

Why aren't all drives built like this? Because then you wouldn't be forced to replace them every 3 years.

(I assume my Plextor was built for industrial customers that don't tolerate being f*cked around like this.)

That's a pretty conspiratorial way of looking at it.

You could also just say that the average customer wasn't prepared to pay the extra cost for an overengineered product that would last vastly longer than its commercial lifetime. You paid over 50% more for Plextor drives than for otherwise pretty decent rubber-band drives from say Toshiba or Mitsumi. If you were intending to keep using it for a couple of decades, that would be a wise investment, but if you expected to replace your whole PC in 3-5 years anyway, which was a very real expectation in the late 1990s, there was no added value in the more durable product - and if most customers wouldn't pay the extra cost, it's very understandable that manufacturers supplied what they would pay for. The fact that we are trying to use the things a quarter of a century on and encounter certain solutions disintegrating more than others is completely irrelevant for economics and morality of the business when these things were new.

If it has a rubber belt then there is a good chance, if fact its odds on, that after all these years it isnt performing as well as it did, even though it might look like its ok.

There are many different reasons why this could happen, but indeed a marginal belt is the most frequent issue.

A few weeks ago, I just repaired a few drives. For the first one, it was a belt issue. The original belt had disintegrated long time ago, so I had a conventional rubber band installed every few years. The problem with rubber bands is, they are too elastic and don't grip the wheels the same way. So even with new rubber bands, I still had issues. Eventually, I came up with a workaround, where I added a thin layer of electrical tape on the spindle top and bottom (motor) pieces to increase the distance between the tray locking magnet (that also holds the CD.) I did this, because I noticed the drive would always eject fine with a CD inside, but not when it was empty. Smacking or whacking the drive on top would also get it to eject, but I'm just not a fan of this method. Eventually when I get some polychloroprene adhesive (commonly known as your orange "shoe glue"), I'll replace the electrical tape with thin plastic cutouts for better durability.
As for the 2nd drive... it also had the original belt and not in the best shape. Unfortunately, since I didn't have any other belts or rubber bands of this size, I couldn't replace the original. Rotating the belt 180 degrees didn't do the trick, because the belt would slip a little sometimes and eventually find its way back in the original "egg-shaped" position again. Actually, the issue with this drive was not that it couldn't eject, but that it couldn't lock the tray by itself once it was closed. If I gave the tray a fast push at the end before it closed, it would sometimes lock OK. But not always. With no new belt available, I had to come up with a workaround again. This time, it was even simpler, but required a lot of observation of the mechanism. What I found is that there were several metal grounding tabs that would ground the laser tray once in the up position (or perhaps these were anti-vibration tabs of some sort??) All of these were probably fine for the tray mechanism to overcome with a new/good belt, but certainly not with a worn belt. Once I found this, all I did was bend the tabs a little so that they exert a little less pressure on the laser tray and that did it! Drive now opens and closes without any issues.
And yes, I did try a complete cleanup and re-lube of both of these above drives prior to attempting the workarounds.
Chances are, if the gears are lubed with white Lithium grease (and they almost always are in my observation), there's nothing wrong with it even after all of these years. But it doesn't hurt to try and add a little more in there. Just don't get any on the belt, because rubber is usually degraded by most oils... not to mention you'd be loosing traction that is already limited (especially with a worn belt.)
Now, I've heard some internet "sources" give recommendations such as boiling the rubber bands to "restore" them back to the original condition. This doesn't really work. At best, boiling the rubber band will just clean some dirt and oils out of it, but nothing more. Of course, in some cases, that might be just enough to get the belt to perform its duty.

Errius wrote on 2024-09-04, 10:00:

Why aren't all drives built like this? Because then you wouldn't be forced to replace them every 3 years.

Actually, it's probably more to do with reducing noise when the drive is opened and closed. I think everyone knows how much noisier a gear-driven drive is compared to one with a belt. Not sure whom it bothered, though (certainly not me.) But I suppose once the trend caught on, it never went back. On a related note, I remember us back in high school bugging this one teacher we didn't like too much with those gear-driven drives. IIRC, we had some beige Pentium 3 Dells with gear-driven drives in them, and they all made quite the noise when being ejected or loaded. All it took for one person to do it, and then half of the class would start doing it. It was like a mechanical noise orchestra of some sort (or early Dubstep music?? 🤣 ) We also had Nokia 17" CRT monitors with built-in degauss buttons. Never was there a day when someone wouldn't walk by someone else's computer and not press the degauss button on their monitor while they are hopelessly trying to focus working of whatever. 🤣

Anyways, back to gear-driven drives... yeah, I really wish all manufacturers would continue to make them like that and not use belts.
Rubber is a consumable material. So in finer applications where a mechanism relies heavily on its specific properties, there's bound to be a problem over time. After 10-15 years, the material degradation can be quite significant, even if it doesn't look like it on the surface.

dionb wrote on 2024-09-04, 13:40:

That's a pretty conspiratorial way of looking at it.

It is also just plain wrong. As I stated above, I have this problem with ALL my Plextor drives. Every single one! But hardly any others.

Personally, a vast majority of all optical drive mechanical failures, across many different makers, were one of three things. Grease drying and turning to glue just like ElectroSoldier said. Belt issues, mentioned by Joseph_Joestar. Or a broken gear or other small plastic bit in the mechanism. The first two are easy fixes, and broken parts were a mixed bag until 3D printing became so widespread and affordable, now they can also be done with relative ease. I've only ran across a couple drives in all these years where the failure was electrical, like the drive motor or laser for example, so perhaps I've just been lucky.

Rwolf wrote on 2024-09-04, 13:31:

I've also had some drives with cracked nylon cogwheels, when new they were pressed tight onto the shaft with some tension, but after years of use they can crack and then the shaft spins but not the cogwheel.

I have a Toshiba XM-6201B with this problem. A small plastic cogwheel is broken and no longer spins with the shaft. I fixed the problem with a drop of glue, but it will no doubt break again in future, so this 'repair' will have to be done repeatedly.