On Tue, Apr 16, 2019 at 11:35:41AM -0400, Brian Brubaker via vcf-midatlantic wrote:
For all the points you mentioned, EVERY hard drive eventually stops working and is thrown away or recycled(aside from the 1% that go to museums).
I think its when we get to drives that only a limited number are left so we are competing that people in this group have problems. For much of the period there are more drives than the collector market can use. More interesting discussion for me would be what have people heard on techniques for repairing drives. I agree all will stop working. When they are plentiful simple repairs and getting another/parts donor is the easy way but that will only work for so long before needing to develop more advanced fixes. So what have people heard on advanced repairs? Somewhat long winded ramblings from a not quite so old-guy. Assuming that the part of vintage computer collecting that wants to run their machine doesn't peter out I expect that some of my machines will be running when I no longer am. I suspect that will include some of my drives. For my RK05's (~1974 14" removable platter) the only parts I see issues with dealing with failure of are the platters and the heads. The bearings look to be reasonably easy to replace for people with access to reasonable tools unlike newer drives. This page about half way down has an interesting repair of RK05 cartridge. http://rescue1130.blogspot.com/2017/11/resolved-copy-and-burn-of-264x-termin... I suspect that I have ran across this when I was reading some of my cartridges in the past but assumed it was damaged platter when I saw the wobble. I still have them so can check and possibly repair them when I get some more of that spare time. Has anybody heard of the oxide binder breaking down on hard drive platters like it does in some tapes and floppies? The RK05 being non contact will help some in tolerance for minor deterioration. If the media deteriorates this could limit how long drives can be kept running. I also think that with sufficient $ or persistence someone will develop a method to re-coat platters if needed. Since the process should work for multiple drive models this may be able to generate enough volume to be worth someone doing this. I would think doing anything about damaged heads will be harder. I heard from one person trying to restore an early synthesizer that was wanting to have a platter swap done to try to recover data from a ST506 with bad bearings that the commercial data recovery places have started to throw out the tools needed to work on these drives due to low demand. From my understanding these places were really data recovery and not drive repair for long term usage. I gave a MFM drive to one of the list members who was looking for drives to investigate bearing replacement. It has the common noisy bearing problem. What was unusual was the second time he powered it on one of the head positioner bands broke. I had never heard of that failure before.
From looking at it I suspect with modern scanning and laser cutters you could actually make a replacement out of the proper spring steel stock.
For repairs that people can't do themselves will costs reach a point where enough people will be willing to pay for them? If not it will be emulators since the costs for them are more likely to go down. I also make an MFM emulator but for me its a tool and not an end goal. That's why I have put so much effort into reading existing disks. I would like to be able to keep the original drives working and if not make sure the data is captured so at least the computer can be kept working. The next step of this I posted about before will be to use my emulator to rewrite the contents back to the disk so hopefully it is usable again. I want to do something more permanent with the shims first. http://www.pdp8online.com/mfm/head_servo/