Tulipcomputer: Difference between revisions
No edit summary |
No edit summary |
||
| (One intermediate revision by the same user not shown) | |||
| Line 11: | Line 11: | ||
[[File:NdeyNGx.jpg|thumb|center]] | [[File:NdeyNGx.jpg|thumb|center]] | ||
After recapping and repairing 12V psu due to blow rail caused by shorted tantalums, everything works (except for the HDD, but i haven't given up)]] | After recapping and repairing 12V psu due to blow rail caused by shorted tantalums, everything works (except for the HDD, but i haven't given up)]] | ||
Update: 14-12-23 | Update: 14-12-23 | ||
I managed to get really close to booting the old beast from a 3.5 inch floppy, reformatted to 720k. But the ancient floppy controller chip on the motherboard just doesn't support 80 track floppies, so it was a no-go. I did however manage to revive the miniscribe 3425 hard drive. I added a ton of additional low-esr decoupling capacitance right at the drive head stepper motor controller, to give it some extra oomph, and let it run warm on its side. As it warmed up i slowly fed maybe about 0.3 to 0.5 mL of sewing machine oil into the outside stepper bearing with a syringe and needle, and that seemed to do the business. The seek error went away. This was quite a lot of oil. I think perhaps some of it may have made it to the second inner bearing. Surprisingly, it now boots reliably. Next task: get the data off before it crashes. | I managed to get really close to booting the old beast from a 3.5 inch floppy, reformatted to 720k. But the ancient floppy controller chip on the motherboard just doesn't support 80 track floppies, so it was a no-go. I did however manage to revive the miniscribe 3425 hard drive. I added a ton of additional low-esr decoupling capacitance right at the drive head stepper motor controller, to give it some extra oomph, and let it run warm on its side. As it warmed up i slowly fed maybe about 0.3 to 0.5 mL of sewing machine oil into the outside stepper bearing with a syringe and needle, and that seemed to do the business. The seek error went away. This was quite a lot of oil. I think perhaps some of it may have made it to the second inner bearing. Surprisingly, it now boots reliably. Next task: get the data off before it crashes. | ||
Update: 17-12-23 | |||
Managed to establish a serial link with the 8088, through norton commander 3.0, which happened to already be installed on the hard drive. It linked with another copy of norton commander running in a dosbox on my laptop. If you ever find yourself in this situation, please take note, norton serial link protocols are not backwards compatible. You must have the same version running on both machines. Nowhere in the (scant) documentation is this mentioned. A pox on their deprecated houses. Top speed was 115200 baud, but in practice 57600 was more stable, with just a blind null modem cable. No loop back handshaking required. | |||
Update: 20-12-23 | |||
All data has been backed up, and a disk image is happily running in dosbox. Only thing to do now is dump the bios eprom and compare it with IBM bios from that era. Apparently Tulip got in some hot water with IBM for just straight-up copying their BIOS in these earlier models, and ended up settling out of court eventually. It would also be fun to attempt to run some modern open source bios on it. | |||
Latest revision as of 03:15, 21 December 2023
| Project Tulipcomputer | |
|---|---|
| Restaureren van een oude tulip compact 1 8088 clone, inclusief dot matrix printer. | |
| Status | Initializing |
| Contact | thomas |
| Last Update | 2023-12-21 |
After recapping and repairing 12V psu due to blow rail caused by shorted tantalums, everything works (except for the HDD, but i haven't given up)]]
Update: 14-12-23
I managed to get really close to booting the old beast from a 3.5 inch floppy, reformatted to 720k. But the ancient floppy controller chip on the motherboard just doesn't support 80 track floppies, so it was a no-go. I did however manage to revive the miniscribe 3425 hard drive. I added a ton of additional low-esr decoupling capacitance right at the drive head stepper motor controller, to give it some extra oomph, and let it run warm on its side. As it warmed up i slowly fed maybe about 0.3 to 0.5 mL of sewing machine oil into the outside stepper bearing with a syringe and needle, and that seemed to do the business. The seek error went away. This was quite a lot of oil. I think perhaps some of it may have made it to the second inner bearing. Surprisingly, it now boots reliably. Next task: get the data off before it crashes.
Update: 17-12-23
Managed to establish a serial link with the 8088, through norton commander 3.0, which happened to already be installed on the hard drive. It linked with another copy of norton commander running in a dosbox on my laptop. If you ever find yourself in this situation, please take note, norton serial link protocols are not backwards compatible. You must have the same version running on both machines. Nowhere in the (scant) documentation is this mentioned. A pox on their deprecated houses. Top speed was 115200 baud, but in practice 57600 was more stable, with just a blind null modem cable. No loop back handshaking required.
Update: 20-12-23
All data has been backed up, and a disk image is happily running in dosbox. Only thing to do now is dump the bios eprom and compare it with IBM bios from that era. Apparently Tulip got in some hot water with IBM for just straight-up copying their BIOS in these earlier models, and ended up settling out of court eventually. It would also be fun to attempt to run some modern open source bios on it.