webdeck Posted October 3, 2020 Share Posted October 3, 2020 Hi folks. I've busted out my TI equipment again in another attempt to copy the data off an ancient Seagate ST-225 hard drive that is on its last legs. What I'm looking for is a program that can try to recover files. From what I can tell, the directory structure seems to be intact, but many files have sectors with errors. I'm looking for something that does a sector by sector copy of each file, with a very high number of retry attempts for any sector that fails, and continuing to copy remaining sectors even if one or more are unreadable. Even better if it can copy directly to the TIPI. I have a TI console, FinalGROM, TIPI w/32k, and a PEB with a Myarc HFDC, so I should be able to run most software. I also have a Geneve, so either a TI or Geneve program would work. Is there anything like that out there? Thanks, -Mike Quote Link to comment Share on other sites More sharing options...
+InsaneMultitasker Posted October 3, 2020 Share Posted October 3, 2020 Are you sure the hard drive is at fault? The errors you describe are equally descriptive of a Myarc HFDC that is not performing properly. Do you only have one hard drive? Has the HFDC circuitry ever been refreshed? Quote Link to comment Share on other sites More sharing options...
webdeck Posted October 3, 2020 Author Share Posted October 3, 2020 I did need to get the HFDC repaired (gate array replaced) because it wasn't working at all initially. Are there other common failures? The hard drive has a hard time spinning up - I have to power cycle it several times typically. It's also the same set of files that have issues, and I can hear the hard drive reseeking at the same points when trying to copy the files, so that's why I suspect the hard drive. I was persistent using ForceCommand to copy one of the failed files and it eventually managed to read the whole file after trying a couple dozen times, which is why I still have some hope that I can recover the data. But it is also possible the heads crashed at some point in the past and scratched the surface to make it unrecoverable. Quote Link to comment Share on other sites More sharing options...
+InsaneMultitasker Posted October 3, 2020 Share Posted October 3, 2020 Yea, the hard drive does sound suspect, provided you are certain the drive is getting sufficient power to spin up and operate. As far as the HFDC is concerned, a few good pictures will help determine what has been done to the board over time. If you are comfortable doing so, snap a front and back picture of the circuit board. Quote Link to comment Share on other sites More sharing options...
webdeck Posted October 3, 2020 Author Share Posted October 3, 2020 Here's my HFDC. I got this directly from Lou Phillips, so it may be a prototype or early model. He had to do some hand work on it because the floppy controller was having issues. The voltage regulators are bypassed (with a very poor soldering job by me from over 30 years ago) because I have a PC power supply in my PEB and provide 5V from that directly. 1 Quote Link to comment Share on other sites More sharing options...
webdeck Posted October 3, 2020 Author Share Posted October 3, 2020 Well, maybe it's both the hard drive and the controller with issues... Today I can't get reliable reads at all. One DIR command succeeds, and the next one on the same directory fails, or returns a partial directory. Quote Link to comment Share on other sites More sharing options...
+InsaneMultitasker Posted October 3, 2020 Share Posted October 3, 2020 The LS31 (write) and LS32 (read) are original stock. Taking into consideration that you bypassed the regulators, the two chips could still be viable, though in my experience its not a matter of 'if' but of 'when' they will fail. I have replaced these two chips on every card that I have updated or repaired for this very reason, regardless of current state. The card itself appears to be the earlier production run with original electrolytic capacitors [possible exception C14] and the QIC-02 circuitry is intact. (The QIC-02 circuit is benign but I typically remove the 3 chips and socket to bring the card in line with the latest revision). U17 typically has a capacitor connected to it on the back side of the board. I am not seeing this in the pictures. I believe it is tied to the hard drive step line circuitry, though I'd need to do some further review to confirm. I doubt it is a contributor to the current problem if the card worked without it in the past. 1 Quote Link to comment Share on other sites More sharing options...
Ed in SoDak Posted October 3, 2020 Share Posted October 3, 2020 (edited) Any way to mount this on a PC? I'm guessing not and can think of several reasons; MFM, proprietary format, obsolete connector, etc. Or maybe snag another drive as a test device, to try to verify proper function of the HFDC card. Sure, easy peasy! I've got one of those hard drive dongles with multiple data and power connectors to plug in by USB and just work. Now wouldn't it be nice if... Edited October 3, 2020 by Ed in SoDak Quote Link to comment Share on other sites More sharing options...
webdeck Posted October 4, 2020 Author Share Posted October 4, 2020 (edited) As it turns out, I had bought a spare LS31 and LS32 last year when I started this project before 2020 derailed everything, since I had heard they often needed to be replaced. So I dug those out of the closet today, desoldered the chips, soldered on sockets, and inserted the new chips. I also redid the voltage regulator jumpers while I was at it. Let me know if there's anything else I should try replacing - maybe the caps? It does seem more reliable (or maybe I just got lucky this time), but there are still files that fail to copy. There is one directory I managed to copy before, but now I can't get a full directory listing on it, so while some things are better, others are worse. I wonder if there are any electronics on the hard drive that need to be replaced, or if it's the platters that failed. Which brings me back to my original question - is there a sector-by-sector copy program that will do some retrying and continue copying even if some sectors are bad? Edited October 4, 2020 by webdeck Quote Link to comment Share on other sites More sharing options...
+InsaneMultitasker Posted October 4, 2020 Share Posted October 4, 2020 Check your cables for proper alignment; the end connector/cable is notorious for being offset just enough so that the cable fingers end up misaligned to the edge connector. I am not aware of any programs that provide the desired functionality you seek. I have on occasion used Randy Moore's sector editor SectorOne to scan through sectors manually to rewrite a handful of bad sectors but that's probably too manual for your needs. @BeeryMiller posted source code to MDM5 not too long ago and it might be possible to modify the read routine to ignore errors/retry. It is also very likely that you could inspect the source code to the HFDC DSR to increase the low level IO retry count, though it would take some sleuthing since the source we have is not the same version H11 base that most cards use. DM2K is maintained by @F.G. Kaal and I believe you inquired at @jedimatt42's FC in another thread, maybe they have a suggestion from their programs' points of view. Quote Link to comment Share on other sites More sharing options...
Recommended Posts
Join the conversation
You can post now and register later. If you have an account, sign in now to post with your account.
Note: Your post will require moderator approval before it will be visible.