• Please review our updated Terms and Rules here

Memory Size Error and Non system disk on 5170

Isn't that the IBM 0665, Type 20, 30 mb. drive like Bettablue just sold? I think that was the original drive in the AT. If so, I have one and it works perfectly.
 
If you're getting a write error you should LLF the drive and then try again.

Even before that you should run Norton Disk Doctor or Spinrite on it. They can possibly correct format flaws and mark those sectors bad that are bad and then you will likely have full access to the drive. If you need either/both lemme know. I can help you there. I say this assuming your floppy drive is working correctly cause without that everything becomes more of a PITA.
 
Stone, it is the LLF that is failing with the write error. I can try Spinrite, but it is that very types of tests in Speedstor and the IBM Advanced Diagnostics that I've been trying to run with those errors also coming up.

Update: spinrite 6.0 won't load on this machine. It was tested in Dosbox before hand. It crashes and needs a softboot. Is there a particular old version I should be trying?
 
Last edited:
Error 1706 is sort of nebulous--"Write Fault", which I'm assuming is the detection of an active signal on the control interface pin 12. The reasons are various--according to the ST412 OEM manual (pretty much the standard):

  • Write current in a head without WRITE GATE active or no write current with WRITE GATE active and DRIVE SELECTED.
  • Multiple heads selected, or no head selected or improperly selected.
  • DC voltages are grossly out of tolerance.

Okay, so it's a catch-all error. So let's use the process of elimination. First off, power the system up and check the power supply output (that might have something to do with the Parity Check you received earlier).

Those being within +/- 5 percent of nominal, the next thing to check is to see that you've set the correct drive geometry in your CMOS setup. Note that neither the controller nor the diagnostics have any way to actually query the drive geometry; they take your word for it. If you're wrong, bizarre things can and do occur.

Finally, note that the first reason given points to a controller issue, not a drive issue. Check your cables.
 
Last edited:
Mark2000 said:
Stone, it is the LLF that is failing with the write error. I can try Spinrite, but it is that very types of tests in Speedstor and the IBM Advanced Diagnostics that I've been trying to run with those errors also coming up.

Update: spinrite 6.0 won't load on this machine. It was tested in Dosbox before hand. It crashes and needs a softboot. Is there a particular old version I should be trying?
Click to expand...
You need Spinrite II... Version 2.0 or something old like that.

Or, like Chuck said, it could be the cable itself. Do you have another set?
 
Mark2000 said:
Stone, it is the LLF that is failing with the write error. I can try Spinrite, but it is that very types of tests in Speedstor and the IBM Advanced Diagnostics that I've been trying to run with those errors also coming up.
Click to expand...
FWIW, Speedstor will test the controller in the Diagnostics section. While you're there try the seek test as well.
 
I don't have the hardware to test the voltages on the PSU. I'm pretty sure all the errors were do to the 370 board as they cleared up as soon as I removed them. I haven't entered any drive geometry. I've just set it up as a type 20 which is what it says on the front of the drive. I just rechecked the cables.

I think the issue is moot, though. On running the advanced Diag disk again after reseating everything and pulling the 720k disk out the HDD gave out a rattle like I've never heard from a disk. It was like there was a loose piece of metal being spun around by the platters spinning. The must be a physical injury in there. I don't think there's a point it messing with it anymore. Stone, I left you PM about that drive you have. Thanks!
 
Invest $5 in a cheap multimeter and check those voltages. Out of spec antique SMPSUs are far more common than you'd think. The added load from the PC 370 cards probably aggravated the situation, not caused it.

<rant>I'll admit to having a peeve--entirely too much vintage stuff gets scrapped because the owner doesn't want to learn to research and troubleshoot it, much less repair it. This is US-made stuff from the 1980s, not today's use-and-discard junk.</rant>
 
Mark, when I tell you in a PM what that other IBM drive (that *maybe* was in working condition) was sold for I'm sure you're gonna do what Chuck has suggested, before you go and replace your drive. :) Also by checking the cables Chuck meant ... replace them with another, known good set. The only way I know if a cable set is good is if it works with a known good drive/controller.
 
Rant accepted and understood. At the very least It would have been good to sell it. But the facts are the techs didn't park the drives before mailing the computer. UPS bumped the thing hard enough that the daughter board on the I/o card disattached and was loose in the case. And, forgive me, but the drive was made in Singapore, not US :). I'm going to test the PSU when I have the chance because that just seems wise in general. I'm only coming up with ATX procedures, thought, both in the forum search and the rest of the web. Do you know a resource for the procedure and what to expect?
 
Last edited:
Mark2000 said:
And, forgive me, but the drive was made in Singapore, not US :).
Click to expand...
FWIW, I have dozens of MFM/RLL drives and they're all made in Singapore or Maylasia. :) :)

I've got lots of drive rails. Can I figure out which one's are for the 5170? I'll try.

I might have a grey 1.44. Can I match the color scheme?

I've been ripping face plates off of FH doorstops, lately. Do you want to cover an empty bay or a full bay?

I didn't send the PM I promised earlier because I figured you weren't done working with your drive yet and I want you to be finished with that project before I induce heart failure. :) When you're ready for the bad news, just lemme know.
 
Sure, leave the power supply connected, although you may need to remove it slightly from the case to get to the motherboard connectors. You can find the voltages at the motherboard connector described here. As far as multimeters, if you're on the cheap you can get them for a couple of bucks at your favorite Chinese tool store--or check some of the big box home stores or even Radio Shack. Good enough for casual use. If you've only got Radio Shack near you, this kit is a pretty decent deal.
 
Sorry, but I'm dumb. When you say keep it connected do you mean to the board or to the wall. I'm figuring the second, but wondering if you might mean both.
 
You want to see the voltages while the system is connected and running (loading by devices can change voltage readings dramatically). You can also get the most important readings by checking a disk drive power plug while the system is operating (red wire is +5VDC, yellow is +12VDC, as referenced to the black wires in the center of the plug). No need to remove anything but the case cover.
 
Been out of state, so I just had time to run the tests now. The PSU checks out within .07v on every prong under full load. I've also found out that it's the Sound Blaster causing the trouble with the Diag disk and not the 370 boards. I wonder if its a resource problem. Now that I know the PSU is good I don't mind putting other stuff in it.
 
You must log in or register to reply here.
Back
Top