IDE bus getting cranky

I'm having a bit of a problem. My empeg seems unable to talk to any hard drives. It's not a case of a bad hard drive, it's simply unable to see the drive device itself on the IDE bus.

RAM disk driver initialized: 16 RAM disks of 4096K size
empeg single channel IDE
Probing primary interface...
Probing primary interface...
empeg-flash driver initialized

(should actually start displaying the drive manufacturer, etc)

I used to have a similar problem, until I recrimped the IDE cable, things were fine for several months afterwards. But no such luck this time, and I've even used a different IDE cable. Same symptoms. I've tried swapping drives, still same symptoms (forget the fact that there's not an empeg build on that drive, the IDE bus will still see it, even if there's no OS on it) I've tried combinations of both, including setting master/slave/cable select and using the other end of the IDE cable. The drives themselves spin up, but they just don't show up on the bus. I've checked all of the solder points for the IDE cable, all fine. It's not the drive and not the cable, and at this point, I'm out of ideas. I've done my best searching the forums, but they all were mainly due to faulty connections that doing a reseat or recrimp would fix the issue. And it actually did boot up once, but when I re-assembled it, no love.

So any suggestions that may be hidden in the forums that I haven't found?

What're the chances of actually getting someone (Rio?) to check it out and fix it?

possible this is just the bad solder joint issue?

If so I bet several forum members can help you there. Or Eutronix??

Checked them all out.. all seem solid. No jiggle on any of them. My previous issue, with all of my removing and replugging of the ide cable, caused one to break loose. Had to solder it back down, that one's solid (probably more so than the others) and none of the others seem to have a problem.. :/

If you were comfortable when doing that one pin, then I'd suggest hitting all the others now. The symptoms you are describing are 100% in line with a bad joint on the IDE header, and I have seen failing header joints that I couldn't spot visually. It really is worth eliminating this as a possibility before spending any more time chasing the problem down blind alleys, and the only way to do that is to retouch every pin.

Chance of getting someone to fix it from Rio: 0%.

Chance of someone here helping to fix it: 100%.

Give us a boot log dump from serial so we can have a look.

I would also check that the drive jumper on the slave drive is making good contact with the pins on the header. Read the FAQ on this.

Sorry for the long delay. Been busy.

I went ahead and touched up the solder points for all the pins. But I still got sporadic results. Since I was going on a road trip, I babied it along. Got it to boot up. Put the empeg back together piece at a time making sure it booted up after each step, and hoping and praying that it still booted up when I was done. It did. Got it in the vehicle, and still booted up. So I at least had music for my road trip. My confidence that it's now solid isn't quite there. I haven't removed it from my vehicle for fear of it not coming up again.

I couldn't find any corresponding solder points on the underside of the motherboard, or am I missing something there?

And I've checked from pin end to solder connectivity, and that's fine, but it's obviously a bit more difficult to check from pin to final destination, so any other suggestions?

Thanks for the help guys..

The IDE header is surface mounted, not through hole. No need to remove the motherboard...

If you've touched up all the pins ok, then it has to be downstream of the header. Either the cable, or the drive select jumper that Rob mentioned.

Tried two different cables, two different drives, and there's only one drive at a time, so there's no slaving, but I've tried one or the other position on the cable, as well as the various combinations of cable select/master/slave.

But after the re-soldering I should probably play with all of the combinations again, just to be sure.

It would be extremely unusual to get an intermittent problem like this that was not mechanical in nature. I doubt that it's the drive, especially as you've tried two.

Yeah, I'm sure it is mechanical in nature.. but where that is and how to find it will probably be the bitch of it all. If it's intermittent, then I would doubt it's the chipset, but I've seen crazier things, and who knows with my luck..

But I'll play with it a bit more and hope for the best.

Thanks!

And you've checked and rechecked every single item in this list?

Yep. Tried recrimping the cable. Using a different cable. Using both cables with a different drive. Re-soldered the pins. Drives spin up fine. No complaints on the chipset. But on failure, the chipset simply does not see the drive at all. No recent changes to the empeg prior to this issue (weeks since last hijack upgrade.) I really did exhaust all possibilities before I came to the forums for help. I personally dislike when people ask questions that are already answered and are just too lazy to do the research first, and hate being one of those people even more

It's probably still the IDE header. Because it is (unfortunately) surface mount soldered, the copper pads on the mainboard sometimes lift -- especially when "repaired" using too high of a soldering temperature. With this, everything looks fine to the eye, but the connection is still very flakey. The repair consists of detecting those bad pads, and then running jumper wires from the header to where the traces originate.

-ml