Attention All 4100 owners

[blue68f100]

There has been some unexplained errors on 4100’s models after HD’s were upgraded. Further investigations show that a modification/repair was made to the original MB, ASSY# 707150048-001. The area of repair/modification was made to 2 areas FB1 & FB2.

FB1 is located between the two Promise Controller. The second area FB2 that has the same repair left of the left promise chip. This one (FB2) points toward the IDE cables.

If you have a ASSY # 70715048-001 or 003 look at the photos below and see if you are affected. 003 seams to be the repaired boards. But so far I have found 1 (-003) that has not been repaired. So you will have to look on the board for the repairs. Now some of the repairs have been done differently than the photo posted. If you are not sure post a photo of the area in question.

This is a -001 board that has NOT been repaired.


This is the 003 board showing locations of both area.


Below are pictures of a 003 A board that has been repaired.

FB1 area.

FB2 area

Notice that 1 component has been replaced with 2 and are bridged together with solder.

For more info please post your questions.

[jontz]

What exactly does this repair do? It appears from the pictures that you have added a second resistor in parallel, thus halving the resistance (assuming it is a resistor of the same value). Is that the case?

And yes, I am finally back...CRAZY around here for several months, but I survived

[blue68f100]

Without this repair your Raid Array will be unstable, data may be in trouble. It only seam to affect the smaller capacity units that have been upgraded. Don't know how may are affected. The 001 board is a 120gig model. The 003 is a 240gig.

[Hallis]

Now that is good info, if somebody were good at surface soldering you could repair your own.

[rpmurray]

Originally Posted by blue68f100

Without this repair your Raid Array will be unstable, data may be in trouble. It only seam to affect the smaller capacity units that have been upgraded. Don't know how may are affected. The 001 board is a 120gig model. The 003 is a 240gig.

Any specific error messages that display in the logs?

[blue68f100]

One of the errors are constant disk checks and resync on restarts. Unless you reboot you may not notice you have problem, then it's to late. Just pop the top and see if yours MB is a 003, if so the mod has likely already been done. If it's a 001 you will need to look closely. Some have and others not. The 001 board was a 120gig (4 x 30gig), did not have the fix. The 003 was a 240gig (4 x 60gig) had the repair. I have not had the chance to look at other models to see if its just the 120gig (and maybe 240 gig) model effected.

This problem only show up after you upgrade the HD, on 120 models. I do not have enough info to know if other models are effected. So far all of the models 240gig and greater that I have seen, has had the mod (003 boards).

[blue68f100]

Updated post original post.

Found a second area (FB2) of repair, done the same way, with what apears to be the same componets. It is located to the left of the Promise Chip. Will update the photo showing the areas.

The resistor measures 10 ohm

[sunfriek]

on my Dell 705n(snap 4100) its board is 70715048-006 which has this mod applied..

mine is 60Gx4 via Seagate barracuda ATA IV (st360021a) 240G.

best regards

[bsmith]

Not to ask a stupid question but, what point is there in adding a resister in series with a ferrite bead. Generally an smt ferrite bead has a dc resistance in the tenths of an ohm. I could see maybe adding a small cap to it to further block unwanted frequencies. BTW I have a 4100 70715048-001 without the mods.I suppose I will try and find a couple of 10ohm SMT resistors in my stash and add them as I'm getting ready to go to 4x160G from 4x30G

[wgong]

Originally Posted by sunfriek

on my Dell 705n(snap 4100) its board is 70715048-006 which has this mod applied..

My 705N also has both mods - it is even newer (70715048-007). It was also a 240GB unit.

P.S. I cannot enter my 705N Service Tag into Dell's support page. I suspect it is because the original unit failed and was replaced under warranty and no one updated the database to reflect the new number.

[blue68f100]

Originally Posted by bsmith

Not to ask a stupid question but, what point is there in adding a resister in series with a ferrite bead. Generally an smt ferrite bead has a dc resistance in the tenths of an ohm. I could see maybe adding a small cap to it to further block unwanted frequencies. BTW I have a 4100 70715048-001 without the mods.I suppose I will try and find a couple of 10ohm SMT resistors in my stash and add them as I'm getting ready to go to 4x160G from 4x30G

Will try to explain why it is needed. The resister is used to slow the response of the coils or damper them. Apparently with larger drives it was not settelting down. Like they were over driving the coils. Causing scatter, which in turns corrupts the raid5 array.

Recovery is very difficult in this condition because the drives are not put the data on the disk where it is suppose to be.

When you are trying to sync 4 drives timing is critical. Even a drive reallocating a bad sector with smart tech can cause problems. It best if you can run a utility to check the surface and have it update the table prior or soon after put in service. This way most of the bad sectors will not be use.

[bn2000]

Mentioned repairs can also be found on MB 70715048-005A.

Nevertheless heavy problems with Snap4100 upgraded from 4x40GB to 4x160GB (4x130GB matched).

System hangs on W2KSP4 while transfering hugh amounts of data to Snap and results in manual reboot by reset button.

Is there any idea ?

System configuration is:
SnapOS V3.0.566
Bios V2.4.437
Hardware V2.2.1

[blue68f100]

Please repost on a different thread. I'm trying to keep this one strictly for MB mods.