hot swapping pcb

[chipmunk]

I have a maxtor diamondmax 9 drive that is not showing its capacity.
The PCB is OK, I verified that.
Now I want to hot swap the PCB.
Is it (absolutely) necessary to power down the motor before hot swapping or can I get away with leaving the motor running?
I assume it would be wise to power the drive from a separate power supply to handle the power surge in a more elegant manner.

[poehere]

Did you read your manual it tells you all the critial information on PCB swaping. Even when you do a hot swap it tell you to turn off the power. If you want to fry the board leave the power on - maybe get some training before attempting to do this find out why you can not see capacity of hdd it is in the books to tell you what the problem is and how to go about fixing it sorry but I am as new to this as you are and I spend more time reading than I do repairing for the moment. I want to know the most I can before attempting to screw up something that does not belong to me I know for a fact in every chaper of the manual for each drive family it repeates the procedue over and over again and each time it tells you what to look for. It will also tell you how to change out your board if it really nees to be changed out. Maybe I can suggest that you spend at least 3 weeks reading and doing some research before trying to do this without the knowledge necessary to do it. I would rather for the moment leaves this to someone who has done it more times than me than try to do it and make a mess. It is your money to buy the parts and if you mess it you should not charge someone for your mistakes. I am very sorry don't mean to be mean but if you check and see how many people make same error as you it might give you an idea to do research before attempting to copy their errors.

[BlackST]

What?!?

[cryoborgofthevenus]

can I get away with leaving the motor running?

Not a chance

[BlackST]

Try it ... You'll see the result... Likely: CLUNK! POOF! And some smoke...

[thatdellguy]

Chipmunk these guys are not telling the complete truth. The trick is to drop the drive from ten feet while its running to stop the motor...then you can hotswap the pcb...

[Harvey]

hmmm, BlackST, very funny

[cryoborgofthevenus]

Chipmunk these guys are not telling the complete truth. The trick is to drop the drive from ten feet while its running to stop the motor...then you can hotswap the pcb...

uh-oh , secret revealed.....

[BlackST]

Naaaaah... it is noisy and if the drive opens it will make a mess (platters rolling all around the room, etc.). To stop the motor safely, the procedure is the same for the clicking Barracuda. I'm sure it will stop either if the drive was clicking or not.

solution-for-clicking-barracuda-t11414.html

Another way to stop the motor safely to do smart and successful hot swap:

Connect a 3,5 HP diesel power generator to a 3,3 V stabilized power supply. Conenct it to the RAM only (to keep data alive, it's a static ram, so when you'll give power to the board again, the microprogram will restart from the point it was stopped , put enough diesel fuel on the tank, check oil if needed, turn on the generator, give power to the drive, remove power while keeping the cords on, do the swap, give power, remove cords... voila !
You have only to find suitable crocodile clips for TSOP package

And now, please don't say we don't help the people starving for recovering their por... war... ehm, uhm... their stuff !

Have a good day !

[cgallery]

I've done hot-swaps w/o stopping the spindle.

[chipmunk]

Hi gurus (skip the pranksters)
I do understand that it is wiser to shut the motor down before hot swapping the PCB.
BUT I would like to understand what goes wrong if we do a truly hot swap.
Option 1: Induction causes voltage on the driver chip to spike and fries the chip (Isn't it protected somehow?). If this is the problem then soldering some stategic diodes on the board before hot swapping should make it immune.
Many years ago I did this on an Apple II computer so I could hot swap the interface cards. (Save a lot of time in testing...)
Option 2: The drive goes very nervous from seeing that the motor is not responding to what it is telling it to do.
Thanks for sharing your knowledge!

[cryoborgofthevenus]

Can explain to us exactly what are you trying to achieve with your idea ?

You already know the risk and also know the correct procedure to do it "safely"

I just don't get it....

[chipmunk]

... because I am trying to understand and to learn.

Also I only know how to turn the motor on and off in the case of Seagate but not with Maxtor for example.

[cryoborgofthevenus]

You mean Z and U using Seagate terminal ??

Well, take a look closer to MHDD commands. Boot your system with it. And carefully check documentation.

I think you need to do your homework

[BlackST]

Simple : many drives give an impulse to the VCM to unlatch the headstack that is kept in parking position with a small magnet and a latch. So without this initial pulse , headstack won't move from parking position.
AND the driving of the spindle motor is 3-phase and closed loop current sensing, plus some voltage is derived from this section with a DC-DC converter. Result : you remove the load, the combo will try to compensate using the remaining windings (yes they are not so intelligent sometimes) and it will likely burn - the IC or something related like the diode stack or the MOSFET. (CLUNK ! POOF !! - smoke... - )
Don't wanna think about the headstack with the VCM bias/driving current removed without properly parking (that is done in case of stop or power failure).... (SCREEEEEEETCHH !!!).
FINALLY, some flags will be obviously waived in the drive moemory (emergency / failure) , so it won't work definitely.

(OH !! THOSE BEGINNERS !!!! )

[chipmunk]

I am practicing hot swapping the PCB on a Maxtor DiamondMax Plus 9 (Calypso) drive.
I let the drive go to sleep before doing the swap so that the motor is not running.
So far this is not working.
I connect the first drive to the computer (Mac - yes I am a macintosh person). I let the drive start and confirm that all is working. The I wait for the drive to go asleep. I swap the PCB onto drive two. I wake the drive up again and try to read from it (using Data Rescue) but nothing can be read. I then swap the PCB back to drive one in same manner and confim if that (still) works. Reading drive one after all this still works. So I can conclude that the procedure at least does not blow the PCB or drive one up.
I did this several times now but I can not manage to actually read any target drive.
Does anyone have any suggestions on what it is that has to match between these two drives for this to work?
What do the letter codes on the drive mean? (N,M,G,A,)

[BlackST]

A-D-A-P-T-I-V-E-S.

[msurgeon]

First of all make certain that the the third line on the large chip of the PCB (i.e. Ardent C8C1-840UK) matches between donor and patient. If they don't, then hot swap won't work. However, there is more to a hotswap that just getting the drive to be recognized by the BIOS. There is drive-specific information read during startup and stored in RAM that is different between patient and donor. Without professional-level equipment this is not an end-user project.

[chipmunk]

The code on the Agere chip matches (C8-C1 in all cases).
I understand that there are adaptives in play, but what is going on here? WHICH adaptive thing is stopping this from working?
Bad block tables will screw the data up (a bit) but should not prevent reading.
Number of heads and platters should match (that is obvious).
The PCB does recalibrate something, because when I mount it back on the original drive it reads very slow at first but picks up steam after a bit and returns to its normal speed. So it did adapt something to drive two and is even willing to adapt back to drive one.

[msurgeon]

T
Bad block tables will screw the data up (a bit) but should not prevent reading.

Not true. Factory defect lists differ for every drive. One of the key firmware elements is the translator module which converts those defect lists into a properly indexed map of the data. In a regular hot swap you are trying to read one drive's data with another's translator module. Not going to work.