Blown diode

[Lithops]

Hi all!

I had a diode blow out on my old trusty Seagate Barracuda 7220.7 200Gb being implemented as a cased external drive. (Image attachment included).

Now I've been reading into what can be done, but i have a few questions:

1. The drive is supposed to be functional when the diode is removed, right? (Actually it removed itself from the board when it failed permanently and i haven't dared to try it as is...) I just wouldn't want to risk any overvoltage issues possibly destroying anything further since a replacement board is not so easy to come by. I do however have a 12V power supply that came with a IDE/SATA to USB-adapter so it shouldn't cause an issue. (The damage happened with another power supply, which was oddly enough rated 12V output as well...).

2. Is there anything else that might have fried in the process? The board looks fine on both sides though, besides the pads of the failed diode as the board coating started to smoke just before i rushed to unplug it.

3. In regards of replacing the diode, anything specific that i should know about?

The HDD was coming to the end of it's service life as it started to have power issues and i was getting ready to migrate my files onto another external drive when it failed and started smoking the second i plugged it in. Now my files are stuck in it and would really appreciate if we'd get it going again for a tad longer. In my country they rip you off even worse than many other places to recover files, so that's not an option...

Thanks heaps in advance!

Attachments

Toast

[fzabkar]

Make sure your PSU is producing the correct voltages at the Molex connector. Then test the PCB on its own. You don't need to replace the diode if your PSU is good. I can help you with voltage test points if you upload a detailed photo of the complete PCB.

BTW, did the "bad" PSU have a mini-DIN plug (+5V and +12V)? If so, then be very careful. The pinouts for these connectors are not standard.

[Lithops]

The power supply used when it broke down came through a circuit board inside the case. Pic attached. The thing was plugged into a cord in which the transformer block stated to give out 12V. I think it came with the device.

I'll photograph the PCB whole when i can get to it. (now tinkering in the garage at another location...)

Attachments

[Lithops]

Oh and just to be clear (can't edit anymore the above post) the board in the image is only similar to the one used. Being stripped from anything useful before recycling. So that you know about the bent connector pins and other missing parts

[Lithops]

Okay. Here again. The PCB in full. Also dug up my multimeter. Been a while since i've needed it.

Also checked the other power supply, the one that i've used successfully in running other hard drives as "stand-alone" with the IDE/SATA to usb-adapter and at least the wire coloring is a match with the one i used when the diode died. So there should be no trouble with pinouts, i guess?

Attachments

[fzabkar]

At the 4-pin Molex connector, confirm that +12V is present on the yellow wire, and +5V on the red wire.

Measure the Vneg, V1, V2, and V3 voltages.

Seagate Barracuda 7200.7 ST3120022A hires PCB photos:
http://www.hddoracle.com/viewtopic.php?f=112&t=109

Attachments

TVS_diodes_2.jpg (86.37 KiB) Viewed 14604 times

Vneg.jpg (181.19 KiB) Viewed 14604 times

[Lithops]

Quick update.

Found the other supposedly good power supply, BUT it seems that i've ran into a tripped diode (or some other short circuit) before. Can't really remember since the thing has been sitting unused for years. The 12V section is dead, reading 0 volts. 5V gives out 5.39V (as if it matters). Must've had a shorted diode before which has blown the fuse on the transformer side.

Have to get me a new one and continue from there. Well, I was going into town anyway...

[fzabkar]

ISTM that the +5V supply is generated on the USB-SATA bridge PCB from the incoming +12V supply. This would suggest that the PSU brick is OK, but that there is an open fuse in the 12V output from the PCB.

Could you show us a full close-up view of the bridge PCB?

[Lithops]

Umm... I know what you're going for but it isn't the case. The HDD is powered directly from the transformer with a cord with a molex-connector and the SATA/USB bridge is a separate contraption.

The bridge PCB in my previous post is not used now. It was the failed one. This is just a kit bought especially for using HDD outside of any casings and PCBs in between...

Attachments

Like this...

[Lithops]

Okay. This project is on hold until i have found a power supply. None of the 6 places i asked didn't have any even scrapped. Have to put an order in tomorrow.

Unless i use my old desktop for power. Wherever that is and if you even can fit to measure anything in the cage...

[Lithops]

Alright, now i'm cooking. Finally got a working psu. And dug up how to run it solo on the table.

Before i start poking with my meter, let me just refresh my memory...

The black probe can go to either ground (pins 2-3) at the molex connector end on the board?

Anything else i should know about before i proceed?

[rogfanther]

First, put pictures of the psu here, before connecting it to anything.

Also, you will need to measure voltages in the psu connector, *before* connecting it to the board, either alone or afixed to the hdd.

[Lithops]

It's just your basic ATX psu... Nothing more to it. And yes yes, of course checking first what comes out the Molex...

Attachments

[Lithops]

Molex gave out 12.01V/5.06V and that's all the time a have for today on this. It's a pain to be dependant on someone else's transportation...

But nothing exploded while the pcb was hooked in, so that's a good thing. Readings to follow tomorrow.

[Lithops]

Here's what I got:

V1 1,49V
V2 4,90V
V3 4,90V
Vneg -4,92V

What say you?

[fzabkar]

Sorry, did I mix up the +5V and +3.3V test points?

V1 is Vcore. Both V1 and Vneg are OK.

[Lithops]

Well, i measured according to the images above...

On a side note, since I was not sure how you intended to measure V2 and V3 as they have two contact points, the 4,90V is what comes to any one of them single. Between the two points in each voltage is 0. If that makes any sense...

[fzabkar]

The voltages should be measured using a ground reference, eg a screw hole.

I think I may have incorrectly identified the voltages on pins 2 and 3 of the "V33DP" ICs. Are the voltages reversed, ie is the voltage I have identified as +5V (pin 2) really +3.3V? If so, then all is OK.

[HaQue]

I have been told you shouldn't run those PS's without a load. It could be just from years ago and someone deciding that was advice, with no scientific basis, or could be true. Ive done it plenty of times without any problems, so YMMV

[rogfanther]

Cheap chinese stuff will probably burn ( they do it even when correctly conneted to a computer ) .

Most psus from respectable brands have minimum load resistors in them, so that regulation works even without *external* loads connected.

Of course, YMMV and all of that. I usually put one or two old hard disks as load just to be on the safe side.