PCB BIOS and FLASH swap

[zvit]

I won't be putting all emails here but when there's progress, I'll share it here so others might benefit if they have similar issues.

**

Ok, we're progressing!

I used your live CD without virtual. Tested a good drive with NULL destination and scan starts and works great.

Using the same settings you gave me for the bad drive started better than before, at good speeds, for a few seconds (I say better than before because unlike before, the first line in the viewer was green!) But then, like before, it got stuck when it hit the bad head. It would have continued every 10 minutes as before but I didn't want to wait. I tried to uncheck "Read bad head" but that didn't work either. (Why would it try to read the bad head when I uncheck "read bad head?")

So all we have to do (before I convert to SATA) is somehow get it to read as much as it can from around the bad head at good speeds. How can we do this?

Here's a short video screen-record of the attempt with the bad drive:
https://youtu.be/qrvRzujmYaA

Attached is the log from testing the bad drive (called test-4)

Thank you very much!

Attachments

test-4.txt

(6.52 KiB) Downloaded 633 times

[maximus]

Now that the direct USB mode is working, we can try changing the timer settings. I had you turn some of the timers way up for troubleshooting, now they need to be turned back down. Set the soft reset to 1000ms, the hard reset to 800ms, the reset timer to 8000ms, and the general timer to 120000ms. See what happens with that. If that doesn't work, try the options below.

It also looks like the bulk only reset is not doing what it needs to do, so you you can try disabling it and just sticking with the port reset. And now that it is not virtual, you could try enabling either the soft or hard reset, although I still suspect that the drive will reject those reset commands as seen previously.

[zvit]

IMPRESSIVE!!! The timing settings seem to be right on the money! Takes a very short time to skip over the weak head and data is read at 85MB per second, which would scan the disk for phase 1 in a little over 3 hours.

One issue left that needs to be solved.
Every 30-40 seconds, terminal shows a "Connection timed out" and disk gets disconnected and I get a popup saying "Failed to perform identity device command." When this happens, it won't reconnect unless I physically disconnect and reconnect the USB cable (drive is connected directly to USB port on back of mb with short USB cable - no hubs used.) Then I have to again select it as the Source drive in the software. If I don't select it again as source, it won't connect, even though the drive is connected again.

I played around with various combination of the 4 enable\disable reset options but nothing helped.

Note: If I try to select it as source again BEFORE I physically disconnect the drive, one of two scenarios happen: (probably depends on the combination of reset options I had - but I didn't think to check that at the time) Either the device is missing from the drive list (obviously), or the device is shown in the list but instead of full name (ex: "Western Digital Passport 075A") it just shows "075A". Either way, it won't accept it as a source until I physically reconnect it first.

Attached test-5 log. (Since this was progressing nicely, I don't think there will be a test-6; I will just continue working off the test-5 image.)

Ok, so how do we proceed? If we can get the scan to be continues, that would solve it.

Attachments

test-5.txt

(15.42 KiB) Downloaded 723 times

[Lardman]

One issue left that needs to be solved.
Every 30-40 seconds, terminal shows a "Connection timed out" and disk gets disconnected and I get a popup saying "Failed to perform identity device command." When this happens, it won't reconnect unless I physically disconnect and reconnect the USB cable (drive is connected directly to USB port on back of mb with short USB cable - no hubs used.) Then I have to again select it as the Source drive in the software. If I don't select it again as source, it won't connect, even though the drive is connected again.

This is where a power relay would help you out. You can only go so far with resets sometimes a physical power cycle is just needed. Not sure what the power cycle would do with the usb / drive ids as I've only used it on sata.

[zvit]

This is where a power relay would help you out. You can only go so far with resets sometimes a physical power cycle is just needed. Not sure what the power cycle would do with the usb / drive ids as I've only used it on sata.

Are you talking about something like this? (Don't laugh - this might have nothing to do with what you're talking about )
If not, can you tell me what a power cycle device looks like\is called, so I can research it?

[Lardman]

Yes - That's sort of thing, There are guides on the hddsuperclone website under "USB relay" and section 2 "relaying USB" (I can't post links) showing you how to wire them up. Best to check with Maximus about the usb id effects as I said I've only used them for sata they're not exactly expensive to test.

[jermy]

May I ask what type of files you are after ?

[zvit]

2: that pcb is just fine, although it has different serial flash on it. (only the package is different).

I'm reading up on the U12 and U14 chips and I'd like to ask a follow-up question about these chips for better understanding and clarity.

I'm clear about the main concept:
1. U14 contains and handles the bridge firmware.
2. U12 is for the drive MCU (so called ROM chip.)

As you saw in my previous screenshots (I'll post new ones below), my PCB has two chips; the U12 is a bigger one and says "FLASH" on it. The U14 is smaller and says "winbond." The donor PCB has two identical chips for U12 and U14. Both are nameless but have identical serial numbers.

I would like to make some statements based on the above information; please tell me if I understand correctly.

1. Even though the U12 and U14 on the donor PCB are identical in size and have identical numbers, they are still different chips because as mentioned, the U14 handles the firmware and the U12 is for the drive MCU. (I'm guessing the components inside the chips are different?)
2. Even though my U12 chip is larger and says "FLASH" and the donor U12 chip is smaller, they are the same ROM chip and are the ones that would be swapped.
3. Both the U12 and the U14 are technically both BIOS ICs (ROM chips) because they both contain non-volatile ROM memory, although each one's ROM is for a different purpose.

If all three statements are correct, I'm good!

MY PCB


DONOR PCB

[zvit]

May I ask what type of files you are after ?

The 1TB drive was about half full will all types of files; pictures, video, documents, etc. I'm trying to get as much of them as possible.

[jermy]

You might get away wite ducuments and picturs but videos hmm mmm.... not so much (depends of the size of the videos)

[pepe]

The flash chips do not handle anything. They store firmware components for the bridge (U14) and for the main MCU (U12). They can be of different brands, even differenc capacity, the important thing is the data they store (and present when the controller demands it).

pepe

[maximus]

I may make multiple posts to reply to some things, as I will likely think of more things after posting. Please check for additional posts after any post I make to ensure that you have all the data I am trying to provide.

The direct USB mode is a different kind of animal, and may need some timer setting to be increased to keep from locking up. You can try increasing the times for soft reset, hard reset, and the reset timer. But if that doesn’t work, it may need to be power cycled when it locks up. Please refer to the following links for information about using a supported relay with hddsuperclone.

http://www.sdcomputingservice.com/hddsu ... /usb-relay
http://www.sdcomputingservice.com/hddsu ... -relay-usb

Also, there was information in another thread about another relay device that is designed for controlling power to USB devices, which is of interest, but I have not tested it. It is possible to call terminal commands from hddsuperclone to control a relay, which is explained in the user manual. Here is the link to the talk about the other relay device.

viewtopic.php?f=7&t=38188&start=92

One other thing that may help is to enable skip fast in the clone settings. That will make the head skipping algorithm more aggressive with less reads in the bad head, which should lead to less power cycles.

[maximus]

Also another thing to try is go to advanced settings and change the "action to perform on major error" to "call command", and enter the command "sleep 5". That will try to disconnect the drive, sleep for 5 seconds, and then reconnect and resume. This sometimes helps with SATA drives, but not sure about USB.

[zvit]

I may make multiple posts to reply to some things,

I already ordered a USB relay yesterday. I will try your suggestions and update. Thank you.

The flash chips do not handle anything.

The flash is U12 on MY pcb. I didn't say that U12 handles anything.
Anyway, can you confirm that these are the two U12 chips that would be swapped? (Even though mine says "flash" and the donor one doesn't)

[zvit]

Ok, after a few trials and errors with the timings, I got it to scan without disconnecting. Took about an hour to scan 10% so I'm assuming that phase 1 will take about 10 hours.

Since the drive was only half full, I'll only do phase 1 for now and see if the extracted data is satisfactory enough. Even if it is, I'll still try phase 2 after I receive the usb relay amd sata pcb in a few weeks, just for my learning experience.

@maximus, if I have a project that has "finished" as the "current status" (so it won't continue scanning - pops up window, "scan is finished for the current project") is there a way to get the status back to "phase 1?" I tried changing it in the project file but it didn't help.

[pepe]

Zvit, yes, U12 is what you need to move if you wanna swap to SATA pcb.
You could simply 90091e for the PN of those chips, perhaps you would have found out they both do have the same PN (Pm25LD020), perhaps you have had been able to find out from the datasheet they are eventually the same chip in a different case. Do your homework, please.
Dunno how other people are with this, i get tired of such spoon-feeding threads. The amount of time you and those nice people replying here invested into this project is already about ten times the time the recovery would take. Wasting of time (mainly others')... And you are pretty much lucky your drive still did not kill any of its heads and surfaces during all these efforts.

pepe

[maximus]

Ok, after a few trials and errors with the timings, I got it to scan without disconnecting. Took about an hour to scan 10% so I'm assuming that phase 1 will take about 10 hours.

Since the drive was only half full, I'll only do phase 1 for now and see if the extracted data is satisfactory enough. Even if it is, I'll still try phase 2 after I receive the usb relay amd sata pcb in a few weeks, just for my learning experience.

@maximus, if I have a project that has "finished" as the "current status" (so it won't continue scanning - pops up window, "scan is finished for the current project") is there a way to get the status back to "phase 1?" I tried changing it in the project file but it didn't help.

Phase 2 is a compliment to phase 1, and should be completed for the best recovery of the good data in the good heads. Not performing phase 2 will leave good data unread. That is because phase 2 catches the data on the back edge of the bad head that is leftover from skipping.

To reset the status there is the function in the tool menu to reset the status. There is also a function to reset the log, which resets more things. Read what they do in the user manual.

[zvit]

Phase 2 is a compliment to phase 1, and should be completed for the best recovery of the good data in the good heads. Not performing phase 2 will leave good data unread. That is because phase 2 catches the data on the back edge of the bad head that is leftover from skipping.

Phase 1 finished but scan said 82% and image is full size: 931GB. I am guessing that the 931GB is pre-allocated space reserved for the image.

Since the image is 931GB, leaving only 5MB free space on disk, if I open the project and choose this image as destination and change the scan to phase 2, I'm wondering if it will be able to add data to the image or will it say that there is no more space on disk? (If I were to guess, I would assume that phase 2 knows to write data without additional disk space, in the pre-allocated space - the space marked by phase 1 that is free to write to.)

Anyway, I'd rather not run another 15 hour scan on a weak head. I will first scan the image with R-Studio and see what it spits out. If the data is satisfactory enough, I will then wait a few weeks for the USB relay and SATA PCB to do phase 2 for educational purposes alone, without having to fear if a head crash happens.

You could simply 90091e for the PN of those chips, perhaps you would have found out they both do have the same PN (Pm25LD020), Do your homework, please.

No need to 90091e; the PN (Pm25LD020) is clearly the same on both chips in the screenshot. I saw that already; I was confirming that even though one said flash and one didn't, that they were identical chips in different cases, which you confirmed.

However, I am still questioning (out of curiosity alone) the U14 chip on the donor board which is identical to the donor U12, with the same PN number. We've already established that we don't take the donor U14 - got that. But if that's the case, that means that the donor U14 is NOT the same chip as donor U12 (otherwise you'd say that I can swap either one.) So isn't it strange that two different chips (donor-U12 and donor-U14) would have identical casings and same PN number?

Also, believe me, I didn't plenty of Google research too. Since hddguru is such a popular forum, Google brings us here on many DR questions. So if you then send me back to google, I'd be doing a forever looping dance

Dunno how other people are with this, i get tired of such spoon-feeding threads. Wasting of time (mainly others')...

That's the beauty of a public, open forum; we don't force you to spend time on us. You choose how much time to spend here. You decide who you enjoy helping and ignore the rest. That way you don't "waste" time on things you don't enjoy. I enjoyed helping people on https://www.sevenforums.com/ for a long time, until they reached out to me to become part of their admin team, which I did. The only reason a left the team after a year is because I didn't enjoy the required 1-hour skype meeting we admins were required to attend to. I didn't feel it contributed enough to make it worthwhile for my time. So I continued spoon-feeding as a "private citizen."

When I volunteered as an ambulance driver for 20 years, not once did I turn to the guy in the stretcher and say "Would you stop complaining about your pain! We'll be at the hospital soon and you can complain to your doctor!" Volunteer work is just that - volunteering; you can walk at any time.

And there's nothing wrong with also making some money from volunteer time. As a suggestion, maybe create some DR software like maximus did (maybe something with SD card functionality.) I will definitely be purchasing his lifetime license so his time paid off and I would purchase from you too if it were something good. maximus handles his time by limiting his responses to one a day - which is smart. Learn from him, maybe do something similar.

The amount of time you and those nice people replying here invested into this project is already about ten times the time the recovery would take.

First, I already said that I'm doing this for a learning experience. Also, I work in the wedding industry and don't have much work now because of COVID so I am spending my time on this drive instead of watching Netflix. No one is obligated to reply so I am not wasting time of others.

And you are pretty much lucky your drive still did not kill any of its heads and surfaces during all these efforts.

Agree. God must be watching over me. I am grateful.

[zvit]

UPDATE!

R-Studio seems to be able to extract all of her 513GB of data, folder structure is intact! Still a few hours left for getting the data out of the image but I checked what already copied and all images and video play fine; documents (word, pdf, excel etc) open fine!

I am amazed at how much data was extracted with a bad head. She must have been very lucky that since the drive was only half full, probably most of her data wasn't on that part of the drive. Who knows, maybe that head got weak years ago and wasn't being used to write data at all, and it finally shut down. Also lucky that the head didn't crash during scans!

I am very grateful to all of you who helped me here, even you pepe!

I want to give a shout-out to maximus who made a great program and gave great support. The $20 spent was worth $1 million to the penny! I will definitely be getting the life-time license for the HDDSuperClone. I doubt that with over 3000 resets, that ddrescue would have been able to help me. Let alone the very pleasing GUI HDDSuperClone has!

As I mentioned earlier, I will still be running phase 2 after I get the USB relay and SATA PCB for educational purposes, so I will probably be back with questions (sorry pepe )

May God bless you all and once again, thank you for all of your support.

Zvi Twersky

[BGman]

I am amazed at how much data was extracted with a bad head.

The head is not bad, the surface is scratched. Can you make the difference?

Normally it takes less then 24 hours to recover the data from such a drive by the way proposed by @Northwind..