Can I Retrieve the DATA from this Sony Digital Voice Recorder?
Вставка
- Опубліковано 15 кві 2023
- Here we have a Sony ICD-PX312 digital voice recorder sent over from Malta by Kate and Theo. It came up with an error whilst recording and hasn't worked since. There is a voice file on there that they would like to get back, can we get this working again and retrieve the voice file? It isn't recognised by a PC when plugged in via USB so I can't use data recovery software to pull off the files. Can the hardware be fixed?
Let's find out.
If you would like to support these videos, please click here / mymatevince
Remember that this is just for entertainment and I am not an expert in these repairs. The processes in the video may not be the best way, the correct way or the safest way to fix these things. I do love fault finding and trying to fix broken things, so I hope that comes across in this 'Trying to FIX' series. Many thanks, Vince. - Розваги
Spoiler Saver
This one was painful as it didn't AND still doesn't make any sense to me. Hopefully once the comments come in, the mystery will be all cleared up and maybe I can do a revisit 👍👍
Heat the whole board up and see what happens.
What about giving it 3 volts where you think it should have the power and take data down that way .
It would be faster to move 3 chips to a new board. Even if that takes you 2-3 hours, its still much faster than what you are doing now.
Just familiarise yourself with the BGA memory chip, and use minimum viable heat for minimum viable time on it.
You will need to reball it.
Oh, if you are scared about the BGA memory chip - don't be. Just unsolder the main chip first like you'd normally do and then use heat from the underneath the PCB, that way solder balls melt before the chip gets too hot.
Then mount these chips in reverse order.
@@DIYRepairHour He needs an entire new kit of bga stencils, support to hold the chips, bga soldering balls of 0.3, and other things that I can't remember right now.
Vince, Most electronic components have a power management chip or current/voltage sensing circuitry. The reason why the transistor at 10:43 was switching off is because the aforementioned circuitry is detecting an overload somewhere and thereby switching off the device this is to protect the device from further damage.
Generally NAND Flash like in this dictaphone will be formatted with a filesystem, much like on a desktop system. It's similar to how SD cards are supported, meaning that the folder structure and meta data is stored on the Flash device.
This would mean that swapping the Flash chips between both devices would theoretically work, but NAND Flash has the annoying tendency to rapidly lose its stored charges (and thus data) when heated up. I'm not sure there's a safe way to do this, and it might be easier to dead-bug wire the target NAND Flash IC up to the working board without desoldering the former.
Depending on the NAND Flash interface and file system used, it might even be possible to wire it up to just about any 3.3V MCU for reading out.
Heat definitely shortens the retention time of the flash memory, but not too much (from years, to days at 200°C). You don't want to heat it for long, but it's usually less than a minute of high temperature.
Wiring each data line and power from the bad board to the good board is a great idea (removing all related components from the bad board). I have no doubt that would be successful, seems the only way it wouldn't is if the flash itself is bad or it has a cracked solder ball which is not fixed by pressing on it. Quite the pain in the butt though! One would have to desolder the good chip and trace out every important net to find a good place to attach to. Sheesh. Maybe if the recording is VERY important 😄
Too bad we never have gerber or design files for stuff like this. Doing this kind of thing with the original design files brings it more into the realm of "stuff we do all the time when debugging a new design".
@@ilya_mzp This is incorrect, especially for modern MLC flash. There are many papers on this subject from the field of forensics. I suggest checking out "Improving the reliability of chip-off forensic analysis of NAND flash memory devices" (DOI 10.1016/j.diin.2017.01.011), which showed severe bit error rates with a desoldering temperature of 250c. The case with this recorder may even be worse because the storage strategy used by the MCU is likely to be much more naive than an SSD controller.
Given that such work would not even be attempted unless the recording was very important, Maya's suggestion is much safer than removing the chip.
Is there a possibility this is a NOR flash?
@@kriswillems5661 Highly unlikely. NOR Flash is very expensive per bit, to the point where it's competing more with MRAM than with NAND Flash.
NOR Flash also has similar issues with data retention at high temperatures, due to it using the same fundamental storage method.
Watching this video on my phone in bright sunlight I had to strain my eyes. Then a cloud came and that's when I realised: it wasn't the sun, it was the yellow mat all along.
Hahaha, the yellow mat strikes again 😂
@@Mymatevincewhere did you bought it?
🙃
Remove NAND and try to recover the data directly. It's likely a standard file system on the NAND so if you just get a raw dump then data recovery software might be able to find the sound files. I highly doubt there is any encryption of data for something this old.
If you are going to swap, just the NAND should work, you shouldn't have to swap the CPU along with it (that might be the fault anyway). You can even dump the NAND without swapping at all if you wire up the chip itself. There's only like 4-5 points you need to power and read the chip.
Great effort Vince! I've done the 5am thing in the past too!😂😂😂 I'd think any user data would just be stored in the flash chip including the file indexes. I don't think the unit would store that in the microcontroller, so if the flash isn't corrupt or at fault, just swapping that might work...
I was thinking the same 😉
Thank you Mick
Hi, Vince, Theo here! Firstly, thank you for your perseverance on this, I can see that it wasn't easy at all! Not sure if you saw the device schematics I found, maybe they'd help, so I asked Kate to send it over to you. Anyway, I found this video fascinating and I'm reading all of the comments to see what others suggest. Keep up the excellent work, you're a legend! 😊
Nice one Theo, no I haven't seen them as of yet. I will check my emails later. Once all the comments come in over the next few days I will have a good read through and then see if a revisit video can be done 👍👍👍👍
@@Mymatevince No worries at all, Vince! You do what you gotta do 😊💪
Absolute fault finding endurance, well done !!
Vince is amazing as always, kudos for your total dedication, wow ❤ I love your videos and I can see how hard you are working, no wonder people find your channel one of the best things to watch! Keep up the good work 👏👏👏
I actually have one of these. They're pretty good recorders for the price
I use it to have an external audio source for filming videos
I used them for a number of years for field recording. You’ll eventually want something more advanced, but they’re great for folks who want something better than camera audio.
Great endurance test Vince. 😊😊
I had one of these, slightly older in silver ICD-P620, loved it before I switched to using a smart device. Was glad I got all my recordings off it before something like this happened. As much as I love the focus that a dedicated device like this gives (no temptation to look at some other app), there is something to be said for recordings that transmit via wifi straight to a computer with version control.
Another great video. I'm pleased the yellow mat came out because that blue one looks really skanky now. Mind you it shows its done years of good service.
Hey Vince ! I really enjoy watching your videos and would absolutely love to see you revisit this with swapping over the chips.
I enjoyed watching your video very much. Keep up the great work!
Can we have a limited edition T-shirt or Mug saying I stayed up for the masochist Sony voice recorder sessions! Fascinating video Vince keep up the good work.
Glad to see your fingers better Vince.
Looking at 8:45, you can see some corrosion on the mini USB and some capacitors, possibly resistors, nearby. I would start from them. Ceramic capacitors have a tendency to short, maybe that's the reasons that voltage is missing.
2$ usb nand reader, take chip out and put it on nand reader to take out data like a usb as encryption chip is also onboard 😊
Revisit...we've come this far... we're so close - I can feel it. The time's spent and I'm all in. Are you with me? I'm with you...come on...swap them...please?
You did give it a great shot. Maybe test by taking off the chip on the working 1 and then putting it back on the working 1 to see if the heat destroys the file
Go for it, there’s nothing to lose in trying 😊
Good job mate 👍
I used to have a couple of these same recorders back when i was part of an amateur paranormal investigation team they were pretty good for EVP
Something you might be interested in getting are some multimeter leads with needle tips. They are much smaller than the common size.
Go for it!
It's a very interesting.
For me it is a reliable recorder for Panasonic microcassettes.
I still used this.
lol.. vince using some of his old BT wire stock to connect the device. Njoying your Vids Vince :)
Wow I jus woke up and this video still going 😊😂😂
Cool video
Give it a go Vince why not 🎉🎉🎉🎉
Swap the chips mate... What have you got to lose... As is... It's no working... You might get lucky... Great effort.. ❤
Mine, despite malfunction, was able to transfer recordings to the computer, and I think I remember installing the Sony application that came with it in order to do that. On the computer, It had it's own little file menu for copying to hard drive along with play buttons and stuff. The malfunction was fixed with a simple "Yes" to the format screen accessed by the unit's menu, of course that wiped all the recordings off the unit. I simply put them back in from the computer back up.
it could be worth a try to warm the entire board to drive off any moisture. I encounter that issue with pid controllers for furnaces regular, amazing what a warmup can do.
It's a shame you couldnt fit this one! I think power management issue somewhere! The flash chip probably holds all info about the files stored on it. There probably isn't a relationship regards data structure and data contained on the ARM chip, but there may be encryption that is specific to the ARM chip fitted?!? Worth a shot to swap the NAND flash over!
I think you should try just the memory chip 1st. The ARM chip is probably the processor and just reads whats on the memory chip. I think kinda they way a computer reads a usb stick. Just a thought. Keep the videos coming. They are my therapy. 😊
I think you should switch out the flash chip. That's one of the first things I would have tried since that's the only part that's important.
The file info is probably stored on the flash, and it's probably just using regular files on a FAT16 or FAT32 file system. No sense in reinventing the wheel after all. Switching out the microcontroller shouldn't be necessary.
It would maybe help to know how important that recording is. Is it like "Moms last words" or "Aunt Helen's cake recipe" ? So you can know how experimental you can be, or if you need to be really careful.
Vince you surely epitomise 'Mad dogs and Englishmen'.
The ARM chip is the processor. It stores no data regards the memory.
Simply swapping the memory chip would work. However there's a risk to this of losing all data (part and parcel of working with with memory)
9 Times out of 10 it's successful.
The filesystem on the memory will be compatible as when installed it would of been formatted as such they will work the same
Yes I absolutely would swap the writeable memory first and test then the main "confuser" and test again. However that may be poor advice. Sometimes an IC is looking for a trigger to generate a voltage. Such as what is produced with the on off switch. In my head I think the triggered output could remain in a constant state until something external produces the exact trigger it's looking for on its input pin to change state. It is possible by swapping components you actually caused the trigger that changed the state of the I/O. For myself, I believe the problem is most likely with the main IC.
I get that, cheers Shawn. I suppose some pins may have latched on (or off) with my replacing of the chips which led to the strange readings. So confusing! Thanks for watching 😎👍
@@Mymatevince it's only a guess. With as many times that power was removed that should have been enough to normalize/reset I/O states.
My other wild guess is you have a bad B- line and something is storing a voltage and not dissipating it.
If you swap those memory chipa you will loose them both, to me it looks like the firmware is broken
@@technixbul interesting, do you think the memory is volatile or by swapping them over damage will occur? Genuinely interested.
The flash memory is likely formatted as exfat or fat32 and their are usb thumb drives that use flash memory in that package. Might be an idea.
After all the testing, I wonder if a reflow oven would have helped. You did mentioned the possibility of a dislodged solder ball. Lead free solder can crack, decay.
I would definitely try changing the memory chip first. Most probably it works like a computer's SSD drive and stores the file system on itself. Another suggestion I second is to take out the memory chip and then insert it into an external reader. This way you have to desolder/solder the chip only once.
Memory chips look like different sizes to me. 23:32
@@cfq.tufanuf7601 But I think the board layout will be the same, I think if you swap both chips it should work
Vince is not up to doing BGA work yet
@@UserUser-ww2nj My suggestion is not reballing. Just remove the memory chip and stick it in a nand reader to get the data
@@tolgainci Vince doesn't have a nand reader, he would probably not have use for one other than for this one use so it's not worth the expense to buy one
Revisit for SURE. I wanna see you swap those chips.
I was certainly going to ask you (before your coda comments) whether it might be worth trying to swap the memory (or both) large chips. I think that WOULD be interesting (in terms of how difficult, and also whether it is actually feasible that data can be recovered). And I suppose if one researched enough about the storage chip, it might be possible to jury rig circuitry to force it to "disgorge" its data content. I have to say, I find your "failure" videos just as interesting to watch as the success ones, and I think you might still be able to crack this one.
Yes yes yes yes yes please please do this. I walked out of the room for a moment and came back in and sat you had a second device and was like BRILLIANT he's gonna swap the memory to a working one .. But no.
Vince, I have no doubt on your skills, however my first thought was no way he fixes it. Was a nice try, looking forward to see if you swap the chip. Those variations of Voltage, maybe is due to a capacitor . You took something off the capacitor drains, off etc.
inject 2.8v on the joint where you measure lower voltage on the dead board. Something on the board causing it to drop on the board. Unfortunately what I believe there is a partial short inside ARM microcontroller. If I were you I would replace that chip and give it a try.
@My Mate Vince
Do the board revision differ? The nand chips have a different size.
Hi Vince, that chip with the surface mount inductor connected to it, is the inductor ok, I'm guessing this is a switching psu boost type ic.
on the last chip you check the 4th pin i think is a feedback line, that tell the chip the line is ok or not. feedback line have resistors to their path. if you follow the via maybe you find something. and yes you can change the 2 chip (nand and controller) if they are good you can recover the data
You can look up the datasheet of the ARM processor to check if this controller has internal flash. If not it's probably sufficient to swap the flash. But is the processor actually running? Does it have a clock? Is something happening on the address bus? You have a scope, but I don't know if it's good enough to check these things.
I didn't watch every second of the video so you may have mentioned it, but it seems to me they are slightly differnt units.
If you go to 23:11 the the two large chips are differnt sizes.
Those are the flash storage chips. It's pretty common to have multiple options on a board for different chip sizes. You can even see the solder points for the larger chip around the smaller one. That way they can use whatever supply they have.
Here's a crazy idea. If you really want to recover the data. Transfer the nan flash to the working one. Make sure any flash controllers are transferred as well that way the data can be recovered
If you ever buy one of these, always get one with an SD card so at the very least, if it breaks down, you can still access your recordings.
Good attempt, 5am is above and one-step beyond. Pure guesswork on my part, but looks to be power related (rather than flash etc). The component at 33:44 is a MM3404 regulator and chip enable seems to be missing, though of course this can be due to multiple other components. I'm sure you don't want to hear that a service manual is available (though you may need to brush up on your rocket science).
As a fellow electronics tinkerer/repair/rework person I can only say that they're fn weird, haha. They are just weird.
Transistors deal with doped metals and sometimes billions of gates where anything from heat to a small electric charge to bumping them can make them go from behaving one way to another.
Some components want EXACTLY 3.3v and not a bit more and others have tolerances of over a volt.
Some are fragile like glass, others can be hit with a hammer and still work.
Gremlins, Vince. Fn gremlins.
id say the smaller chip above the arm chip is the controller for the storage.
I think it's the main processor chip the arm chip you could try swapping them first so you don't have to heat the memory chip up at least not as much but then you run the risk of shorting out the good processor if it is something bad on the board either way though swapping the memory will work it's just like swapping a hard drive out on a computer
Definitely do a follow up.
My guess is, that the control information about the recordings is also written to the non-arm chip.
So only the storage chip has to be moved to the new device.
Bit this is a guess only.
If it's a shit or bust situation, you may as well try swapping the memory chip over 1st, then the processor if it doesn't work.
Maybe stick a heatsink on them if you are worried that heat will damage the contents.
I had a similar one (ICD-PX720) and a couple of things worth noting: First, it saved in a proprietary format. Second, the software that decoded that proprietary format ceased to be supported by Sony (trying to install it, it called home and the mothership said NOPE).It must have a built-in self destruct and will refuse to work after a certain date. The device suffered the same fate. It said ACCESS and refused to ever work right again. It's replacement saves in .mp3 format like the other one should have from the start. I think trying to power it by the mini usb port (it was an option) may have led to its demise.
I would give it a shot on swapping the 2 chips.
In my experience memory chips can be quite volatile. Also you'd need to check if their is any encryption or security linked to the main chip and the memory. I mean I guess you could do it, but as said memory chips can be quite volatile so the data may get wiped or chip damaged when you try anyway. Seems like a nightmare.
I'd love to see a follow-up on this one if you could. I still have hope
My guess is that the flash chip also stores the firmware for the device, and that at some point the flash chip got corrupted. Swapping over the chip might not help in that case. I think it also stores the firmware because I dont see another flash chip for that purpose.
You can only take NAND chip from the faulty board and put it in the NAND reader and it becomes like a USB flash drive and successfully transfer data 🙃
Hi Vince, Can we have a demo of the Etephon ? and was it a kit ? Could be interested in getting one ,thanks
I can only put your voltage changing scenario down to heat either affecting a thermal fault on the circuit or a component or components have failed due to heat from soldering... other than that defo Gremlins. As mentioned by others putting the NAND flash onto the other working board (they are fairly resilient to heat in the short term) just to get the data off it. I can't believe you stayed up until 5.00am what's wrong with you?(Actually I can believe it - I've done it more times than I would care to mention 😀). Makes for an interesting video, even though things don't work out. Enjoyed watching. Edit: Also as mentioned below: - I would think that providing the NAND flash doesn't get destroyed, the contents will be read from it by the other device. The flash should be thought of as a non-volatile memory store with similar properties to an SSD, the data is stored and registered on the device and will tell the CPU what is on it when the firmware instructs the CPU to read it. It will have a directory pointing to the start address of each string of data to be read. A Table of Contents (if you will). That is - a file system.
you should only need the NAND chip as works just like an SSD on a pc and as long as you do not go crazy hot with the hot air the data will be fine as long as it was not already damage
It's very unlikely that the SoC has any sort of persistent storage in it other than maybe some EEPROM. Swapping the NAND chip will probably be just fine. The problem is that they are quite large BGAs and getting them off one board and onto another without destroying them is not easy.
I'd simply swap the flash chip over to the other one and see what it does, and if that won't work make an image of it by example putting it on a donor USB stick.
It's very likely that the whole filesystem is on the memory chip, so worth trying swapping
Thank you Mike, hopefully I will do a revisit video on this 👍👍
nice video
I would try to swop the QFP first and see what happens
Thanks Anonymous👍
i would only swap over the memory chip. I think the device will see the files on them if the chip itself is good. The ARM ist the main CPU, so you don't need to swap it.
HI Vince I think it will work swapping the memory chips
If those chips are gonna get damaged because of heat, then how did they solder them in the first place (specially the BGA one)?
I have the 440 version and my problem is that the power switch doesn't turn on the unit when it's held down. It has an inbuilt slide out USB connection that enables you to slide it into a PC USB port and the unit then boots up allowing me access to its storage folders etc etc.
My guess is there's something wrong with the on-board power switch on my device.
It is highly unlikely that any internal storage on the CPU/SoC is used for regular read/write access. At most it would just contain the device firmware. The file system and data should all be on the flash memory chip.
I have one of these recorders the ICD-PX333 and have a 16GB card in it for my recordings in case this happens. Then I'll just buy another one. I don't even bother trying to fix this surface mount crap. There is one of these with a built-in FM radio that i'm trying to find.
Vincent my.. Well not my mate cept yt.. You are a funking god with your aptitude🤯🤯👌👍👍
Your Quad amp vid👌, have a look at N.A.D & Mission all 10/10 amps
I would try Just swapping Flash memory .. Nice new meter DM166 curious why ain't you using that one For measuring ?
Ohh cool I use a similar one as a ‘digital minidisc recorder’ it can record music with line in socket and does not have a silly, unneeded gigantic microphone attached to the top 😀💪 it uses an SD card and keeps the number of recording on it so maybe swapping the NAND over will be the same! Is it worth a try?
what a trying thing vince
Order some of the leads that TheCod3r has - very sharp
Chance it! 🙂
Maybe stick that memory chip on a usb stick with the same style bga chip? Just shootin in the dark here, but nothing to lose.
as long as the memory chip is non volatile you can swap it to the working one to get the data
As others have suggested, remove the flash chip. You can get a cheap nand reader from China so you don't have to reinstall it into the new board
The memory chips look different sizes so you wouldn't have been able to swap them regardless. Maybe the manufacturer changed design because of this or similar faults
I think you might of missed something at time index 35:16it looks like there was a trace damaged - I mean those traces are so darn small, but to your left of your pliers I can see a trace with a dark spot that looks suspicious and might be open, maybe take a good long look to see if that might be the problem and check the rest of the board more thoroughly under the microscope.
They appear to be different memory types so I don’t think swapping the other chips would work in this instance.
The first thing I thought was could you swap the memory over to get the info off it. Dont know if you did do that in the end. Another thing I thought is, what happens when you plug it into a computer as an external drive? Would there have been a way to maybe just read off the memory inside it even if it wont turn on? I mean, usb thiumb drives dont have any power except what is plugged into them and they only allow reading.. Thats probably super naive but anyway.
Hi Vince. My vc is exactly the same as the first 1 u opened an old Sony voice recorder and hasn't worked in years but I need the recordings that are on there. There is no visible damage it just stopped coming on but I remember I left some batteries in and they leaked slightly but again no big damage as I could see.can you help ?kind regards David I'm sure there must be away of retrieving my recordings even if I have to pay .
Why don't you remove the Toshiba flash chip and pop it into a chip reader to retrieve the data that way?
Just use lead solder on legs and remove components with hot air (small nozzle), it will take a second and board will never get hot enough to damage data on NAND. It's not like these chips never get warm under use and lost data as soon hot summer winds start to blow. I know that you like to make your life complicated but but there is enough other things you already do to accomplish that, this one you can skip :))
would definitely try swapping memory chip
You could move them both?
The memory chip has a file system that tells what files are where. The main cpu doesn't know what files are where.