FNIRSI-1013D AS MUCH USE AS A CHOCOLATE FIREGUARD.

[Gannet]

I HAVE A YEAR 2020 FNIRSI-1013D scope which has gone haywire.
The ID on the memory card is 2020 ZFSN3-4

It does boot up but sometimes not.
It intermittently switches itself off.
When it is on I often have no time base running
I have checked the battery it seems ok
I runt it from my desktop PSU and it draws around 1.35Amps at 5VolsDc

I have posted two pics, one with the timebase (Sort of running) the other is just junk.
It would seem to me that there could be a power issue, I dont know.
Would an updated firmware help, I know nothing about Linux.
Where do I start, any help would be most welcome.
Regards all.


If this post is in the wrong area would Mr. moderator move it to the correct one, please.

[pcprogrammer]

Hi Gannet,

welcome to the forum.

It is always difficult to diagnose something like this without having the device in hand, and yours is the old version by the looks of it. The firmware running on yours looks different from what I reverse engineered, with the blue and red traces, so can't tell if what I made for the scope will work on yours.

The schematics of the newer version can be found here: https://github.com/pecostm32/FNIRSI-1013D-1014D-Hack/tree/main/Schematics For as far as I know, the only difference is in the FPGA used in the older model.

Did you read the thread about the scope here on the forum? https://www.eevblog.com/forum/testgear/fnirsi-1013d-100mhz-tablet-oscilloscope/msg4525100/#msg4525100 There is lots of information in it.

I have not measured the current being drawn by mine, but I think the 1.35A is a bit high.

Since your firmware looks different I can't tell for sure, but the newer version is for sure not based on linux, and I doubt that yours is.

A picture of the PCB can help in identifying your model.

[Gannet]

Thanks for the reply,

I have attached a pic of the PCB

Hopefully that will help identify which version I have.

[pcprogrammer]

It looks the same as the one here https://github.com/pecostm32/FNIRSI-1013D-1014D-Hack/blob/main/Pictures/Old%20version/Main_board_old_system_large.jpg

This picture was taken by Dave and he did a review of his version. https://www.eevblog.com/forum/blog/eevblog-1317-$140-2ch-100mhz-fnirsi-tablet-oscilloscope-review/

But his version has the yellow and green traces, so yours is a bit special. Would be interesting to get the firmware out of it. I wrote software to do it but it involves writing to the SD card. Let me know if you are willing to do this, and if you have a SD card reader/writer you can attach to your computer and also how big the SD card is that sits in your scope.

[Gannet]

The board looks the same as the one you pointed me to.
The SD Card is 1Mb
I have an SD card reader.
Can I load the firmware to a new card rather than overwrite my existing one.
Is there any special software that I need to use to copy the image of the firmware to the new card? Where is the lates working version located?
I have seen reference to taking one of the two isolated plated through holes down to ground (See clip), I have also seen other saying, no not that pin the other one, which is correct.
Thanks

George

[pcprogrammer]

Hi George,

The original firmware is not located on the SD card, but in the FLASH ic near the MCU. The SD card is only used to store images and wave files.

The point on the PCB that you are pointing to is to reset the MCU and is not really needed.

The firmware I wrote can be written to the SD card and does not overwrite the FLASH. The same for the backup program I wrote.

The instructions to get these programs on the SD card are based on using linux, of which you pointed out to know nothing about. But with basic computer knowledge you can come a long way. It is not that hard. Linux can be used from a live USB stick or in a virtual machine.

You write that your card is 1Mb, but most likely you mean 1GB, because that is what the early models were shipped with. Mine has 8GB, but that does not matter. The new firmware uses less then 300KB, so no problem with a smaller card.

To make the backup of your scope the binary plus instructions can be found here https://github.com/pecostm32/FNIRSI-1013D-1014D-Hack/tree/main/Test%20code/fnirsi_1013d_firmware_backup/Binaries The actual file to load is fnirsi_1013d_fwb.bin

More instructions on prepping and loading the SD card can be found here https://github.com/pecostm32/FNIRSI_1013D_Firmware

What this does is make a backup of the FLASH and touch panel configuration and stores them on the SD card. I made a video about it.



With the scope connected to the computer you can copy the three needed files to your computer and post them here.

Let me know how it goes.

Regards,
Peter

[Gannet]

Thanks Peter,

I have two old Raspberry PI's, Maybe I'll resurrect these and install Linux, what do you think?

Regards

George

[pcprogrammer]

That is a possibility, but you might be just as quick by installing linux on a usb stick and boot your computer with it. Perform the needed actions and when done remove the USB stick and restart your Windows (I assume that is what you use) machine.

There are lots of examples on the net on how to do that from Windows. Even with the installer for linux mint you can test drive without actually installing it on your computer.

[Gannet]

That sounds like plan Peter, I'll give it a go.

George

[Gannet]

Hi Peter , I downloaded Linux Lite v6.2 and it appears I have made a bootable usb Drive.

Please see screenshot.

I'm a bit wary of trashing my pc so gently does it what next.

Cheers ,

George

[pcprogrammer]

All you needs is a command terminal and access to the file to transfer to the SD card.

If all is well the linux you have running from the USB stick has internet connection, so you can go to the repository and download the file on to the USB stick. Then open the command window with "ctrl - t" or "ctrl - alt - t". At the command line you can then type the commands, but make sure you select the right target.

So first connect the SD card reader/writer with the SD card inserted into the computer. This will mount the SD card automatically if all is well. At the command prompt type "lsblk" and then enter. This will list the block devices.

Code: [Select]

NAME   MAJ:MIN RM   SIZE RO TYPE MOUNTPOINT
sda      8:0    0   3,7T  0 disk
└─sda1   8:1    0   3,7T  0 part /home/peter/Data
sdb      8:16   0 111,8G  0 disk
├─sdb1   8:17   0  55,9G  0 part
└─sdb2   8:18   0  55,9G  0 part /

It will show something like the above. The SD card might show up as sdc ..... 1G 0 disk and the line below will show ---sdc1 .... 1G part .......

This gives you the name of the device you need to work with.

Maybe it is best to save a screen shot of the listing and post it here. (You can use the PrtScrn button to make the capture and then post it here)

We can then take it step by step, just to keep your system safe. But it is about my bed time, and tomorrow I'm a bit busy, so be patient

[Gannet]

Thanks Peter, as they say Rome wasn't built in a day.
Have a good weekend.

George

[pcprogrammer]

I tried my scope with a bench power supply and at 5V on the battery terminals it draws ~1A.  Yours has a different FPGA, but 350mA is quite a bit extra, so you might want to check the different voltages in the scope and feel if parts get extra warm.

[Gannet]

Hi Peter, I feel like I moved mountains. Hopefully my copy of my screenshot is attached.
Nothing inside the scope is getting unduly warm.
No how do i get that file onto my usb drive sdc/sdec1

Cheers

George

[Gannet]

Welcome to Linux Lite 6.2 linux
 
Sunday 20 November 2022, 13:47:29
Memory Usage: 1063/15913MB (6.68%)
Disk Usage: 483/7GB (7%)
Support - https://www.linuxliteos.com/forums/ (Right click, Open Link)
 
 linux  ~  lsblk
NAME        MAJ:MIN RM   SIZE RO TYPE MOUNTPOINTS
loop0         7:0    0   1.9G  1 loop /rofs
sda           8:0    0 931.5G  0 disk
└─sda1        8:1    0 931.5G  0 part
sdb           8:16   1     0B  0 disk
sdc           8:32   1   3.7G  0 disk
└─sdc1        8:33   1   3.7G  0 part /cdrom
nvme0n1     259:0    0 465.8G  0 disk
├─nvme0n1p1 259:1    0   100M  0 part
├─nvme0n1p2 259:2    0    16M  0 part
├─nvme0n1p3 259:3    0 465.1G  0 part
└─nvme0n1p4 259:4    0   593M  0 part
 linux  ~  

Hmmm, that's better :-)

[pcprogrammer]

By the looks of it /dev/sdc is your USB stick with linux on it.

/dev/nvme0n1 is your internal m2 disk which will hold your Windows. It is not mounted and best leave it that way. /dev/sda seems to be a 1TB disk, that also is not mounted. But there is some presumably ramdisk mounted at /rofs.

This will be where you can download the needed file from my repository. It most likely will do this automatically when you use firefox (I assume that to be the default browser that came with Linux Lite) There probably will be a download folder.

But the thing to find out is at which device the SD card will be connected. So insert the SD card into it and connect it to the computer after you booted into Linux Lite. It should mount it. Then use the "lsblk" command again and show me the resulting list.

[pcprogrammer]

Nothing inside the scope is getting unduly warm.

Well that is a good thing. Lets see if the scope will do the backup and if it will run my firmware.

Regards,
Peter

[Fungus]

Seems to me like it was actually very useful for the first two years.

Otherwise you wouldn't care that it broke. 

[Gannet]

I did like it.

George

[Fungus]

I did like it.

George

I'm just confused by the title of this thread...

[Gannet]

Well! as it is it's as much use as a chocolate fireguard or even a teabag without holes, pardon my weird sense of humor 

[Fungus]

Well! as it is it's as much use as a chocolate fireguard or even a teabag without holes, pardon my weird sense of humor 

The expression is used for things that are useless by design, not for things that broke after years of being useful. Hence the confusion.

Anything can break.

[Gannet]

I do appologise if I have offended you, just frustration on my part. I do expect devices such as these should not go down after such a short time. This BB with all the messages we see and for all the conjecture created does not in my opinion reflect that of a quality piece of equipment.

[Gannet]

Hi Peter, I have ordered a new 8Gb Usb memory card.
I'm assuming I can use that to load the files you are pointing to and that there is no need to install linux lite 6.2 onto the new card.
I will still need managing once I have the card, so I hope you don't mind if we continue our dialogue.

Best regards

George

P.S looks like you may get a Haggis for your tea if were successful

[pcprogrammer]

Hi George,

no problem. I check the posts regularly.

There is no need to install linux on the SD card. Just need to copy the backup software to make the backup of the flash and then the new firmware. The new card most likely will be formatted as needed, but if not I will guide you through the process.

Regards,
Peter

P.S. Rather have biscuits with my tea

[Fungus]

I do appologise if I have offended you

No offense, it just seemed very weird. 

[pcprogrammer]

I do appologise if I have offended you, just frustration on my part. I do expect devices such as these should not go down after such a short time. This BB with all the messages we see and for all the conjecture created does not in my opinion reflect that of a quality piece of equipment.

Don't let Fungus mess with you. He knows this scope is not the best on the market. A bit of a shame yours only lasted a couple of years before giving problems, but truth be told mine arrived with a touch panel problem. It was that what got the fun started

Now I have reverse engineered every part of it,the  schematic to some extend, the firmware and the FPGA programming.

[Fungus]

... does not in my opinion reflect that of a quality piece of equipment.

Apple used them to develop their M1 chip, you can see them all around the lab in this video:



If it's good enough for Apple...

[Gannet]

Got the 8GB card Peter and this is as far as I got.
Somewhere on the 8Gb card is the name 'Scope' that's the label I gave it when I formatted in windows.
Seems OK or isn't it?
Now gently gently where do I go from here

Thanks

George
PS Going for a beer  Screenshot.jpg (86.7 kB. 640x480 - viewed 141 times.)

[pcprogrammer]

Hi George,

the scope does not look at the label, so it will be fine as long as the format is fat32 and not ntfs, exfat, ext4 or any of the others out there. Unfortunately the compression on the screen shot is to much. Can't read what is shown on the screen.

So please try again.

Peter,

P.S. you might find me weird, but I don't drink alcoholic beverages, lemonades or coffee

[Gannet]

Hi Peter, here are a couple of higher res snapshots.
The 8Gb usb is formatted FAT32 with a disc label Scope
Goodness knows how I got the FR** file on there, maybe you'll have to go gently through the process.
Screen1 is just showing I can be in Linux and access www.eevblog. and hopefully the state of my USB card
Lest see where we go from here.
Thanks
George

[pcprogrammer]

Hi George,

I can see that you used Windows to format the card The tell tale sign is the "System Volume Information" folder.

The fnirsi_1013d_scope.html file is indeed the mystery of what you have done. It is a web page. Not needed for this quest.

Best is to first see if the card works in the scope. Just stick it in the SD card slot and power up the scope. If all is well it starts up normally. Do a picture or waveform save on the scope. When it works as normal, turn it off again and go back to the linux machine with the SD card. There should be a couple of new files on the card, made by the scope.

Then it is time to do the needed steps.

Check with "lsblk" if the SD card is still attached to /dev/sdb. (You probably used "sudo fdisk -l" to list the partitions in your screen shot?)

Then open the github repository that contains the binary needed for the backup (https://github.com/pecostm32/FNIRSI-1013D-1014D-Hack/blob/main/Test%20code/fnirsi_1013d_firmware_backup/Binaries/fnirsi_1013d_fwb.bin)

Download it and open the download folder to see if it is there. Then change your path in the terminal window to where the file is. (cd Downloads)

Now comes the part to load the code onto the SD card. But first it has to be unmounted.

Code: [Select]

sudo umount /dev/sdb1
Be aware that this is with the 1 at the end indicating the partition. And this is assuming it is at /dev/sdb. It can also be /dev/sdc, /dev/sdd, etc.

Then the file has to be copied onto the SD card, but to a dedicated sector.

Code: [Select]

sudo dd if=fnirsi_1013d_fwb.bin of=/dev/sdb bs=1024 seek=8
This is done to the device not the partition, so no 1!

And that is all there is to it. Insert the card into the scope and turn it on. You should get what I show in the video I posted before. At the end the needed files are on the SD card. You can use either the scope with the USB connection or your SD card /reader/writer to get them onto your computer. This part can also be done under Windows if you prefer.

Let me know how it goes.

Peter

[Gannet]

Hi Peter, I'm nowrunning Linux lite in terminal mode. a few more files were written to the card, I can see them now. (good init)
we have 1.bmp, 1.wav,piclist.sys.pic_system.sys,wavelist.sys and wave_system.sys

besides others we have one partition
sdb/sdb1 8:16, 1 , 7.4G 0 part/ media/linux/scope

Its the next part I'm strugling with, opening the github repository and how to download the file and save it to my 8Gb card whilst in linux terminal??

Or have I missed a step?

**************

OK it looks like I saved the file finirs_bin to home/linux/downloads/


The CD command will not work for me ? thats odd

[pcprogrammer]

With linux running from the USB stick it will not write the things you download onto the USB stick. It will just hold it in a ram disk. And that is not a problem, but means it has to be done in one go.

Ok the scope works with the SD card, since it wrote the files to it.

The partition you mention is on the SD card, and that can't be used to hold the file that needs to be written to sector 16. This because it has to be unmounted to use "dd". I write file, but it is not like a normal file within a file system. The binary data from the file is written with the dd command outside the file system.

You write it is downloaded to home/linux/downloads, and that sounds right. Within the terminal window you are most likely already in the home/linux folder and as linux is case sensitive the command to switch to the downloads folder is then just "cd downloads" To use the full path it would be "cd /home/linux/downloads"

So the step is to download it into the downloads folder, which firefox does automatically, as you already found out.

Then enter that folder in the terminal window. When that is done perform the two actions mentioned in the code sections in my previous post. The unmount first, and then the "dd" (disk duplicate) command.

But make sure that the SD card is still attached as /dev/sdb.

I can't stress that enough, because "dd" does not check, it just writes where you tell it to write, and using the wrong device can f-up your system.

[Gannet]

I did manage to change directories as you can see.
But I can't seem to unmount.
But it's wine time, I'll pick up tomorrow

Cheers George

[pcprogrammer]

Hi George,

copy paste is your friend in this case, but needs to be done with the right mouse button.

The mistake you made is that the command is "umount", so no "n" in it.

You can select what I typed just as in Windows, and then copy and paste it in the terminal window, but to do the actual pasting you have to use the mouse right click and then the menu item paste. You can't use "ctrl - v" in the terminal window.

Cheers,
Peter

Edit: And there must be a space between the command and the parameter.

[MrAl]

Hi,

So i guess this is the same as the 1014D just in a different package?

I thought i saw a web site selling chocolate that melts at 100 degrees C but then found out it really melts at 30 degrees C they lied (ha ha).

[pcprogrammer]

So i guess this is the same as the 1014D just in a different package?

Yes, but with a simple AWG, and the hardware user interface added to it. For the rest the same crappy front end, same FPGA, ADC's and MCU.

[Gannet]

Today's effort, I'm sorry I'm being very slow Peter, sometimes I get very very tired, I'm not sure if my oxygen machine is helping me at all I arose at 9 this morning and went straight back to sleep in my armchair untill 1pm.

However, I think I'm making progress, I'm beginning to understand that Linux is very 'case & space' sensitive.
On my linux desktop there is an icon showing my sd card which from there I can unmount it. When first loaded it is automatically mounted.
From the next screenshot you'll see that the file fnirsi is in my downloads folder.
Sadly, once more I'm having trouble with the last sudo command 'dd if=fnirsi_ etc ' wont work for me.
Am I in the correct directory or somewhere that I shouldn't be ?
Maybe you can spot it from the screenshot.

Regards

George

[pcprogrammer]

No problem with the slowness

The why of the command failing is because you got the wrong file in the download folder. The open folder shows fnirsi_1013d.bin and you need the fnirsi_1013d_fwb.bin for this part of the quest. (https://github.com/pecostm32/FNIRSI-1013D-1014D-Hack/blob/main/Test%20code/fnirsi_1013d_firmware_backup/Binaries/fnirsi_1013d_fwb.bin)

Also it might be that using the unmount option from the folder also drops the SD card attachment. It is best to use the command line instructions.

To see what is available in a directory you can use the command "ls"

For the rest things look to be correct.

Greetings,

Peter

[Gannet]

Do you think this is it Peter?

Gnight

George

[MrAl]

So i guess this is the same as the 1014D just in a different package?

Yes, but with a simple AWG, and the hardware user interface added to it. For the rest the same crappy front end, same FPGA, ADC's and MCU.

Well thank you for the info this has been very interesting.

Now do i read it right when i see that the ADC's being used in these scopes are only rated for 200MS/s and not 1GS/s as advertised?
200 megasamples per second VS advertised 1 gigasamples per second that is.

[pcprogrammer]

Do you think this is it Peter?

Hi George,

yes that is the right file, but in the command you types you forgot the "dd". You used "sudo if=......" instead of "sudo dd if=......"

Code: [Select]

sudo dd if=fnirsi_1013d_fwb.bin of=/dev/sdb bs=1024 seek=8
Try the copy and paste method I wrote about before

Good day,
Peter

[pcprogrammer]

Now do i read it right when i see that the ADC's being used in these scopes are only rated for 200MS/s and not 1GS/s as advertised?
200 megasamples per second VS advertised 1 gigasamples per second that is.

A single ADC is only rated for 100MSa/s, but with two ADC's in a single package you get 200MSa/s by using different clock phases for each ADC. (180 degree phase shift)

All the information about the scope and the reverse engineering process can be found in this thread https://www.eevblog.com/forum/testgear/fnirsi-1013d-100mhz-tablet-oscilloscope/

[Gannet]

Well Peter, you're not gonna believe this! it's up and running. My firmware must have been trashed after all. When I switched on there was lots of relay clicking. I put my probe in and hey presto it seems fine. I do notice that there is a slight screen shimmer.
Also i see under the battery icon it says v0.005 is that revision or volts?

Do we now have the latest firmware fix installed ?

Couple of pics to show are attached. I don't know what all the bother was about (he he) I must hold the record for the slowest installation of your firmware.
Thanks, a big bunch Peter, you'll get your thanks in heaven.

Best regards
George

[pcprogrammer]

Hi George,

yep, that is my firmware, and this also shows that your display is different. The colors are supposed to be yellow and magenta cyan for the traces and green for the trigger level indicator. The background is supposed to be black.

This probably means that the firmware in the flash is not different then the one in mine.

It might be that the display connection is poor, or has it always been like this?

The v0.005 is the revision indicator, and that is indeed the latest version that I made. It does lack the FFT that the original has, and there is no roll mode with the large time per division settings.

Does the fact that you now have my version of the firmware running, mean you skipped the step of making the backup of the flash? In other words that you did not load the fnirsi_1023d_fwb.bin file?

Hope you enjoy having the scope running again.

Cheers,
Peter

Edit: had the wrong color in mind.

[pcprogrammer]

Here is a picture of what the startup screen looks like on mine. It is a bit blurry because I broke the glass when I changed the touch panel and replaced it with an acrylic sheet. The protection foil is still on it.

[Gannet]

I think you're correct Peter, my colors are a wee bit washed.
I missed out the 1023d.bin thinking if my firmware in the flash was scuppered then there was no need to worry and just go ahead and load your firmware
Unless of course I can still do it and that you may want a copy?

I'll check my display connections just in case, but for now it's working thanks to you and your tenacity in sticking with me.
Have a great weekend.

Kind regards

George

PS. Mr moderator, Peter deserves a goody sticker on this board.

[pcprogrammer]

Hi George,

since it is most likely the same old firmware, now that it seems your display is a bit of, I don't need a copy of it.

The display shows the inverse of the intended colors and I'm not sure if this is an option that can be selected. Have to dig up the datasheet for it.

But at least your scope seems to be working again. Just report back when new problems arise.

Regards,
Peter

Edit: did not see a mode where a display would do this, so it is a bit weird. But for example the yellow is made of 255 red, 255 green and 0 blue, and it shows as blue on yours, so 0 red, 0 green and 255 blue.

[Gannet]

Thanks Peter if you hear of any more on the Color’s, please let me know,

George

[pcprogrammer]

Hi George,

about the colors, when you bought the scope was it already like this or did it show the yellow and cyan traces on a dark background?

If so, this means some part of the system broke down causing this, but what broke down, that is of course the question.

Cheers,
Peter

[Gannet]

Good morning, Peter, I don't think the colors I have are what they used to be!
Additionally, although it appears to be working, the scope will often 'Switch off' and sometimes it will take one or two cycles of the on/off button for it to re-awake.
Hmm, maybe I do have an onboard fault after all.

Regards

George

[pcprogrammer]

Hi George,

with the problems recurring it indeed points to some defect. I have searched on goolge for inverse colors on LCD, but it did not give much to go on.

I have no idea if a power supply issue can cause a display to do what yours shows. Can you measure the voltages with a multi meter? Especially the -6.3V, 15.8V and the 10.3V.

It will be a process of exclusion to determine the root of the problem.

[Gannet]

Hi Peter, here's what I get with the voltage check.

16.12v
10.62v
-7.16v

I attached another screen shot, Did you notice that the upper part of my grid/graticule is missing

Cheers
George

[pcprogrammer]

Hi George,

these voltages are within range of what is maximally allowed going by the datasheet I found for a similar panel.

Did not notice that the grid was missing in the upper half. Start to think that the display is the problem here. I wonder if you see a large current drop when you boot the scope without the display connected.

I bought this display and tested it with my scope. https://nl.aliexpress.com/item/32693491775.html I expect yours to be compatible with it. Maybe worth a try.

Don't know how the older model is constructed, but replacing it in the newer model is a bit tricky since they used double sided tape to attach it to the touch panel.
Someone used special wire to cut it loose in his scope. It is mentioned somewhere in the FNIRSI-1013D thread.

Regards,
Peter

[Gannet]

I’m willing to give the display a try Peter, so I’ll order one up, what’s bothering me that the scope is sometimes very difficult to ‘turn on’ and often turns off at random.  So there’s a bigger issue somewhere?

Praps I was a bit quick on celebrating.

Cheers George

[pcprogrammer]

Hi George,

the turning of could be a thermal protection of the main regulator, since it is drawing ~30% more current. The difficulty in turning it on might be related and depends on when it happens.

As you can see in the schematic, the power switch only controls an enable to the SPX29302T5, which makes the latter the actual power on part. It could have problems turning on when there is too much initial current being drawn. But this is not my field of expertise.

Did you check the other voltages in the scope, like the 1.2V, 2.5V, 3V and of course the 3.3V? Another one is VREF which should be ~1.25V.

Regards,
Peter

[Gannet]

Where are the schematics Peter?

I didn’t check those other voltages yet

Cheers

George

[Gannet]

Sorry Peter, belay the last. I have them so I'll check all voltages and get back to you.

Cheers

George

[pcprogrammer]

Where are the schematics Peter?

I didn’t check those other voltages yet

Hi George,

had other things to do today, but going by your next post you found them. (The link was in my first post in this thread )

Cheers,
Peter

[Gannet]

That's fine Peter, I found loads of other info too. You say check the voltages, my goodness I can't even see the pins.  A breakout board of some kind would indeed help!
I looked at the file containing all the parts on AliExpress, some interesting stuff there I see. any particular item you'd recommend?

I have noticed that when the scope shuts down, well it doesn't, all that seems to be happening is that the display goes blank.

Is there a silk screen overlay anywhere with component Id's and locations?

Thanks

George

[pcprogrammer]

Hi George,

  I have the same problem. Need a microscope to work on these things.

About the goodies on Aliexpress, it depends on your hobby. I play around with microcontrollers and well scopes, and for that the things like a CH341 programmer or a ST-Link are very handy. But availability is troublesome at the moment causing the prices to be way to high.

I will mark the measuring points for you on one of the photos I have, because there is no overlay with component numbers.

About the shutdown the 3V3 should go down to 0, but there is a small delay to allow the MCU to write settings to flash (original) or the SD card (mine)

Regards,

Peter

Edit: Attached the power supply part with voltages mentioned. VREF is not 1.25V but it is for measuring the battery level with the MCU, so it varies and is not to important. The 3V is missing on this picture, so another one is coming.

You might have to download the second image and zoom in to read the text.

[Gannet]

Many thanks for the Voltage check points.

All voltages are within +0 and -160mv Peter, and with the display blank, you guessed it, all voltages are just the same, that's pointing to a bad display I guess, what do you think.

I have a display on order with a message saying 'dispatched'

Cheers

George

[pcprogrammer]

You are welcome. Happy to help.

Did you check the current when the display is disconnected?

I'm no expert on displays and have no idea if what yours shows is some typical failure behavior. Guess it is wait until the new one arrives to see if it is the actual culprit. 

Cheers,
Peter

[Gannet]

Hi Peter, I return with not so good results,
Trying to make sense of the display's that I'm getting and most appear gibberish.
I fear that there is a great deal more wrong with the scope than the display, in fact it may not be the display that is at fault after all. The washed-out colors etc could well be due to something other.
I have attached a pic of the waveform from the test point on the scope.

Maybe I should just bung it in the post and give you something to play with whilst you're not enjoying yourself. 

What do you think?

Cheers

George

[pcprogrammer]

Hi George,

which signal are you measuring? Not sure if probing the scope with itself works properly. There is the shifted ground issue to consider.

It would be an interesting object to play with

Did you order another display to test it with?

Cheers,
Peter

[Gannet]

Hi Peter, still waiting on the new display,
I'm probing the calibration input on the top of the scope; I assumed it was a square wave 1V test point.
However, I'm using a small square wave generator, have a look see at the results.

Cheers George

[pcprogrammer]

Hi George,

in both cases it seems to measure some 50Hz signal. Is that the mains frequency in your neck of the woods?

The calibration signal of the scope should be a 1KHz square, so if that is what you are connected to in the first picture you showed, the period time is 1ms, and your scope is set to 5ms/div and 20KSa/s, so it should show 5 periods per division, so yes something is wrong.

In your second picture it looks like it is picking up just the mains on an open probe.

Weird things happening indeed.

I just noticed that the pictures saved on the scope won't open on my Ubuntu machine. So have to do some digging of my own on why that is.

Regards,
Peter

[Gannet]

Hi Peter, yes I believe it’s pretty goosed, do yo fancy having a go of for me ?

I think it’s beyond my level, I’ll bet you a Haggis that the display is not the issue and if it’s not I’ll let you have it when it arrives

Yes we are on 50 hz in the Uk
Cheers

George.

[pcprogrammer]

Hi George,

You might be right that there are several issues with it than just the display. After you try the display when it arrives, and it does not solve the problem, I'm willing to give it a go. Having a working one and several other scopes besides it, at least gives the option of probing them to compare the signals and hopefully trace the problem.

I don't bet because I never win

Regards,
Peter

[Gannet]

Good idea Peter, I’ll let you know when the display arrives,

Keep safe

George

[BillyO]

I would argue that the chocolate fireguard is of far more use than the Fnirsi.

Let's say that fireguard was the firewall in a small airplane and you crash in the northwestern wilderness of Canada.  That chocolate firewall will serve to distract the grizzly bears while you make good your escape.  The most you could do with the Fnirsi is try to bat the bear about the head with it.  That would just annoy it.  Also, before the bears show up you could stuff some of that chocolate into a backpack for quick and tasty energy for your trek out of the woods.  The Fnirsi would just be deadweight and not the least bit nourishing and it probably tastes as badly as it works.

[Gannet]

[Gannet]

Hello Peter, the new display arrived and made no difference to my washed out colors.
I thought we were onto a winner but sadly no.
So if you are willing to have a go at it for me I'd be eternally grateful. You can have the new display as a Christmas gift. :-)

Let me know what you think.

Regards
George

[pcprogrammer]

Hi George,

that is disappointing. That makes me wonder what else can cause this. A side by side comparison might provide the answer.

I will send you a personal message to about sending it over.

Regards,
Peter

[perdrix]

Less use than a chocolate fireguard (you can after all eat the chocolate fireguard)!

D.

[Gannet]

That's a good Idea, that way no one will be able to trace where it's gone. Trace he he 

[pcprogrammer]

Received the patient and started investigations on it.

Apart from the inverted display there is also a, what looks like, random resetting of the processor. Some relays click, the screen goes black and the current drops. The screen going black is due to the brightness signal being turned off, but there are also moments that it stays on and the screen is just white.

Think this has to do with how far the processor can run through the code before running into trouble. The strange bit is that sometimes it starts up fine and works stable for a while, but eventually it turns of again.

The current is around ~350mA when the back light is off and no activity on the LCD signals, around ~750mA when the back light is on, but still no activity on the LCD signals. It goes up to ~1A when it is working.

The MCU clock shows normal signs of live, and the voltages are up to par without garbage on them. My guess is that the MCU is broken somehow. Ordered up a new one and when it arrives, I will see if it is indeed the culprit. It is a bit of a challenge with the ground pad, but some youtube videos show it can be done with just a hot air station.

[pcprogrammer]

Unfortunately I messed up. QFN is not my thing. The old one came of easily, but I did not notice that a pad also came of

Tried to solder the new one on and the solder did melt and the ground pad made proper connection, which can be measured via small copper bits on the corners of the chip, but I think a lot of the pads did not. My hands are also not so steady anymore so while trying to tap it down it moved of the pads 

Finally managed to get it straight and under a microscope it looked like the pads had connections after multiple attempts with a soldering iron running past them, but it is dead. So either I fried the chip or there are still some bad connections. I can see on a scope that the clock is running, and on startup there is some activity on the SD card signals.

So I have met my match. 

The picture shows the pad that came loose.

[Chris02]

Hey pcprogrammer,

cool project to make a new firmware for the scope!

If you want post or send me some high resolution pics and I can try to check the soldering.
And did you check the current consumpting after changing the cpu? It maybe should give you some hints, e.g. if there is any short shircuit from soldering.

I also has got one question. Supports you firmware (I guess original one doesn't do) simple "+" operation? I have to check some RS485 signals and need to add CH1 and CH2.

[MrAl]

Hi,

One of the things i think is strange about some of the new low cost scopes (like this "one", and i have to put the word "one" in quotes ha ha) is the wave generator options.
Who would design a wave generator where you can not vary the output amplitude?  Not me.
If you are going to go though the trouble of putting a wave gen inside another instrument that has lots of buttons and knobs already why not build in an amplitude level adjustment?  How much more design work would that require, about an hour or two?

[pcprogrammer]

If you want post or send me some high resolution pics and I can try to check the soldering.
And did you check the current consumpting after changing the cpu? It maybe should give you some hints, e.g. if there is any short shircuit from soldering.

I have checked the soldering under my camera microscope and for as far as I can see the connections are there, but it might well be that underneath the chip there are some shorts. Hard to tell.

The current drawn is lower then normal, but higher then without the MCU. I suspect the MCU is fried due to the multiple tries to get it right.

I also has got one question. Supports you firmware (I guess original one doesn't do) simple "+" operation? I have to check some RS485 signals and need to add CH1 and CH2.

No, the new firmware even lacks the FFT and has no new features except the selection of the sample rate.

[pcprogrammer]

Hi,

One of the things i think is strange about some of the new low cost scopes (like this "one", and i have to put the word "one" in quotes ha ha) is the wave generator options.
Who would design a wave generator where you can not vary the output amplitude?  Not me.
If you are going to go though the trouble of putting a wave gen inside another instrument that has lots of buttons and knobs already why not build in an amplitude level adjustment?  How much more design work would that require, about an hour or two?

You are hinting about the 1014D, because the 1013D does not have a wave generator. But implementing proper amplitude variation requires additional hardware with a second DAC. With only 8 bits there is not much room to keep a proper signal at lower amplitude. It is how the Hantek DSO2D1x does it though.

They could have done it in the 1014D, and maybe donwulff can implement it when he converts the new 1013D firmware to the 1014D 

[daemon17]

Hi pcprogrammer! 
I bought this device on 01.03.2023
I want to reflash it and I have a few questions.

How can I make a FLASH backup without breaking the scope?
Do I need to use sunxi-tools for this?

Explain what is the order of loading the scope? It first refers to the SD-card or FLASH Chip.
I work with Linux, so there are no problems with it.

You've done a great job. I admire you!

[pcprogrammer]

Hi daemon17,

welcome to the forum.

First of all to use the new firmware I wrote you don't have to write to the FLASH. It works from the SD card.

But making a backup of the FLASH never hurts, and can be done with the backup tool I wrote for it.

The procedure for making use of both these is the same.

The binary for the backup is here: https://github.com/pecostm32/FNIRSI-1013D-1014D-Hack/blob/main/Test%20code/fnirsi_1013d_firmware_backup/Binaries/fnirsi_1013d_fwb.bin

The binary for the new firmware is here: https://github.com/pecostm32/FNIRSI_1013D_Firmware/blob/main/fnirsi_1013d_scope/dist/Debug/GNU_ARM-Linux/fnirsi_1013d.bin

Information on how to load the binaries to the SD card are given in the readme here: https://github.com/pecostm32/FNIRSI_1013D_Firmware

I also would suggest to check the thread about the hacking and development of the firmware here: https://www.eevblog.com/forum/testgear/fnirsi-1013d-100mhz-tablet-oscilloscope/msg3807689/#msg3807689

Success and have fun with it.

Cheers,
Peter

[daemon17]

Thanks Peter for the help.

I'd like to share exciting news.   
I got everything

Best regards, Alex
             

[daemon17]

Hi Peter!
How can I restore the original firmware from the saved files:

• FNIRSI_1013D_full_flash_backup.bin,
• FNIRSI_1013D_tp_config.bin,
• FWB_FSI_1013.bin

Would love to know about it

[pcprogrammer]

The flash can be restored if needed by using the full flash backup and a flash programmer like a CH341 programmer, because that file (FNIRSI_1013D_full_flash_backup.bin) is an exact copy of the flash chip on the board. (Either 2MB or 4MB)

The FWB_FSI_1013.bin is a copy of just the firmware part of the flash, and has to be renamed (don't recall to what) and written to the SD card as a normal file. Rebooting the scope would then write this firmware to the flash chip. This only works in the original firmware! Also not very useful in your case since it is the same version as already there.

The file FNIRSI_1013D_tp_config.bin is not really important. It is just a copy of the configuration in your scopes touch panel. I added that to the backup because there are different touch panels in circulation and I needed to get my hands on one that I messed up in one of my scopes 

To revert back to the original firmware you can delete the new firmware form the SD card, or stick another blank formatted SD card in the scope.