FNIRSI-1013D "100MHz" tablet oscilloscope

[Sleo]

Michal Derkacz (github user ziutek) made an update for the firmware. He improved the RMS measurement and created a new version v0.006 for it. I merged it into the repository after looking at the code. Did not test it myself and took it on face value.

There is no new image file made like for version v0.005, so windows users who fail to use the binary in the repository have to hope on someone posting a new image file here on the forum.

I'm too lazy to make an image for each new firmware version (as well as for different SD cards), so I made this little utility for Windows users.

This utility can load/verify/delete binary firmware directly to the USB-connected oscilloscope or to an SD card in a card reader. It can also read/write/clear/compile/fix LCD/TP configuration sector and load/save it as a file.

The utility writes directly to the physical sectors of the selected device, so there is a potential risk of data corruption. As usual, use it at your own risk. But I tried to make it safe enough - it won't work with devices that aren't USB, recognized by the system as hard drives, have a non-512 byte sector size, containing a system volume, etc (and won't even put that devices on the list). Also it can read but will not write to physical sectors that falls within MBR and existing partitions of any type (except of RAW).

The "F1C100S" device (i.e. Allwinner CPU of Fnirsi 1013D/1014D oscilloscope) will be the selected device by default if it is present, then the software will look for SD/SDHC cards and then for other USB storage devices. Side effect: USB flash drives will also be on the list because I don't know of a reliable way to distinguish them from noname USB card readers.

The utility requires administrator rights to directly access the physical device. If you are using an administrative account, elevation will be performed automatically (you may get a UAC warning), for a non-administrative account, you will be prompted for administrator credentials.

This software does not have any undocumented, trojan or destructive features, and does not use any third party libraries and components. It does not write anything to the file system or registry. It's made with Delphi VCL and pure Win32API. Software compatible with systems from Windows XP SP3 to Windows 11 (both X86 and X64). Any bug reports and suggestions are welcome.

How to use: For convenience, place this .exe in the same folder as the firmware binaries (and configuration binaries, if needed) and run. Do not worry about safely removing the device, the program writes data immediately, ignoring system caching. I added everything that is needed at the moment to the zip - two firmware binaries (v0.005 and v0.006) and configuration sectors for 1013D/1014D (without swapping TP, do it yourself, it's easy).

PS: Sorry for not implementing automatic light/dark theme switching, ;-) maybe in the next version...
PPS: For curiosity and possible future use: disabled input fields can be unlocked by double-clicking on the label above it.


By the way. With Windows and original firmware, the oscilloscope mounts as a disk device one second after selecting "USB Connect". With PCProrgammer's firmware (v0.005 and v0.006 tested) this takes some time, maybe a minute. Just wait, then reselect device, if not in the list.

[cool2000]

It seems that RMS measurement implemented in the version v0.006 requires some changes for the saved WAVE files. Currently it doesn't show correct value.
Possible solutions could be: save calculated RMS value with the other WAVE file settings or recalculate RMS value from the file samples after loading.

[Evi]

@Sleo
What's wrong?
Launched with admin permissions
Tested on two PC and two card-readers

[Sleo]

@Sleo
What's wrong?
Launched with admin permissions
Tested on two PC and two card-readers

Sorry, it's not widely tested yet.
Either the specified sectors belong to the mounted partition, or the user does not have write access to the physical device. The third version is of course my mistake somewhere in the code.
i will have to add diagnostic log output for such cases.

Your card may be formatted with not enough space between MBR and first partition. Try, please, press "Clear" button (it rewrites only one sector), does it get the same error?

[Evi]

Your card may be formatted with not enough space between MBR and first partition. Try, please, press "Clear" button

Maybe you reccomend the right formating?
Clear did nothing.

[Sleo]

Your card may be formatted with not enough space between MBR and first partition. Try, please, press "Clear" button

Maybe you reccomend the right formating?
Clear did nothing.

"Clear" does not mean formatting the card, erasing the partition, MBR, etc. It clears (deactivates) the previously written firmware, returning the oscilloscope to its original firmware. More precisely, it writes zero to the first byte of the first sector of the firmware, which does not allow the bootloader to recognize the presence of firmware on the SD card.
If this operation not raises an error, then you have write access to the physical device and the first sector has been successfully overwritten. So it seems to me that the assumption of insufficient "hidden" space is correct. I will write recommendations for such a case a little later.

[Sleo]

Your card may be formatted with not enough space between MBR and first partition. Try, please, press "Clear" button

Maybe you reccomend the right formating?
Clear did nothing.

Well, I reproduced this situation.

You can check the offset of the first partition using the command line:
DISKPART
LISt DISk
SELect DISk * (use appropriate disk number here)
SELect PARtition 1
DETail PARtition ("Offset in Bytes" is the value you are interested in)
EXIt

Offset 1048576 bytes is the default for 8 GB and larger devices, partitioned using Windows. This offset leaves about 1 MB of free space for oscilloscope firmware.
But if Windows formats an already partitioned device with less free space, it DOES NOT CHANGE the existing offset. You need to delete the partition and create it again.

Command line way:
DISKPART
LISt DISk
SELect DISk * (use appropriate disk number here)
CLEar
CREate PARtition PRImary
EXIt

Disk Manager way:
Find your card, right click on partition, select "Remove Volume", confirm.
One more right click, select "New Single Volume", next, next, next... You may leave everything as default.

After creating the partition, you need to format it as FAT32. Then use my loader, it should work without errors.
Please note that the default offset for 1..4 GB devices is 65536, so there will not be enough space for oscilloscope firmware without special repartitioning.

I think I should at least add a check for "hidden" free space on the SD, what do you think?

[Evi]

Now that works!

[juan3211]

Hi guys, I am newbie with oscilloscopes.
I have an 1013D between Aliexpres and my home.
I have a doubt about this.

I have read that with the 10x probe, this oscilloscope is capable of measure up to 400v

In oscilloscopes' specs, are these values peak or peak to peak (I mean, from -400 to 400)?
I have found different answers

This is important. If 400v specification is from -400 to 400, then I could use 10x probe to check pure sine inverter wave form

If 400v is peak , then I can't

Another question is: imagine that the previous answer is that I can measure only 400v. Is it better a 100x probe or a 400/40 little transformer ? Does transforme change/disturb/smooth the output sine wave of an inverter ? Even with load connected to inverter ?

Could you guys please ask me first answer? And if I need a 100x probe, for my use may be better/cheaper a little transformer ?

Thanks a lot.

[wolferl1210]

By the way. With Windows and original firmware, the oscilloscope mounts as a disk device one second after selecting "USB Connect". With PCProrgammer's firmware (v0.005 and v0.006 tested) this takes some time, maybe a minute. Just wait, then reselect device, if not in the list.

Thanks for the V0.006 release.

Screen 'USB connection'
Finger tip on the ' ON / OFF ' button and the screen returns to the scope view.
On my Windows PC (Win10 x64) the USB connect never shows up as a device

Checked with the device manager in Win10. There is a flick of the screen by tapping ON/OFF, could not find out if and what has changed.

[pcprogrammer]

Checked with the device manager in Win10. There is a flick of the screen by tapping ON/OFF, could not find out if and what has changed.

On the scope side nothing changed for as far as I checked. I don't use windows for development, so can't help there.

[pcprogrammer]

....

Best to start a specific topic with the question about measuring high voltages. I'm no expert on this subject, but would not trust this scope to be save with a 10x probe measuring mains voltage.

But if you do make sure there is no USB connection to anything. Power adapter or computer.

[juan3211]

Thanks a lot. I will open a new post.

Anyway, does anybody already do it ? 10x probe to measure a pure sine wave of 230VAC ?

Thanks.

Best to start a specific topic with the question about measuring hi

Best to start a specific topic with the question about measuring high voltages. I'm no expert on this subject, but would not trust this scope to be save with a 10x probe measuring mains voltage.

But if you do make sure there is no USB connection to anything. Power adapter or computer.

[Sleo]

Anyway, does anybody already do it ? 10x probe to measure a pure sine wave of 230VAC ?

Of course somebody did 
Just do not connect the scope to the computer with the USB cable during this measurement. The probe itself is limited to 600V peak-to-peak (at 10x) and the scope input is limited to about 40V (400v with 10x probe), so it's suitable for 230V power line measurement.
230V is an RMS value (measured at the line terminal with respect to the neutral terminal or vice versa). The peak value is 1.41 * Vrms = 325V. Both values are displayed relative to zero, which is the ground of the scope. On the one half of the period, Vpeak is +325V, on the other half -325V, so the total sine waveform sweep is 650V. But the maximum peak-to-peak voltage is only 325V with polarity reversal each half of a period.
This oscilloscope has a maximum v/div = 50v (with 10x att) and 8 vertical divisions, so the maximum screen space is 400V. The 650V wave will be off screen with 10x attenuation. To see the full wave, you need a probe with 100x attenuation. You can move the wave up or down to see the bottom or top half, but there is a bug in the alternative firmware with displaying vertically shifted waveforms.

Pictures 1 and 2 - alternative firmware v0.006, pictures 3 and 4 - original Fnirsi firmware.

[juan3211]

Thanks a lot @Sleo.

Of course I will use without conneting any PC or usb cable.

First, thanks a lot, I had a doubt about the 40V limit. I didn't know if it was -20 to 20V or -40 to 40V.  As I have read you, the answer is -20 to 20V is the real limit.

Why have you not broken it with a 10x probe? You are really measuring -32.5 to 32.5V aren't you? Or does the 40V value of the specs a RMS value? (so what a "incorrect" or "misunderstable" specs)

Second, I understand what you say about I won't see full wave.

Do you know if I use a step down transformer to reduce 230VAC to .... 24VAC por example ... will it disturb/smooth/.... the wave? what about FFT (armonics) ? are they affected?

I want to measure the "quality" of different waves from inverters (pure sine, modified, comparation between several pure sine waves or between modified ones)....

Thanks a lot.

Of course somebody did 
Just do not connect the scope to the computer with the USB cable during this measurement. The probe itself is limited to 600V peak-to-peak (at 10x) and the scope input is limited to about 40V (400v with 10x probe), so it's suitable for 230V power line measurement.
230V is an RMS value (measured at the line terminal with respect to the neutral terminal or vice versa). The peak value is 1.41 * Vrms = 325V. Both values are displayed relative to zero, which is the ground of the scope. On the one half of the period, Vpeak is +325V, on the other half -325V, so the total sine waveform sweep is 650V. But the maximum peak-to-peak voltage is only 325V with polarity reversal each half of a period.
This oscilloscope has a maximum v/div = 50v (with 10x att) and 8 vertical divisions, so the maximum screen space is 400V. The 650V wave will be off screen with 10x attenuation. To see the full wave, you need a probe with 100x attenuation. You can move the wave up or down to see the bottom or top half, but there is a bug in the alternative firmware with displaying vertically shifted waveforms.

Pictures 1 and 2 - alternative firmware v0.006, pictures 3 and 4 - original Fnirsi firmware.

[Sleo]

I had a doubt about the 40V limit. I didn't know if it was -20 to 20V or -40 to 40V.  As I have read you, the answer is -20 to 20V is the real limit.

No, No. The answer is 40V difference, regardless of polarity. You can measure -40V relative to ground, or +40V relative to ground. But not +40V relative to -40V. This is not the case for an AC power line, where +V and -V present on L-wire relative to N-wire at different times. Maximum peak voltage between L and N for 230VAC line is 325V (absolute value). But the sine waveform is 650V in height due to polarity change during period.

Why have you not broken it with a 10x probe? You are really measuring -32.5 to 32.5V aren't you?

40V limit was found in old version of scope manual. It's not a hardware limit, but just a scope screen height at max v/div. The hardware limit is very high, at least some hundred volts even with 1x probe. The input resistance is at least 1.2MOhm, allowable input current of operational amplifier protection diodes is 10 mA, so it needs about 12 KV on the input terminal to break the input amplifier. The real limit depends on the break voltage of passive components - resistors, capacitors, relays, PCB, probe, etc.

Do you know if I use a step down transformer to reduce 230VAC to .... 24VAC por example ... will it disturb/smooth/.... the wave? what about FFT (armonics) ? are they affected?

Why not to use just a resistive divider? Two resistors 200-500 KOhm @ 0.5W should be a good choice. Transformer has an inductance and magnetization effects, that may affect your waveform.

[1audio]

HV should be a separate subject. However 1) get an X100 probe rated for 1 KV. They are not expensive if you don't need 200 MHz (you don't). 2) a transformer is a very good and safe way to monitor the power line. You get full isolation and some HF and LF filtering of stuff on the power line that is basically noise. A commercial product for that was made but expensive and 120V circuits only.

[cool2000]

1) get an X100 probe rated for 1 KV. They are not expensive if you don't need 200 MHz (you don't).
2) a transformer is a very good and safe way to monitor the power line.

I have one x100 from AliExpress claimed rated for 1KV. After some time it simply stop working, so I have to disassemble it. There was some problem with the central wire. It doesn't look reliable and safe enough. Resisters used in voltage divider definitely not rated for 1KV and isolation itself is very poor.
So I think the transformer should be better option.

[Mortys]

Hello, can you help me when doing the sleo method, I get the screen displaced to the left

[Sleo]

You definitely need to write something to the configuration sector. I don't know which configuration is right for your oscilloscope, but you can try all known ones. The only problem is, if the LCD settings are wrong, it may not be so easy to press the "connect" button on the touchscreen. Nevertheless, you can use a card reader to write a different configuration or clear the configuration sector.

[Atlan]

Is the oscillation up and down caused by interference in the power supply?  Even the battery charge level fluctuates, wouldn't it help to filter the power supply better? Thank you

https://youtube.com/shorts/-fPYNPpwxRg?feature=share

Is there a bug list of the original and this alternative firmware?

[Atlan]

The power supply solution is a disaster.  From the distribution of power paths to the super attack of the connection of analog and digital power on the AD converter.  The use of power supply IOs that have guaranteed parameters at an input voltage of 4.3V and higher is laughable when it is powered from a 4.2V lipo.  It would be appropriate to provide a step-up converter to 5V for the oscilloscope circuits (except the backlight).  Otherwise, on my new version, u17 and U13 are powered by 3v3 (not 2.5v as shown in the schematic)

[Atlan]

The increase in the level of interference when switching from 100 Msa to 200 Msa/s is caused by what?
https://youtu.be/QJjuiEPd3dA

[pcprogrammer]

Bad hardware design.

The sampling done by the FPGA is done at the set rate divided by two. So on 200MSa/s the ADC clock is 100MHz. The two ADC's per channel are sampled interleaved leading to the 200MSa/s. The higher clock rate will lead to more power usage and can bring more noise on the signals.

That is what you get with a cheap scope.

In the original software they added a lot of filtering on the waveform captures to make it look like a nice and pure signal. In the open source software there is no filtering. The captured data is displayed as is, and shows the imperfections of the system.

[Atlan]

 Please Fix this mistake

and fix this ?
Similarly, the cursor moves when selecting the trigger (how does it work, the cursor moves to the selected place depending on ch1 or ch2, but if I select a menu and change my mind or click on ch1 while ch1 was previously selected, it causes the trigger cursor to move.

and The menu item (automatic confirmation) does not work. automatic confirmation is on, it is necessary to constantly tap the screen after calibration for confirmation.

At what voltage does the oscilloscope actually turn off, is the function of automatic shutdown at low voltage implemented in the software?

Vawefrom and picture error

Well thank you