FNIRSI-1013D "100MHz" tablet oscilloscope

[Atlan]

Possible bad calibration, bad touch on ch1 during calibration.

v0.020p checking the limits of calibration values. If the calibration is bad, it will say calibration failed

[Evi]

I got every time Calibration fail until I switched to stock FW and made there Baseline Calirbation.
Returnig back to new FW Calibration ended successfuly.

[Evi]

I guess it'll be comfortably to have ability to make partial calibration:
e.g. only baseline or to stop on 3, 6 or 7.5V.
You do not allways have all available voltages on site.

[Atlan]

Which is the stock firmware?  The calibration of the input dividers needs to be performed only once.

[Evi]

I made a fast switching between SW(described some pages ago). What is version I do not remeber it is one of the Fnirsi's releases.
I do not know does it needed be done once or twiсe, but the fact is fact.

[Atlan]

Evi upload firmware 0.021, go to the main menu. Push the usb connection. On the blue screen, tap on the PC screen. Take a photo of the displayed values and post it on the forum.

[Evi]

Can I ask about purpose?

[boojum]

Quite strange that the header is not found to be correct, because it is based on the guide lines given here.

Maybe this will help fix the problem:
I look Pillow code. It check header size and raise error because header size is 56 bytes.

After add value 56 to check header size condition Pillow successfully opens a 768070 bytes bmp.
Maybe you can try just increasing the BMP header to 64 bytes?
As I understand it, 56 is a non-standard header size, which causes a problem in many cases.

I found easy working solution for fix BMP format
Just change only one byte - 56 to 40 here (i use binary hack for .bmp, change one byte at 0x0e offset )
After that Firefox, Telegram and Python successufllly recognized screenshots.

[Atlan]

won't I be missing some data in the header?

v0.021a  here I changed the //Size of DIB header
   40, 0, 0, 0, //56,0,0,0

Can I ask about purpose?

I want to know what the tolerance is on the input part that did not pass the calibration.

I hope that you did the first part of the calibration without USB and with the probe disconnected, and then set the voltage on the test and confirm OK

[pcprogrammer]

won't I be missing some data in the header?

Who knows.

Nice those semi standards. According to the wikipedia page about bitmap file format it is a specific header format to use when additional information needs to be specified. In this case for the correct color mapping.

I tried the modification of the single byte on a file I have on my system and Firefox failed on it before, but opened it just fine after the change. The default image viewer on Linux Mint has no problem with either.

[boojum]

won't I be missing some data in the header?
v0.021a

Position of pixel array still correct - it not changed.
The last 16 bytes of the DIB header just won't be used, I think.

[pcprogrammer]

The last 16 bytes of the DIB header just won't be used, I think.

And yet these bytes tell the layout of the color information within the 16 bits per pixel. That is why I choose this header format.

Ah well, if it works it works. 

[bffargo]

fnirsi_1013d.bin v0.020j.bin
Something wrong with CH1 below 500mV after calibration

I had similar results yesterday pulling 20J.; though mine was that it didn't work correctly in higher V/div settings only. I recalibrated 3 times (which is a huge pain as I had to hack some interesting BNC splitters and bench DMM probes to detect exact voltages at the inputs and tweak with analog power supply dials). Manually fiddling with volt range on each channels got signals showing correctly.

But auto set is completely broken it doesn't detect anything right no matter channels being on or voltages or waves applied.

[Atlan]

Fargo, an image with data as I wrote above. (auto settings not work in DC mode?)

I will take one BNC T-piece, insert it into ch1, connect it to ch2, and from the BNC T-piece I will stretch the cable with alligator clips to the measuring device and the power source.

Tokar: I know about your request for DC mode, I'll come back to it later.

 I am now working on scrolling the record in memory. 

And somehow there is quite often a request for roll mode or long recording. 

I would appreciate some help with auto DC mods if anyone has time.

[bffargo]

Fargo, an image with data as I wrote above. (auto settings not work in DC mode?)

I will take one BNC T-piece, insert it into ch1, connect it to ch2, and from the BNC T-piece I will stretch the cable with alligator clips to the measuring device and the power source.

Yeah that's pretty much how I did it. DMM'd one side of the T and through various interconnects tied CH1 to CH2 over two T's. I just don't have exact parts to make it line up perfectly, but once I got it all it was stable and same power going into both sides. But the numbers seemingly below show that it was inconsistent for some reason? (just reading into the variances and seem to be correlated to having problems on one end for CH1 and other end for CH2)

[Atlan]

Ch1 5V and ch2 100mV are out of calibration.  For calibration, use a properly filtered voltage, possibly a 7815 and a potentiometer.  You have to remember that the oscilloscope takes a sample and when there is a peak, it calibrates according to it.

50mV is just 100mV * 2

I will modify the software tomorrow.  It would be a problem to calibrate this way.  I'm not sure that this software will let you calibrate. 

Now the only option is to record the 00h.bin mentioned about 2 pages ago into the oscilloscope, it will erase everything.  And you will recalibrate... 

That's why I wanted the loader to be supplemented with a button that confuses the configuration data.

[bffargo]

Am using a SMPS bench power supply with Owon bench DMM connected metering the voltage right at the BNC. I can switch to a Linear one if you think the frequency of the SMPS is causing an issue

[Atlan]

I wanted to use the rd6006, but when I saw what it had on the output, I used an analog source.  See text up.

[Atlan]

  V0. 021b mistake   

  0.021c ok 

Added possibility to reset calibration values. (if a bad calibration occurs) Added notification of a faulty calibration if the offset cannot be set (only if there is some hardware problem, bad ad converter, missing ad converter, problem with the offset voltage setting)

Input divider calibration is enabled when the oscilloscope is turned on for the first time and baseline calibration is used.

Calling up the calibration of the input divider is possible only through the menu by resetting the calibration values. (or by deleting the data on the sd card where the configuration data is located. NEED SDCART READER No It can be done with a card stored in the oscilloscope. )

After successful calibration of the input divider and switching off the oscilloscope, the data is written to the SD card. (Unless you change the burnt components on the input circuits, it will no longer be needed - that is, if you do not erase, do not replace the sd card)

If the calibration is started again, only the base level calibration will be performed. i.e. 0 is found. (What sometimes needs to be done)

By uploading the firmware, the calibration of the input divider DOES NOT CHANGE.

[facekim]

With the version updated to v0.021c, resetting the calibration in the USB menu does not work.
At least in my case.

Thank you

[Atlan]

Does the button respond to touch?  Make a video.

[facekim]

After returning to the factory firmware and putting the modified one again if it comes out.

Sorry
Thank you

[tokar]

With the version updated to v0.021c, resetting the calibration in the USB menu does not work. At least in my case.

I confirm.
 Procedure: v0021a, “Baseline calibration”, calibration of 0.3-15V dividers was successful.  I closed the menu, turned it off and turned on the device.  I click the menu - USB connection - PC window.  We get a table of constants. 
I'm flashing it via loader v0021c.  Notification enable, Baseline calibration, I get "Calibration successful".  Click the menu - USB connection - PC window.  No reaction.  I fill sector 700 ""00h". I repeat the above steps - there is no expected result.
I fill sectors from 16 to 2047 "00h". Turn it off - turn on the oscilloscope, the Fnirsi screen saver. Calibrate. Flash v0021c. I repeat the above steps - there is no expected result.

[Atlan]

It is no longer necessary to enable the notification.  Resetting the values ​​does not work, because they are written again during a correct shutdown.(use external sdcard reader) The option is to click on the screen starting in usb mode.  You have to click right in the middle of the pc screen.  Very precisely in the center of the pc screen.

[tokar]

Yes, it works. Thank you.