FNIRSI-1013D "100MHz" tablet oscilloscope

[Atlan]

First you need to disconnect the probes from the oscilloscope, then the basic calibration will start and it should finish successfully, then a message will appear and you will be prompted to connect voltage to both channels of the oscilloscope at once. Or record a video so we can see what you are doing.

Do not enable DEV mode. It is not intended for regular users.

[Sorgelig]

First you need to disconnect the probes from the oscilloscope, then the basic calibration will start and it should finish successfully

Probes and USB were disconnected (i know this requirement as i use this Oscilloscope for many years). Baseline calibration is always failed. I tried to use "Clear / Return to factory FW" then upload 0.26v7 again. Still failed to calibrate.

Do not enable DEV mode. It is not intended for regular users.

I've tried with dev mode to see if i get any additional info. Unfortunately, no any additional clues provided in dev mode. Of course i've tried with dev mode disabled as well.

Or record a video so we can see what you are doing.

I'm not sure what exactly i should record. It's just starting Oscilloscope, then going to menu -> Baseline calibration -> failed. Everything is standard. If there is more info i can provide, tell me where i should look for that. I wonder why calibration sector is always empty. Can it be something to do with SD card? Although original firmware has no problem with calibration. What additionally your firmware does which original FW doesn't? Then may be it will give some clue.

Is there a way to get extended info about cause of calibration failure?

[Sorgelig]

I just quickly checked the source code and one thing caught my attention. It is adc2rawaverage which seems gets some previous/setting value and if new calibrated value is too far then it produces a fail. So if settings are cleared or corrupted then possibly new calibration will always fail.

[Atlan]

If that wasn't a problem for you, a video would help, for sure.

So you don't even get to the prompt about connecting voltages to the oscilloscope inputs?

The SD card must be 100% functional. Maybe your reference voltages are out of tolerance. If you can wait a bit (a few days) I will upload the firmware with a wider calibration range here.

[Sorgelig]

If you can wait a bit (a few days) I will upload the firmware with a wider calibration range here.

Sure i can wait. Thanks to your firmware which adds multi-boot so i can continue to use an original FW.

If that wasn't a problem for you, a video would help, for sure.
So you don't even get to the prompt about connecting voltages to the oscilloscope inputs?

I will wait for your new FW first and if it won't help then will make a video. I didn't even know it would require some reference voltage to be connected at later stage. It fails at first stage of baseline calibration right after that relay ticking finished.

My original motivation to go to custom FW was due to a bug in original FW when baseline get screwed at some point, so i often have to re-calibrate a baseline. I hope this bug is fixed in your FW, as it's the main my concern.

[Sorgelig]

Ok. Here is the video.
https://drive.google.com/file/d/1h6sM0e-1uivAornCN9yfDmN4-hHIrJaU/view?usp=drive_link

[Atlan]

Don't you mind, Fnirsi, that at 50mV in long mode it shows half as much  and that it doesn't show below a certain signal level 

The video is not accessible.

When the battery is discharged, the signal line starts to move and calibration is necessary. At a certain battery voltage, even this does not help. It is possible to install a step up converter, see the firmware manual.

[Sorgelig]

For me it doesn't look like "drifting", it suddenly jump either half of the screen up, or even beyond the screen. While other (usually inactive) channel still shows baseline in correct position. Because it never been like slowly drifting but sudden jump, i didn't think about battery level causing that.

[Sorgelig]

Ok, added step-up converter according to pic.
One thing need to mention for that mod: battery trace connected to converter is always live, so converter will suck the battery even if oscilloscope is off. To eliminate it (or at least greatly reduce) you have to use converter with EN(enable) pin. Connect this pin to Pin 1(EN pin) of MIC23902. Alternatively, converter's IN+ can be connected to the middle pin of power switch.

Waiting for your updated FW as baseline calibration still fails.

[Sorgelig]

Here is the picture of better DC-DC converter. It has dual-inductor design which allows true power off when EN pin is low.
It also can do both step-up and step-down conversion (which is actually not important in our case).

[Atlan]

Try this firmware 0.26v9, it has an added description during BASE calibration, fail1 and fail 2 write down which one it says on which channel.

ATTENTION: the calibration data will also be erased from the configuration sector for the display, i.e. sectors 709 and 710.

[Sorgelig]

Both channels give "Fail-1" error

[Atlan]

that's about this error
FAIL1
//Check if the average reading is outside allowed range
if((calibrationsettings.adc1rawaverage < 125) || (calibrationsettings.adc1rawaverage > 131))

FAIL2
  //Split the difference on the two ADC's
  calibrationsettings.adc1compensation = compensationsum / 2;
  calibrationsettings.adc2compensation = -1 * (compensationsum - calibrationsettings.adc1compensation);

  //Check if the found result is within limits
  if((compensationsum < -20) || (compensationsum > 20))

I suspect a bad battery, or a bad contact on it. If you run the calibration with the USB charging connection, will it not be successfully? The problem will be in the hardware of your oscilloscope.

Try this firmware.
Let the base calibration run (it should be successful), if you see a prompt to connect power, tap outside the calibration screen. The window will close. And upload a photo of the diagnostic screen here

[Sorgelig]

FW 0.26vA still fails with "Fail-1" error, so calibration doesn't advance to "connect power" stage.

I suspect a bad battery, or a bad contact on it. If you run the calibration with the USB charging connection, will it not be successfully? The problem will be in the hardware of your oscilloscope.

I've tried with USB connection and without - same result.
I would believe in hardware problem but original firmware works ok. Baseline calibration always succeeded. Do you have source code of original FW? May be try to roll-back calibration procedure to original one to see how it will work? I remember first versions of custom firmware had no problem with calibration.

[Sorgelig]

Can you make a FW which will output raw ADC value in real time, so i can provide such info?
May be add this to diagnostic view?

[Sorgelig]

btw, i don't know if there is a config interface to setup ADC, so problem may lay not in reading ADC but in programming some chip? May be interface (such as i2c or SPI) uses too high speed and fail to program some chip/adc. Even if i see a baseline near the top of screen, it doesn't show any waveform regardless what signal i try to feed to it. So it looks like something else is not correctly set, not ADC.

[Atlan]

Download the schematic that PCprogramer has here. And start measuring.
Would you mind showing me a picture of what the signal looks like when it's not working?
In long mode, signal interference does not occur?
Cold junction on reference voltage?

Can't you upload a video to YouTube?

[Atlan]

I don't know if I uploaded the correct version, so one more time B

Don't download it or use it, it's just a version for him.

[Sorgelig]

0.26vB is the same as previous, fail-1.

for some reason i can't upload to youtube.
here is the video:
https://drive.google.com/file/d/1GO7Ut3oyfyw4nR3mZlJ8cwAyqW6VKBP5/view?usp=drive_link

You can see both your FW and original FW work. As you see Original FW is fine, so there is no problem with HW.

[Sorgelig]

May be my oscilloscope is very old and has other FPGA bitstream? FPGA chip has no name, just long number. Probably AL3?

[Atlan]

I can't watch the video, it keeps saying I should ask the owner for access. Don't you have it publicly accessible?

A photo of the entire PCB from the side where the FPGA and uP and AD converters are located.

Spustite PECO firmware-LEN z baterie, ( nastavte kanaly na AC,50mV, 20MSa/s 5us/div, trigger do stredu obrazovky)chodte do diagnostiky, tam stlacte RESET, ADC comp by sa mali nastavit na 0. DAjte Exit. A ulozte priebeh ako WAVE. Urobte to pri viacerych nastavenia casovej zakladne napr 200KSa, 50MSa, 200MSa/100ns/div. Celkovo 4 wave subory tu nahrajte.

Run PECO firmware - ONLY from battery, (set channels to AC, 50mV, 20MSa/s 5us/div, trigger to the center of the screen) go to diagnostics, press RESET there, ADC comp should be set to 0. EXIT. And save the waveform as WAVE. Do this for multiple timebase settings e.g. 200KSa, 50MSa, 200MSa/100ns/div. Upload a total of 4 wave files here. Thank you

[Sorgelig]

sorry, for link.
Now it should be accessible to everyone.

Later I will try to do what you asked.
Can you give me link to exact PECO firmware version i have to use?

[Atlan]

the version doesn't matter, one of the latest. Feel free to upload the one you have there now. Otherwise, where did you get the oscilloscope? Couldn't the input part be damaged? Why is the test signal uncompensated since the probe is in 1X mode according to the FNIRSI firmware.

[Sorgelig]

I'm not sure why you call it PECO firmware. My Oscilloscope is FNIRSI, and i don't see any other brand there. Do you mean original FNIRSI firmware?
The firmware which works is the one i get when i choose Clear/Restore FW in LOADER. What version get uploaded in that option?

[Atlan]

PECO is the name for the alternative firmware. Can you explain why in the video provided (original fnirsi firmware) the rectangular signal from the test output is distorted?

Additionally, in your video, the signal is obviously shifted relative to the channel axis. So, as I said, download the schematic and check the reference voltage values.