FNIRSI-1013D "100MHz" tablet oscilloscope

[Atlan]

But this one with the SD card, we are not interested at all. Especially when it comes to desoldering. The flash for the FPGA is at the bottom, but that part is not visible in the photo. So take a photo of the bottom part as well, ideally with the flash marking.

[mishail71]

Thanks, Atlant... I just flashed the wrong SPI... Now everything works... However, I have this picture

[TriNitron]

Hello everyone! I have an EFL2 FPGA. I tried desoldering the flash memory and soldering four wires, but the firmware isn't active.

Please tell me how exactly I can reflash the EFL2 FPGA, which pins should I solder the wires to, what inexpensive programmer is suitable, and what Windows software is recommended?

And most importantly, what will this do? Will the FFT work? Will the frequency counter increase its accuracy to three decimal places, at least to 100 kHz? Will it improve signal stability?

[Atlan]

See posts 3480 and 3483 for the latest files.

 EF2 is not currently supported (edit: it is already supported) And it won't be until the CH1 issue is fixed.

That's why I needed some testers.

[TriNitron]

Thanks, Atlan!
I can become an EFL2 tester, but I need to figure out how much an EFL2 programmer would cost, and whether it's possible to solder wires to specific pins instead of desoldering the entire chip.

What I don't like about this oscilloscope:
1 - Low frequency meter resolution. For example, the quartz crystal frequency is 32768 Hz, but the oscilloscope shows 32.7 Hz. With such accuracy, you could measure the frequency with a mercury thermometer))

2 - Horizontal signal jitter.

3 - Half-wave asymmetry. For example, on a sine wave,
Vmax + 510 mV,
Vmin - 470 mV,
but they should be the same.

These are fundamental aspects.

The rest, like FFT, oscillator, and so on, don't make much sense in the context of the oscilloscope's purpose as a measuring instrument. Why do I need a generator if the signal is jittering, and I don't understand whether it's jitter or an oscilloscope glitch, and what steps I need to take next... look for the problem in the circuit, or return the Fnirsi 1013d to a toy store for children, where it should have been sold in the first place.

[Atlan]

Otherwise, the things you mention don't require changing the firmware in the FPGA.

You will need a JTAG programmer. You can find some here for a regular uP stm32F103
https://github.com/AnlogicInfo/anlogic-usbjtag

Then the AnLogic TD program which you can download from the AnLogic website after registration.
license key from anlogic website The license is valid until January 2026, so we don't have much time.
https://www.eevblog.com/forum/fpga/reverse-engineering-anlogic-al3_10-fpga/

1013D with EF2 has a JTAG connector on the PCB, download the datasheet and see what the pinout is on the JTAG connector

Just do everything at your own risk

If you decide to go for it, download the FLASH FPGA content and post it here on the forum, I can generate alternative firmware and compare it to what you post here. and possibly provide you with the generated firmware so you can try it on EF2.

I don't know if you should stick to your last sentence.

[pcprogrammer]

The license is valid until January 2026, so we don't have much time.

Hopefully they will renew it before it runs out. I got this current one from their website download section earlier this year.

[mishail71]

 doesn't work scope v1.00n3.bin
When I go to the oscilloscope, I get a black screen...

[TriNitron]

Atlan, didn't you extract the EFL2 source code?

There's an FPGA/EFL2 folder on your Github page.

And why does connecting the JTAG connectors with wires on the board work for the AL5 but not for the EFL2, if the standard is the same? Is the pinout different?

Yes, the AL5 has a separate flash memory, while the EFL2 has it integrated into the chip, but is the JTAG standard the same for both?

I simply modified my board as required for flashing the AL5 firmware: I desoldered the flash memory with the Fnirsi operating system and soldered four wires between the JTAG connectors on the board.
When I tried to extract FPGA.bin, I saw missing files, or the DOWNLOAD button was inactive. Does this mean the EFL2 firmware is read-only, without an AES key?

Just buying a clone of the Altera USB Blaster programmer costs about $20-40. I bought this oscilloscope for $80 and will probably sell it for $50.

And in the context of all this, I want to know, is it even worth starting this journey if, with a reflashed EFL2, the frequency counter will remain the same XX.X, the vertical signal jitter will remain, and so will the asymmetry? What's the point then?
Purely for fun?

[TriNitron]

FPGA EFL2

I have updater ver0.02v
the fnirsi spi option is gone, leaving Fnirsi FPGA. If I download it, it downloads a 2MB FPGA.bin file, but it's actually Fnirsi spi, downloading from the Fnirsi flash chip.

I installed
scope ver. 1.00n3

The generator button works, the internal interface responds, but the actual signal from the pin is 1 kHz, regardless of the settings, whether enabled or disabled.

However, the oscilloscope inputs are inconsistent; I've changed the from 3000 to 15000. When changing the horizontal scale, the frequency counter readings change significantly: 1.00 kHz/659 hz/1 hz, and so on.

I tested both channels separately. Sometimes the oscilloscope waveform blinks red, but it's within the measurement range.

Dots works, PRO - it's unclear what it does.

[Atlan]

 doesn't work scope v1.00n3.bin
When I go to the oscilloscope, I get a black screen...

Please download version 1.00n3 again. It may have been corrupted during download. I will upload the full package later.

[Atlan]

FPGA EFL2

1. I have updater ver0.02v
the fnirsi spi option is gone, leaving Fnirsi FPGA. If I download it, it downloads a 2MB FPGA.bin file, but it's actually Fnirsi spi, downloading from the Fnirsi flash chip.

I installed
scope ver. 1.00n3

2. The generator button works, the internal interface responds, but the actual signal from the pin is 1 kHz, regardless of the settings, whether enabled or disabled.

3. However, the oscilloscope inputs are inconsistent; I've changed the from 3000 to 15000. When changing the horizontal scale, the frequency counter readings change significantly: 1.00 kHz/659 hz/1 hz, and so on.

4. I tested both channels separately. Sometimes the oscilloscope waveform blinks red, but it's within the measurement range.

Dots works, PRO - it's unclear what it does.

1. In order to download the flash for the FPGA, it is necessary to unsolder the flash for the uP and make the connections 4-wire, BUT THIS ONLY APPLIES TO AL3 and not EF2!!!!

2. It only works with alternative firmware, which you don't have there because it is stored INTERNALLY in the FPGA. See the post I posted yesterday on how to change it.

3. Not supported in native FPGA firmware

4 I know about it, it needs a small software modification to support the original FPGA firmware.

[Atlan]

New FPGA firmware with version support, new DSO firmware with fpga firmware detection. Diagnostic menu is information about fpga firmware

[Atlan]

Atlan, didn't you extract the EFL2 source code?

There's an FPGA/EFL2 folder on your Github page.

And why does connecting the JTAG connectors with wires on the board work for the AL5 but not for the EFL2, if the standard is the same? Is the pinout different?

Yes, the AL5 has a separate flash memory, while the EFL2 has it integrated into the chip, but is the JTAG standard the same for both?

I simply modified my board as required for flashing the AL5 firmware: I desoldered the flash memory with the Fnirsi operating system and soldered four wires between the JTAG connectors on the board.
When I tried to extract FPGA.bin, I saw missing files, or the DOWNLOAD button was inactive. Does this mean the EFL2 firmware is read-only, without an AES key?

Just buying a clone of the Altera USB Blaster programmer costs about $20-40. I bought this oscilloscope for $80 and will probably sell it for $50.

And in the context of all this, I want to know, is it even worth starting this journey if, with a reflashed EFL2, the frequency counter will remain the same XX.X, the vertical signal jitter will remain, and so will the asymmetry? What's the point then?
Purely for fun?

The folder is there but it is not updated. There is no point in producing code for EF2 if it is not debugged for AL3. I do not need to have a jtag programmer connected to AL3, I can change the FPGA content directly from the oscilloscope interface. Which is convenient. Most importantly, I can always return the original fpga firmware.

But the oscilloscope with EF2 has a connector for jtag. The oscilloscope with AL3 has a connector for SPI flash 25vq80 programming.
The programmer doesn't cost that much and can be used, for example, with Stm32, see the post above on the page.

Because you can program whatever you want. The oscilloscope turns on extremely quickly. It's a shame that the input part is so bad. The whole design is.......
 It would be possible to place a frequency counter on the screen somewhere. By taking more samples, it is possible to look at individual bits, for example, on an i2c signal.

As I already wrote with. The smaller display for the 99E is an oscilloscope that works and has bus decoding. https://www.dreamsourcelab.com/shop/oscilloscope/dstouch-ds4t252/

[mishail71]

I installed new ALL6 files and saw these artifacts on the screen...

The probe is connected to the first channel...1 kHz

[Atlan]

Where is the diagnostic screen image?

[mishail71]

Unfortunately, the translation can't convey what you said. I'm at a loss to tell you...

[Atlan]

Have you read the firmware manual? Just click on the link in the description below.

[mishail71]

Aha, I figured it out a bit!!!
For some reason, my FPGA file is not activating?

[Atlan]

Use a x10 probe at higher generator frequencies

[Atlan]

Upload a video. I'm sick, so I don't feel like watching it.

[Felipe Lacerda]

Here are a few observations regarding the current firmware ver:

- measurements on both channels are mostly correct with memory depth set to 3000.
- the internal generator shows abnormal behavior when +duty is below 10% and the frequency is above 1 MHz - the output starts pulsing and the waveform shape becomes incorrect.
- at 4 MHz and +duty 10%, the output frequency jumps from 4 MHz to 2 MHz (see attached screenshot).
- occasionally, it is not possible to move the waveform vertically on the display (this occurs randomly, I cannot reliably reproduce the exact conditions).
- at frequencies above 5 MHz and +duty below 5% with MEM3000, the measurements start to become unreliable on both channels.
- with MEM settings above 6000, the measurements are unreliable on both channels at any frequency and duty cycle.

In general, the lower the duty cycle and the higher the frequency, the more unreliable the measurements become.

- the displayed waveform is horizontally unstable, even though the signal on the reference oscilloscope is perfectly stable with no horizontal shifts.

Could this issue be why the manufacturer limited the memory to 3000?

[TriNitron]

Atlan, get well soon!
Your interface is working well, everything is user-friendly. It's a shame that EFL2 is so difficult to update.

Maybe it would make sense to create separate scope.bin versions for AL3 and EFL2 as well?

So, the current stable version for EFL2 is scope_v0.27n?

And judging by the fact that the project is developing with the AL3 FPGA, EFL2 owners shouldn't bother checking this out for the next couple of months to six months. After all, you've done everything that could be done without changing the FPGA, as I understand it.

[Atlan]

fixed overflow when the number of samples exceeds 3000, now the frequency and time calculation time+ and time- should work correctly, as should duty.

Report any other bugs if you find them.

[Atlan]

The touch must be moved on the signal axis - according to the cursor to the right