Help Wanted!

[gusmanb]

Hi all!

I'm preparing something that is going to increase the number of protocol analyzers massively, a bridge to allow the usage of Sigrok decoders 😄

The base concept is already working, the system is capable of loading the decoders, execute the analysis and represent the data, is still basic but it works. Here is a small example of the SPI decoder.

As you can see in the image the number of decoders supported is really high compared to the original ones, basically all the non-stacked decoders are supported (in the future I might also add support for the stacked ones, but let's go step by step 😃 )

I have created the bridge from the ground up just using the documentation, there is no code reused from Sigrok and thus, I surely have errors on the implementation.

And there is where I need your help.

The decoders for what I have data do work, but I would need data from more protocols/devices to test as many decoders as possible.
If you can provide a capture for any of these protocol/devices please, upload it to this thread, I will update the list of missing ones as the people provides them.

This is the list of decoders and the status of the data requirement:

Protocol/Device	Status	Provided by	Notes
Audio Codec '97	Missing
Aosong AM230x/DHTxx/RHTxx	Provided	fbgentle1
Advanced User Debugger	Missing
Atmel Program and Debug Interface	Missing
Digital calipers	Missing
Controller Area Network	Missing
HDMI-CEC	Missing
Compact Joint Test Action Group (IEEE 1149.7)	Missing
Edge counter	Provided	Dr. Gusman
Digital Addressable Lighting Interface	Missing
DCF77 time protocol	Missing
Digital Serial Interface	Missing
RFID EM4100	Missing
RFID EM4205/EM4305	Missing
FlexRay	Missing
Gray code and rotary encoder	Missing
Guess bitrate/baudrate	Missing
Inter-Integrated Circuit	Provided	Dr. Gusman	Tested with 7 bit addresses, 10 bit capture required
Integrated Interchip Sound	Provided	fbgentle1
IEEE-488 GPIB/HPIB/IEC	Missing
IR IRMP	Missing
IR NEC	Provided	fbgentle1
IR RC-5	Provided	fbgentle1
IR RC-6	Missing
Sony IR (SIRC)	Provided	fbgentle1
Timing jitter calculation	Missing
Joint Test Action Group (IEEE 1149.1)	Missing
NXP LFAST interface	Missing
Low Pin Count	Missing
SEGA Maple bus	Missing
Intel MCS-48	Missing
Management Data Input/Output	Missing
Microwire	Provided	fbgentle1
Miller encoding	Missing
Morse code	Provided	fbgentle1
Interpret bit patters as numbers or state enums	Missing
1-Wire serial communication bus (link layer)	Provided	fbgentle1
On-off keying	Missing
Parallel sync bus	Missing
Padded Jittering Data Link	Missing
PS/2	Missing
Pulse-width modulation	Provided	DeeJay & fbgentle1
Qi charger protocol	Missing
Remote control encoder	Missing
RGB LED string decoder (WS281x)	Provided	DeeJay
SAE J1850 VPW.	Missing
Siemens SDA 2506-5	Missing
Secure Digital card (SD mode)	Missing
Texas Instruments SDQ	Missing
7-segment display	Missing
Signature analysis	Missing
SLE44xx memory card	Missing
Sony/Philips Digital Interface Format	Missing
Serial Peripheral Interface	Provided	Dr. Gusman	Only mode (0,0) tested, more mode captures required
Sitronix ST7735	Missing
Stepper motor position / speed	Provided	fbgentle1
Serial Wire Debug	Missing
STM8 SWIM bus	Missing
RFID T55xx	Missing
Timing calculation with frequency and averaging	Missing
Texas Instruments TLC5620	Missing
Universal Asynchronous Receiver/Transmitter	Provided	Dr. Gusman
USB Power Delivery	Missing
Universal Serial Bus (LS/FS) signalling	Provided	mengstr	Low speed and full speed provided
Wiegand interface	Missing
Zilog Z80 CPU	Provided	fbgentle1

Once the ammount of decoders tested is enough to have confidence that the system is more or less stable I will release a beta to the users that request it. I will also announce it here.

The faster we get the captures the sooner that I will be able to release the version, so if you can provide anyu capture for one of the untested decoders, please contribute 😄

Thanks all!

[fbgentle1]

Hope this is what you are after.
I had to put a txt extension on the sigrok files
Cheers Bruce

RC5_Code.sr.txt
RC5_Code.pvs.txt

[gusmanb]

Hi!

Thanks, what I need is the capture done with LogicAnalyzer, not with sigrok :)

[fbgentle1]

Hi,
Hopefully this file is better :-)
Here is an RC5 capture once again I added .txt to the lac file so it could be uploaded.

Cheers Bruce
RC5_Code.lac.txt

[gusmanb]

Excellent! This is exactly what I need.

Thanks a lot :)

[fbgentle1]

Hi,
Here is an IIS trace for a Microphone with data on the Left channel and nothing on the Right.
Hope this is of some use.
Cheers Bruce
IIS_Left_Channel.lac.txt

[gusmanb]

Of course it is 😄

Thanks to these captures I'm fixing the implementation of the bridge, this capture for example raised a bug as I assumed that each annotation should be always in a row but the I2S decoder does not expose any row :)

Now it is fixed ;)

[fbgentle1]

Hi,
Here are a couple more IR protocols

Cheers Bruce
IR_SIR_Video_8.lac.txt
IR_NEC_Extended_Protocol.lac.txt

[fbgentle1]

Hi,

A few more for the collection.

Cheers Bruce
DHT11.lac.txt
OneWire_LinkLayer_DS1822.lac.txt

[gusmanb]

Excellen!

Thanks a lot :)
With your help things are evolving a lot faster.

[fbgentle1]

Hi,

Here is another one.

Cheers Bruce
Microwire_93C46_Read_Addr21_Data64b4.lac.txt

[fbgentle1]

Hi,
Here is a Rotary encoder E6B2-CWZ6C 1000 pulses per revolution trace
Cheers Bruce
Rotary_Encoder_1000pulse_rev_A_B_Z output.lac.txt

[fbgentle1]

Hi,
A stepper motor controller trace this time
Cheers Bruce
stepper_motor_controller.lac.txt

[fbgentle1]

Hi,
Here is a morse code trace.
Cheers Bruce
MorseCode_TimeUnit_0.01.lac.txt

[fbgentle1]

Hi,
Here is a PWM trace.
Cheers Bruce
PWM.lac.txt

[gusmanb]

Here it is: https://mega.nz/file/CPxnVIzJ#kmvk6tJaLVad2T17vGAXL30eoVOFeaJZygsTFYIbZXc

I have included a small txt with info on how to get it working.

Hope you like it (and don't find too many bugs 😆)

[fbgentle1]

Thanks,
I have downloaded it, will let you know how I go once it is setup and running.
Cheers Bruce

[fbgentle1]

Hi,
I have been trying a few of the decoders and they all work fine so far and look great.
I did have one issue with the Pico W board where it wouldn't capture any channel above 6 but I think it was to do with my system as reloading the firmware has fixed the issue.
I will continue testing but great work
Thanks Bruce

[gusmanb]

Perfect, thanks for the report. I have solved some bugs that I found that may be related to what you saw on the W, is it possible that you left a channel "empty"? a hole? There was a bug that caused problems in these cases.

I will send you this weekend a new version with all the fixes and also new capabilities (now you can hide/show the channels and the tooltips with all the info are ready, also I have changed the color parlette to something that I think is more readable).

Cheers!

[fbgentle1]

Yes, I definitely had an empty channel between the captured channels.
I look forward to the update.
Cheers Bruce

[sslupsky]

Have you had a look at the ngscopeclient project? Might be of interest to you as it appears to be actively maintained.
https://www.ngscopeclient.org
https://github.com/ngscopeclient

[gusmanb]

I think that with V6.0 everyone is going to be more than happy :)

I managed to get to work the full decoding system from Sigrok even the stacking, improved the rendering speed and a lot of new functionalities :)

[mengstr]

A KA and a standard packet sequence from a LS USB from a mouse. It's only sampled in 8Msps, but that should be enough to decode properly.

USB-LowSpeed.lac.txt

[mengstr]

USB Full Speed captured at 100Msps from an firmware update of a Black Magic Probe

USB-HighSpeed.lac.txt

[gusmanb]

Perfect, this one works much better :D

[mengstr]

There's no smartcard ISO 7816 (wired) entry in the table above, but sigrok do have one - https://github.com/svenso/sigrok_iso7816

It's very hard to capture any decent amount of data considering the standard clock/data ratio is 372 clocks per databit. The reader I used has a clock at approx 5MHz so I sampled this file at 25Msps. Hopefully there's enough data in the ADPU to have the decoder at least start doing its job.

ISO7816.lac.txt

[gusmanb]

The list is extracted from the included decoders on the PulseView distro, that particular one is not included.

In any case I added it manually and it seems to be working :)

Also, I have already implemented the pinning of the preview and the info of the current pulse :D

[mengstr]

Nice! This is definitely a case where a much larger memory buffer is needed, or a synchronous clock (but that would only get 5x more samples which is probably not enough to capture the entire ATR message).

[fbgentle1]

Hi,
Here is a Z80 capture, it doesn't have the address lines captured but they are optional.
The decoding was correct and couldn't see any issues but thought it may be of interest as it is a little more complex capture.
Cheers Bruce

Z80_CPM.lac.txt

[DeeJay]

I think you are homing in on the issue.

I cannot run the tests you propose.

From a Command Prompt session -

Check python

Microsoft Windows [Version 10.0.19045.4780]
(c) Microsoft Corporation. All rights reserved.

C:\Users\David>python
Python 3.9.2 (tags/v3.9.2:1a79785, Feb 19 2021, 13:44:55) [MSC v.1928 64 bit (AMD64)] on win32
Type "help", "copyright", "credits" or "license" for more information.
>>>

Check python3

Microsoft Windows [Version 10.0.19045.4780]
(c) Microsoft Corporation. All rights reserved.

C:\Users\David>python3

C:\Users\David>

The latter fails to invoke a python interpreter and brings up a Microsoft Store window proposing the installation of Python 3.12.6

Separately, I have done a Windows file Explorer search for 'python3*.dll' and found 6 matches.

2 are for the version in AppData\Local\Programs\Python: ...\Python39\python3.dll and ...\Python39\python39.dll

2 are related to Thonny - a python-oriented IDE that provides its own python interpreter - python310

2 are related to LibreOffice - an MS Office 'clone' - python35

But I suspect it is the simple fact that 'python3' does not map to a python interpreter that is the problem for LAv6.0

I don't know how I innocently created this situation. But maybe other potential users will have done the same thing.

[gusmanb]

I think is not you who caused this situation but MS, I bet you they added some kind of hard link to the store if you invoke python3 and is not installed with that name...

But now it's clear what is happening and I know how to solve it, instead of once an executable that matches is found, stop trying the others and fail, if is not correct I'm going to change the procedure to do the full detection on both possibilities, this should solve this problem.

Once I get the new detection routine ready I will upload it here.

Thanks a lot for your help 😄

[gusmanb]

Here it is:

https://mega.nz/file/yOYFyJqD#XUafuycj8lXLk3lilIS4rCwb00AsYSp-Xfi8GxG7nnY

This one will test both interpreter names and even in the case that there are two valid Python 3 installs it will choose the best one that can find.

Also, the log is now called "PythonInitLog.txt", I think that until this is well tested I'm going to leave it on as it is really useful.

Cheers.

[DeeJay]

Starts up with no warning dialog and with Sigrok decoders loaded into a panel at the Right side of the window.

Log file shows -

10/09/2024 11:35 -> Initializing python system...
10/09/2024 11:35 -> Initializing python installation detection...
10/09/2024 11:35 -> Target Interpreter: python,  Initializing Python Installation Detection...
10/09/2024 11:35 -> Target Interpreter: python,  Python version found: 39
10/09/2024 11:35 -> Target Interpreter: python,  Python version parsed: 3.9
10/09/2024 11:35 -> Target Interpreter: python,  Finding python library on Windows...
10/09/2024 11:35 -> Target Interpreter: python,  Script output: C:\Users\David\AppData\Local\Programs\Python\Python39\Python39.dll
10/09/2024 11:35 -> Target Interpreter: python,  Final python path: C:\Users\David\AppData\Local\Programs\Python\Python39\Python39.dll
10/09/2024 11:35 -> Target Interpreter: python,  Python path found: C:\Users\David\AppData\Local\Programs\Python\Python39\Python39.dll
10/09/2024 11:35 -> Target Interpreter: python3,  Initializing Python Installation Detection...
10/09/2024 11:35 -> Target Interpreter: python3,  Python version not found, aborting.
10/09/2024 11:35 -> Valid installations found: 1
10/09/2024 11:35 -> Selected version: 3.9
10/09/2024 11:35 -> Stablished path: C:\Users\David\AppData\Local\Programs\Python\Python39\Python39.dll
10/09/2024 11:35 -> Initializing decoders...
10/09/2024 11:35 -> Python initialization completed.

Tested with contents of unpacked archive only. I did NOT copy across the python.cfg file, and one was not present after the app exited.

[gusmanb]

Excellent.

Thanks a lot for your help 😄

[DeeJay]

Returning to the topic of getting trial data for device decoding...

RGB LED string decoder (WS281x), captured with LA60pre. (The Sigrok decoder in LA60pre is able to interpret this capture.)
(I can provide a v5.1 version of the capture on request.)

RGB_LED_WS281x.zip

The pattern sequence is intended to be [OFF, OFF, State_A, State_B, State_C, State_C, State_B, State_A, OFF, OFF] . Part of this sequence is correctly shown in the capture.

[DeeJay]

Pulse-width modulation

I think you have had one offer, but it is still marked 'missing.'

Here is another offering: 4 channels, all 100Hz base frequency, with different Duty Cycles as implied by the Channel name tags.

Capture performed by released version 5.1 of LA. The LA60pre2 code can Open this capture and correctly decode it.

PWM_duty_cycles.zip

[henrygab]

There are many, many traces of RFID (EM4100, EM4205/EM4305 included) here:

https://github.com/RfidResearchGroup/proxmark3/tree/master/traces

The .pm3 files are all single-channel traces:

• One sample per line
• Each line is ASCII, encoding a single digit, typically in range [127,-127].
• For EM41xx/EM42xx/EM43xx, the sample rate is 125kHz.

Direct links to relevant traces

Can you convert those traces to what you need for RFID (EM4100, EM4205/EM4305) samples?

• https://github.com/RfidResearchGroup/proxmark3/tree/master/traces/lf_ATA5577_em410x.pm3
• https://github.com/RfidResearchGroup/proxmark3/tree/master/traces/lf_EM4102-1.pm3
• https://github.com/RfidResearchGroup/proxmark3/tree/master/traces/lf_EM4102-2.pm3
• https://github.com/RfidResearchGroup/proxmark3/tree/master/traces/lf_EM4102-3.pm3
• https://github.com/RfidResearchGroup/proxmark3/tree/master/traces/lf_EM4102-clamshell.pm3
• https://github.com/RfidResearchGroup/proxmark3/tree/master/traces/lf_EM4102-fob.pm3
• https://github.com/RfidResearchGroup/proxmark3/tree/master/traces/lf_EM4102-thin.pm3
• https://github.com/RfidResearchGroup/proxmark3/tree/master/traces/lf_EM4305_fdxa_destron.pm3
• https://github.com/RfidResearchGroup/proxmark3/tree/master/traces/lf_EM4x05.pm3
• https://github.com/RfidResearchGroup/proxmark3/tree/master/traces/lf_EM4x50.pm3

[gusmanb]

Will check it, thanks for the info.

[henrygab]

There are a number of traces specifically showing various modulations:

Biphase, Manchester, FSK (frequency-shift keying), PSK (phase-shift keying), NRZ, ...

filename	description
lf_Q5_mod-*	Q5 configured to emit 00 01 02 03 04 05 06 07 08 09 0A 0B under various modulation schemes

If you have a script to convert .PM3 files to your preferred format, that should give quite a nice selection of patterns for testing decoders.

Note: In past, I've noticed sometimes the ASCII value in .pm3 files slightly exceeds the [127,-127] range; I recommend checking if the values are >95% in that range and 100% within the range [135,-135], and if so, cap the values to range [127,-127].

[henrygab]

[FYI update (no response expected)]

I've submitted PR 2518 to that proxmark repository, which was merged.
Now all the .pm3 files keep their values in the range [127,-128], so each line can be converted to a single signed byte value.

[omicronns]

Here is SWD (Serial Wire Debug) sample, captured on STM32F0308-DISCO, when connecting with the target using STM32CubeProgrammer.
swd_sample1.zip