carstenscharlemann / climb-obc-hardwaretest Goto Github PK

To get recovery from bit flips in CPU Configuration registers there should be a mechanism to re-initialize all needed (IO/Configuration-) Registers.
It would also be nice to have a check prior to re-initialization. (Not sure if this can be done with all registers but where it works it would be another useful indicator (error counter) to record during the radiation tests.

U701-U703 24AA64T-I/MC 64Kb (8K x 8) EEPROM

to be used by Thruster, COM (StacieA-C IF), CLI and/or Stacie-S IF.
It should be usable with Uart0 .. 3 and have a configurable TX ringbuffer.

Interconnect UART-A to UART-C externally and send simple test strings to verify the functionality of the UART interface and hardware.

In the end we should be able to

• read/set time
• recover a 'epoch time' after resets
• indicate if and why recovery of epoch time was not successful after different kind of resets.

For the radiation tests the epoch time and recovery is not yet important/needed but it would be good to have a robust 'Sternzeit-module' from early on .....

Use an external RS485 convertert at the LogBerry to verify functionality of this interface with simple test strings.

The repository ado-chip-175x-6x has new adapted Layer1 layer2 routines.
The Idea here is to use the source code of the other repository here as a base library (git sub module ?) and change the OBC Hardware test image over to make usage of this modules.

Add following message to be transmitted regularly from the OBC to the LogBerry: "Supervision watchdog feed". This resets the external watchdog on the LogBerry which triggers a power cycle after 5 minutes without receiving this message.

IC403 - 1Mbit MRAM

maybe this is an viable option if we need massive external data collection (for dosimeters or other payload).

Tests with the clkOut command first time showed no Output when switched to "clkOut RTC".

After touching the Pins of the RTC Xtal it was ok.
Following tests with several PowerDown/Up cycles and also with Debug sessions restarts and so on I never could reproduced the problem a second time.

Raditaion test: Make validation of content in RTC - GPR[0] ... GPR[4]

On the OBC we have a RGB LED WS2812 connected to PIN67 (IO Port P2[6]).
It would be nice to have a function available to control this LED by means of setting RGB value.

(on the LPCX board there is a RGB led with 3 output pins able to get 7 different colors. We also could write an additional 'Hardware abstraction' here for having same routines work on both boards...)

During development of radiation tests, I tried to use Flash and Mram memory at the same time and run into errors with bus blocking. I think the logic from IRQ - Layer1 - Layer2 signalling - especially in regard to busy bus / wait for unload sender in IRQ - is not working stable yet!

Workaround for radiation test: use all connect devices (MRAM - Flash1 - Flash2) strictly sequential!

Implement CLI commands to set/reset GPIO pins for commissioning tests. Especially important for CS and GPIO pins, but may also be used to test the solder connection of arbitrary LPC pins which are accessible for measurement.

A command in the way of test(PX[xx], state) would be nice. X is the Port number, xx the pin number and state is the output state 0/1. The command should initialize the PIN as GPIO, set the state accordingly.

at this time we have EEprom 1-3,FRAM, Mram, Flash1, Flash2.
still missing. #20 and optional SD-Card.

It seems somebody messed around with fimware update. It shows an MBED Drive (comming up if connected) which disconnects-reconnects and then shows a fail.txt file containing

SWD Error

In order to get good result data on RTC - calibration and XTAL timing of the OBCs under test the timestamps logged by the LogBerries should be as accurate as possible.

TBD.
The feature list for the thermal vacuum tests with a thruster has to be defined and implemented.

Extend board.c by all relevant pins including their initialisation for EM1.

As I understood what we want, I am going to implement the 'Quick Setup Example' from Datasheet Page 8-9. This will be:

• 'High Sensitivity' Mode
• 1 Second measurement window
• Ref Osc 50kHz

@andreassinn : is this ok?

In order to get a command Line interface we can use the UART Side B as debug interface for commanding and output. There should be a command loop similar to PEG but with better cmd structure (better fitting later COM Packages).

constantly validate checksum over internal FLASH memory

When MCUX is used under Linux this repos does not compile. Reason are several 'typos' mainly within include directives asking for .h files with wrong capitalized letters in it!

Readout and
transmit TMP100 temperature from I2C regularly.

Make a module controlling the Radiation test. It should be able to

• have (compile time) Configuration of different timings (e.g. Read/Write cycles per min/hour and per memory type)
• Make Block read tests on all Memory types (with 2 or 3 different patterns?)
• Make Block write tests on all Memory types (with 2 or 3 different patterns?)

The block read/write will be provided by 'Layer2'-modules. (see and start with EEPROM which is also available on the LPCX board)

When the software does not touch the WATCHDOG_FEED pin at all the watchdog reset is never triggered.

The green LED does light up for a very short time, but it seems that the first monoflop (IC5-1) is not triggered during normal power up operation (via USB - LPC-LINK2).

The idea here is to provide

1. Some Indication to the outside world, that software still operates nominal - (message on UART or regular toggling of dedicated IO-Pin
2. Some Indication of the various memory test results (checksum ok-failed counts operational read/write errors ....) -> I propose messages on UART B (TXD2/RXD2 pin 64/65)
3. optional commands to control some aspects of the tests during test run. (e.g. Enable/disable some parts of tests - control speed/treshholds of tests, ...

The concrete content here is TBD and I herewith open the discussion under this issue ;-)

Add measurement readout and regular transmission temperature and supply currents, acquired by ADC.
Code is available.

To prepare COM tests we should port and adapt PEG Software to make a module structure ready for testing communications with STACIE hardware.

checksum tests for IC401/IC402 - S70FL01GSAGMF I011 - 2x1GBit Flash

@alexsous : I think we need another session (possible with skype) to get you started on this one 😃

Regularly transmit state of following signals:
RBF
Supply Rail
Bootloder Selection

Pin definitions and code available

Implement remote access (e.g. Teamviewer) using a PC provided by Carsten, which will be placed in the control room at Seibersdorf. The three LogBerries are connected via a switch to the PC. SSH and FTP to access the LogBerries should be sufficient.

@andreassinn : This could be our base for adding features in regard to OBC board commissioning.

At this moment only CLI runs (on UART2) and nothing else gets initialized or tested.

My plan is to move more and more modules from the test_adolib_lpcexpresso1769 project to the lib and make them available here. Pls. let me know if and what you are expecting first.

The Layer1 I2C job queues and Interrupt routines can be adapted to work without RTOS in this Hardware test image.

This buses are used to connect the different memory chips:
SSP0:

• Flash2 (IC401)
• EEC Flash (IC7)
• MRAM (IC403)

SSP1:

• Flash1 (IC403)

IC55 - CY15B128J 128kbit F-RAM

Tha aim is to have possibility for everybody to use this evaluation board.

The main issue is to find a free debug UART (for printf...) and to decide which UART to use for Thruster test.

Stuck in its 'Disabled' position. Maybe still operational with a fine instrument (needle?) ....

I tried to set up a Wiki page with the very basic instructions how to get started with the MCUX SW and work with the debugger boards.

However there is no "edit" or "new page" option available for my account. Maybe permissions have to be set...