Burns,
I just got back home for a day long visit to Dallas and UTD to take care of my other volunteer projects.
Sounds like all good news. I am very happy to see this!
Comments back from below:
1. Good on the temp sensor reading. The temp sensor you are reading is the one down next to the AX5043 transmitter chip. There is a second temp sensor at 0x91. The code will need to be changed to support this one eventually. The second temp sensor is next to the CPU chip up between memory and the CPU. You should be able to read this one also by just changing the address in the code from 0x90 to 0x91. Chris can add the second one later.
We can move the I2C address to another address within the range of addresses provided on a later build.
We can coordinate adding an external A/D if needed but there are several A/Ds built into the TMS570. Three of these are being used now to monitor supply voltages. The A/Ds are using 3.3V and GND for reference. The 5V A/D is AD1/N_12 and should be at 2.5V for the middle of the range. The actual 5V supply may be running a little low though so don’t be surprised if lower. The 3.3V supply is on AD1N_4. Again this should read mid range on the A/D which will be 3.3 / 2 or 1.65 volts. The 1.2 V supply is on AD1N_13 and should read 1.2 / 3.3 * (2^12).
This is something you can work on testing but should work as designed.
There is an option later to use two of the A/Ds on a critical measurement and then use the redundant capabilities of this CPU to make sure the readings are accurate. We could also monitor other things with more A/Ds but this should be enough for a prototype. The AX5043 readings we might measure with analog readings are available in registers I believe.
2. The one wire pull-up issue should be addressed by activating the Pull up on N2HET1_14. We could add another pull-up on the board later. I thought the idea was to have a security string on the PCB that could be matched with ground transmissions. Dropping chips if fine though as it will save current. Not much saved here though.
3. The SPI lines for the AX5043 are available on J801. We would have to add the memory SPI if needed. J801 is on one of the schematic pages.
We can add the I2C bus to the PC104 if a pin set is defined and the signals are needed there. We will have to coordinate this with others. I could do this with a couple of jumpers if thee is an advantage to it now. We need to be careful with two Masters and adding capacitance to the line. Also voltages between systems.
The box I am building will have the PC104 available from the top for accessing the serial port and other things. I was thinking of making a small PCB that just picks up a few of the pins on the end to make the serial port connection easier and from outside the box. Add a couple of indicator LEDs also.
Please test the memory and see if any of it is working. This will need some work on the SPI pin selection and probably what you have already figured out.
Would like to see if you can read and write anything to each of the 5 AX5043s.
When we get through all of the above then all the CPU related things will have been tested.
You would need jumper cables just to feed signals on and off the AX5043 for RF test so this should probably come later. Just reading and writing to the devices would be a good milestone.
Bob
On Feb 2, 2024, at 4:52 PM, Burns Fisher (AMSAT) <[email protected]mailto:[email protected]> wrote:
I2c works. Pacsat>get i2c I2c device state: (1 is ok) PacSat Board Temp: 1 RTC: 1 Pacsat>get rtc Date: 2000-1-1 06:18:45 UTC Time: 946707525 vs IHU Unix time 1000000132 Pacsat>get temp Pacsat board temp is 92 ( 26.0 degrees C)
Comments for the future:
1) The temp sensors should be at different addresses. We have no ADS7828 I2c A/D converters, but they are very useful and space-proven gadgets for reading voltages, etc. However, the "easy" temp addresses clash with the 7828. It's hard to see becasue Max uses 8 bit values for addresses (with bit 0 set to 0) while the rest of the world uses 7 bit values. So Max's address 0x90 is really 0x48 to everyone else.
2) The one-wire secure memory is totally unused and I never managed to make it work on the RTIHU. I think it needs a pullup (I don't recall why the builtin pullup does not work). But regardless, I suppose we could put an encryption key in it, but it is not especially useful when it is flying in space anyway :-)
3) In the future it would be nice to have SPI and I2c busses available at test points. Or even on the PC104.
73,
Burns Fisher, WB1FJ AMSAT(R) Engineering -- Flight Software