- pacsat-dev - mailman.amsat.org

My Action Items
by Bob Stricklin 15 Dec '22

15 Dec '22

All, Attached is a Excel/PDF outlining the current needs of the Pacsat system. The is based on the previous Fox design concept with some adjustments to match our needs. This is a first pass and the best information will come from testing since this is a dynamic system. In the sheet I indicate the possible modes the TMS570 may operate in. I have not looked at the details of the two principal modes. The mode that requires the most current is the mode with Built In Self Test (BIST). After launch testing may not be useful to us as we will limited ability to correct for problems. The modes can be changed by changing register values. I can see a lot of development time may be needed to understand these modes and test the results of making any changes between modes. I would like to hear some comments regarding the earlier development for this device. We will want to know the impact of making any changes before firming up launch code packages. In the PGE package we can operate the TMS570 up to 160 MHz in the pipeline mode. We can also reduce the clock rate and obtain a savings of 0.85 mA for every 1 MHz reduction in clock rate. Most of the current consumed by this part is provided by the 1.2 Volt supply line. If we drop the clocking rate to 100 MHz we can realize a savings of 51 mA or about 61 mW. This clock rate reduction will be limited by the ability to collect and process the data from the radio modems and memory devices. Testing with software in place will probably be needed to get an idea of what is required. I am attaching a set of drawings for the first pass at a design for the AX5043 system. This includes 4 receivers and 1 transmitter. The transmitter is the same as the design used in the Fox program. The receivers are changed to to use a single antenna, filter with a bandwidth to cover the 2m operating band, preamp, and power divider. More work is needed on all this as well but I wanted to show some progress. There may be enough flexibility in the AX5043 chips channel selection to allow us to program four different receive frequencies while using a single reference oscillator. We may also need a clock reference generating system and I have started laying this into the design to be sure it is covered. I have not yet included in power requirements for these devices. Bob N5BRG

1 0

Trying to write filesystem document
by Corey Minyard 15 Dec '22

15 Dec '22

I don't appear to have write access to AMSAT-NA/pacsat on github. Do you want to do this with clones and merges? -corey

2 4

Summary of Pacsat basic software
by Burns Fisher (AMSAT) 14 Dec '22

14 Dec '22

The master branch will now work with 2 mrams as I said before. It also can successfully talk to a serial line at 32k8 (and you really have to have that to do anything useful) as well as an AX5043. I have managed to get the 5043 to see a carrier on 2m (the RSSI goes way up) but not to get it to decode anything or transmit anything. I may try that more, but I'm not the 5043 expert by far! After talking with Jim McCullers, I converted all the connections to use inner pins (Booster Pack connectors) rather than the outside pins. Here are all the connections that the software is setup to use. You can also use the equivalent pins on the outside of the Launchpad (J10 and J11) but I wanted to make sure it would work with the inner pins. (J2/3 means header J2, pin 3. Pins are counted from 1 on the USB connector end) Power out (to MRAM and AX5043 board) 3.3V Output J2/1 Gnd J5/1 Console serial line: Rx-J2/3 Tx-J2/4 Gnd-J3/2 (Or any other --there are lots) SPI1 (to both MRAMs): Note that the common term for the data lines is MISO and MOSI. TI calls them SOMI and SIMO, the MRAMs just call them SO and SI. MISO: J9/7 MOSI: J9/6 CLK: J67 NCS0: J9/3 (For MRAM 0) NCS1: J9/8 (For MRAM 1) SPI3 (for AX5043--probably for all if we have more than 1) MISO: J5/7 MOSI: J5/6 CLK: J2/7 NCS1 J5/3 (Additional 5043s could use other NCS pins on SPI3. IRQ goes to GIOB[0] J8/8 (I hope I have all that right, but you can double check from the signal names against the Launchpad data sheet). I attached a picture of my setup on a previous mail, but it is hung up waiting for Bill to release it. 73, Burns Fisher, WB1FJ *AMSAT(R) Engineering -- Flight Software*

2 1

One more change to Pacsat basic software
by Burns Fisher (AMSAT) 14 Dec '22

14 Dec '22

The master branch will now work with 2 mrams as I said before. It also can successfully talk to a serial line at 32k8 (and you really have to have that to do anything useful) as well as an AX5043. I have managed to get the 5043 to see a carrier on 2m (the RSSI goes way up) but not to get it to decode anything or transmit anything. I may try that more, but I'm not the 5043 expert by far! After talking with Jim McCullers, I converted all the connections to use inner pins (Booster Pack connectors) rather than the outside pins. Here are all the connections that the software is setup to use: (J2/3 means header J2, pin 3. Pins are counted from 1 on the USB connector end) Power out (to MRAM and AX5043 board) 3.3V Output J2/1 Gnd J5/1 Console serial line: Rx-J2/3 Tx-J2/4 Gnd-J3/2 (Or any other --there are lots) SPI1 (to both MRAMs): Note that the common term for the data lines is MISO and MOSI. TI calls them SOMI and SIMO, the MRAMs just call them SO and SI. MISO: J9/7 MOSI: J9/6 CLK: J67 NCS0: J9/3 (For MRAM 0) NCS1: J9/8 (For MRAM 1) SPI3 (for AX5043--probably for all if we have more than 1) MISO: J5/7 MOSI: J5/6 CLK: J2/7 NCS1 J5/3 (Additional 5043s could use other NCS pins on SPI3. IRQ goes to GIOB[0] J8/8 (I hope I have all that right, but you can double check from the signal names against the Launchpad data sheet). Attached is a picture of the whole thing in my shack. (Ha ha--what a mess!) 73, Burns Fisher, WB1FJ *AMSAT(R) Engineering -- Flight Software*

1 0

Power fail-safe embedded file system for FreeRTOS
by Rich Gopstein 12 Dec '22

12 Dec '22

I found this while poking around. They have both GPL and commercial options. https://www.freertos.org/FreeRTOS-Plus/Fail_Safe_File_System/Reliance_Edge_…

3 6

Finally got TWO MRAMs working together
by Burns Fisher (AMSAT) 12 Dec '22

12 Dec '22

Below is the console output from the current software. "test mram" writes the address into every 4-byte MRAM cell from 0 to 1MB and then reads it back to check. Believe me, I've had plenty of cases where it does not match and it printed it out. (It was essentially always my error!) Pacsat>get mram size Size of MRAM0 is 512KB Size of MRAM1 is 512KB Pacsat> Pacsat> Pacsat> Pacsat>test mram 0KB written 64KB written 128KB written 192KB written 256KB written 320KB written 384KB written 448KB written 512KB written 576KB written 640KB written 704KB written 768KB written 832KB written 896KB written 960KB written 0KB read 64KB read 128KB read 192KB read 256KB read 320KB read 384KB read 448KB read 512KB read 576KB read 640KB read 704KB read 768KB read 832KB read 896KB read 960KB read 73, Burns Fisher, WB1FJ *AMSAT(R) Engineering -- Flight Software*

2 6

Thanks to the group.
by Jim McCullers 09 Dec '22

09 Dec '22

The config worked and now I can run Jonathan's Raspberry Pi code. Now on to 9600 AX.25 on 2-meters. Time to go read both the AX5043 and Raspberry Pi manuals! Jim McCullers WA4CWI

1 0

Pacsat dev meeting #6
by Bill 08 Dec '22

08 Dec '22

Pacsat dev meeting #6 Thu, Dec 8, 2022 7:00 PM - 8:00 PM (CST) Please join my meeting from your computer, tablet or smartphone. https://meet.goto.com/285772213 You can also dial in using your phone. (For supported devices, tap a one-touch number below to join instantly.) United States: +1 (571) 317-3112 - One-touch: tel:+15713173112,,285772213# Access Code: 285-772-213 Get the app now and be ready when your first meeting starts: https://meet.goto.com/install Agenda: Open

1 0

Summary of several bits of work
by Burns Fisher (AMSAT) 06 Dec '22

06 Dec '22

Rich, Chris, and I have had a number of private emails, so this is a quick summary: 1) We have all been able to "talk to" the Pi 5043 board over SPI from the LaunchPad, although one of the boards had a short that required fixing. 2) Chris has managed to get it to transmit. 3) Rich and Burns (mainly Rich) determined that L8 on Pi 5043 board is actually a 0-ohm resistor. That's ok for 70cm. But for 2m, an inductor is needed. Rich has made a hand wound air core coil that seems to work. Burns tried one that did not work. I probably screwed up soldering the big coil to the tiny pad. In fact I probably lifted the foil. I recommend if we are going to continue with the Pi board we get the actual SM 0603 or 0402 or whatever inductors. 4) Eric made up 3 MRAM-on-DIP adapter boards for me to check out, but we think Chris has successfully done that too. I will write the s/w to test more thoroughly and to allow multiple MRAMs to work and to sleep when not in use. 5) The I2c temperature chip on the Pi board works fine with the Launchpad. In summary, I think we are convinced that the chips are all usable with the Launchpad. I'd recommend going ahead with a board that contains a couple of MRAMs and a couple of 5043s. Jumpers are REAL easy to make mistakes with, and jumpers to 4 different boards (2 mrams, 2 5043s) would be a real bear to use. I'm most likely not going to be available on Thursday unless my monthly 6:30 in-person meeting gets out early! 73, Burns Fisher, WB1FJ *AMSAT(R) Engineering -- Flight Software*

1 0

Updated pacsat sw
by Burns Fisher (AMSAT) 05 Dec '22

05 Dec '22

Hi All, I've just merged in a branch (after testing by Chris) which has a HALCoGen configuration that is specifically for the LaunchPad. In addition, I've moved the console to be on a real UART rather than the N2HET emulated one (which has some problems with missing characters). The console serial connections are now: Ground: pin 32 Rx: Pin 33 Tx Pin 34 Chris and I are both able to talk to the 5043 chip on the Pi board, but at some point during initialization, the crystal stops oscillating. As soon as I get a couple MRAMs to use (maybe today or tomrorow?) I'll try to get them set up both to sleep when not in use and to make the fact that there are multiple MRAMs transparent--just specify an address and it will choose the right one. 73, Burns Fisher, WB1FJ *AMSAT(R) Engineering -- Flight Software*

2 3