BOOSTXL-CC1120-90: Fails to shut down radio when lauchpad controller enters sleep mode. Drains batteries.

Can you elaborate? The MISO line? What should I set it to? Low? 

The TCXO is not the problem. The radio itself will not turn off. I'm pretty much at a loss at this point. The TI hardware is kind of crappy, with the errata calibration and now this issue. We have tried everything we can think of but cannot get it to enter the lpm sleep. If we never transmit a message then power stays low indefinitely. As soon as we transmit, we cannot revert to low power.

So how do i check the chip version? I do not see any mapping to a PARTVERSION register in any of the code supplied by ti. I see a ton of different code out there for people attempting to use the CC1120, but where would one find the official latest version of a C library?

We have a crude measuring device, but are seeing .03 to .05 amp after transmit, which is pretty much the same as during.

I have code to check the marc state, but no list of marc state possibilities. From another persons post I was delaying until marcstate was = 0x41, but that did not change anything.

I have tried everything i could think of in the transmission code, including turning off all io, strobing idle and looping to confirm, doing resets, etc before strobing SPWD. No change. I have tried playing with the value of the miso pin as well. Heres some code that is in there at the moment. Id be happy to send you a zip of the project. 

             trxSpiCmdStrobe(CC112X_SIDLE);
             while((trxSpiCmdStrobe(CC112X_SNOP)& 0xF0) != CC112X_STATE_IDLE);


             uint8 writeByte = 0x80;
               cc112xSpiWriteReg(CC112X_IOCFG3, &writeByte, 1);
               writeByte = 0x80;
               cc112xSpiWriteReg(CC112X_IOCFG2, &writeByte, 1);
               writeByte = 0x80;
               cc112xSpiWriteReg(CC112X_IOCFG1, &writeByte, 1);
               writeByte = 0x80;
               cc112xSpiWriteReg(CC112X_IOCFG0, &writeByte, 1);

               trxSpiCmdStrobe(CC112X_SXOFF);

               while((trxSpiCmdStrobe(CC112X_SNOP)& 0xF0) != CC112X_STATE_IDLE);

                   P3OUT &= ~(1 << BIT1);



             trxSpiCmdStrobe(CC112X_SPWD);

Ter, thanks. I had done a global search for PARTVERSION but somehow missed it. Ive been in and out of any code I can find online with SPWD. I would love some kind of git repo with an official library if that exists, but for now will continue with the revE example.

I see the marc states in that document, but do not see them mapped to constants in the sw example, but will double check the marcstate before attempting to sleep. 

We are not doing anything to a CC1190. I assume that is a chip on the CC1120-90 boosterpack. We are only trying to get that boosterpack board to consume low power. For now, the goal would be to returning to consume the same low power as it does before the first transmission (doesnt register on our 0.00 meter). If we can get to that point we can try to turn off more. We cant get under .03amps. If there is something I need to do other than attempting to strobe SPWD then please provide any example code. I am shocked that these are marketed as ultra low power but there is not a single use of SPWD in the example software. 

We have the radio going to sleep finally with the SPWD strobe. What would we do additionally in software to address the cc1190 chip on the boosterpack? We are still drawing 8ma at rest on the entire kit (micro + modem). We have stolen code to shut off all gpio, usb, xt2. but i cant find anything about shutting off 1190 separately. Thanks for your help.

Edit...found this code?:

#ifdef CC1190_PA_LNA
    // initialize the IO
    RF_PA_EN_PxDIR |= RF_PA_EN_PIN;
    RF_LNA_EN_PxDIR |= RF_LNA_EN_PIN;

    // configure idle
    RF_PA_EN_PxOUT &= ~RF_PA_EN_PIN;
    RF_LNA_EN_PxOUT &= ~RF_LNA_EN_PIN;
#endif

TER, So I have made a lot of progress. I wanted to ask you if I am at as good as I am going to get. 

CC1190:
Do I really need to shut this down separately when using the cc1120-90 boosterpack? Setting the pins to low doesnt make any difference in our power consumption. I would think that when you set the radio board to SPWD that it would shut everything down.

Power draw:
The draw measured at the 3.3v isolation bridge is <1ma at rest. The draw measured from batteries is 15.65ma at rest. (powering externally via the 5v header). We need the dcdc converter that is on the debugger side of the isolation block to get to the 3.3v, because we have 5v sensors and only want to use the 5v battery. Surely the dcdc converter isnt using 15ma? What else could be causing that draw? I have removed all other isolation jumpers other than 5 and 3v.

For anyone else having power issues, check out this video and workshop pdf: training.ti.com/msp430-workshop-series-7-12-low-power-optimization

Thanks, Tony 

Thanks, I figured that out last night. So I think I am good to go, but am open to any recommendations on the most efficient way to get both 3.3 and 5v out of a pack to power this thing. The radio and sensors need 5.

I found like four versions of a HAL from TI, with dates all over the place. I wish there was an official git repo for these boards. Having to scour multiple pdfs and zips is a real pain when you are not familiar with the chipset families at all.