LP-CC1312R7: Measure the current draw during operation

I will try to cover the different questions. Please let me know if  I have forgotten something.

Danilo A. said:

Our customer would like to measure the current draw during operation of the wireless LP-CC1312R7 Launchpad

Please refer to this app note: www.ti.com/.../swra478

Danilo A. said:

s it possible to transmit at more than 14 dBm using the evaluation board LP-CC1312R7 like for example changing the power settings to modify the operation of the DC DC converter or a boost mode?

See  the datasheet "7.11 861 MHz to 1054 MHz - Transmit (TX) ". Due to how the PA is powered in CC1312R7 the output power is close to constant over the full VDD range. The CC1200 has a VDD dependency of the output power.  

Danilo A. said:

I can see spurious on the unmodulated carrier at +/- 370 kHz

I have to check what exactly these spurs are. Are these considered as in issue vs any standard? 

Danilo A. said:

i can't see either in the RFstudio how the modulation should be selected.

For 4-(G)-FSK, see

https://e2e.ti.com/support/wireless-connectivity/sub-1-ghz-group/sub-1-ghz/f/sub-1-ghz-forum/1075632/cc1312r-4-fsk-modulation-support

For GFSK vs FSK, see the description in SmartRF Studio:

Danilo A. said:

conducted gmsk peformance looks really poor compared to competitor solution ..
linkIQ low
(868.03 MHz, 14 dBm, 75 kHz bandwidth, 10kbps, 2.5 kHz deviation, GMSK modulated )

The modulation from CC1312R7 should be very close to what you get it calculating an ideal modulation. Please provide the exact settings used for this measurement. 

for one of our proprietary settings we are using GMSK modulation. I can't find a setting corresponding to this.

I only see FSK (modType 0), GFSK (modType 1) and OOK  (modType 2).

for the preset wmbus mode C, the modulation settings is correct (2fsk), but the text description in Rf studio is wrong

RF studio description is wmbus c-mode 100kbps, 2GFSK, ... while this modulation is 2-fsk as it should be

For these settings the LP-CC1312R7 gives really poor performance close to carrier compared to competitor solutions (use same settings as for 99% obw measurement with ocw of 40kHz)

thomas stichelbout said:

RF studio description is wmbus c-mode 100kbps, 2GFSK

I have written a ticket on the text in Studio. The settings are correct.

thomas stichelbout said:

for one of our proprietary settings we are using GMSK modulation

We normally recommend setting modulation to GFSK and the modulation index  = 0.5 to get GMSK. Does this result in the same spectrum for your prop setting? 

I will check some around the spectrum for the 9.6 kbps setting.

As shown the modulation is far from the "ideal modulation" . A strong noise degradation is visible between +/- 14 kHz and +/- 50kHz from the carrier. The degradation is more than 20 dB 

i used an Agilent ESG 4438C to compare the modulation output from the LP-C1312R7 for the proposed modulation settings (i.e. gmsk, 10kbps, 2.5kHz freq deviation close to 868.03 MHz

The settings shown on theSWCU192 – NOVEMBER 2021, Table 25-175., p 2145 are not as detailed as the table shown above. OOK is not specified for example. 

modulation index can not be specified.

The technical reference manual is typically written and released when the chip is released to the market. Some modulation formats are not verified at this time but added after. SmartRF Studio and syscfg contains the phys we have run a full cz on. 

Not sure if I understand the comment about modulation index. Since it's possible to give both datarate and deviation this is given as a function of these two parameters. 

Hi ,

I took a short look at the application note swra478as you recommanded and from what i understand it mainly focuses on the following point : 

• use a keysight power analyzer (as we also have in our lab) but only focusing as using it as a power source at the 3.3V terminal P1 without any jumper connected as shown on figure 9 and figure 17 .. this will not work when using smartRF studio as the design is not standalone with a code running.

I just would like to check the current draw for a continuous carrier with a given modulation setting in use instead of on a running demo between 2 evaluation boards.

The power analyzer can also be used a amp meter on an existing design (for example at one jumper) but this point is not really covered by the app note.

Current measurement using smart RF studio are not covered either.

To keep the communication running i have kept the jumpers, inserted the power analyzer at M3      , programmed the evaluation board and measured the current in different operation.

I measure a current draw of 2.68 mA when the carrier is off (probably as the MCU is kept awake by RF studio).

Removing the other jumper parallel to M3 (i.e. M1, M2 , M4-M18) don't change the measured current draw. 

Removing the jumper at M13 and M14 don't change the measured current draw either as the diodes are turned off already.

I have assumed this value of 2.68 mA being a continuous offset to compare with the TI datasheet and as shown below. 

even doing so the current draw remains slightly higher than the datasheet for power above 5 dBm.

during the current measurement, the RF output is connected to a spectrum analyzer (i.e 50R) where i also measure the RF power.

I haven't looked at the Energy trace part in the app note swra478  as again i more interested in operation with smartRF studio instead of a given specific custom code.

To see if you measure current correctly (from a setup perspective) I would recommend the empty example (or standby example). If you remove the jumpers to the LEDs the empty example has close to constant current consumption. 

TX power: The current given in "7.6 Power Consumption - Radio Modes" in the datasheet is measured with the MCU off. If you use SmartRF Studio, the jtag domain is on and the MCU will be active. Hence the current consumption you are measuring with this method will be the numbers found in "7.6 Power Consumption - Radio Modes" + Icore "active" + jTAG domin on from "7.5 Power Consumption - Power Modes"

ok thanks.

My intention was to check the spectrum for the 9.6 kbps case this week but I have a cold that I don't want to expose others to. It's on my todo list.

fine. i am also juggling between different tasks on my desk where the performance evaluation of the CC1312R7 is just one of them ;)

I am looking at the schematic at the moment, to sketch a layout and i am wondering if the CC1312R7 support a serial wire debug (SWD) interface for programming and debug ? This interface is more compact than the JTAG interface. I see that the SEGGER programmer are supporting this product but there is not much detail for that part.

Please see e2e.ti.com/.../cc1352r-flash-the-microcontroller-using-swd-interface

ok . so no support for SWD. only Jtag

Spectrum: The spectrum can be changed by changing the loop bandwidth. Changing the loop bandwidth will in effect select the shape of the phase noise (basically should the phase noise be lowest close in or further out). The nominal TX settings for the 9.6 kbps case uses 20 kHz loop bandwidth. If you look in section 13.3 in https://www.ti.com/lit/an/swra682/swra682.pdf settings for other loop bandwidths are listed and you can test if some of them give a better spectrum for your use case.

I assumed that you have checked if passing the TX mask you want to be compliant to is possible?   

hi,

I am still missing to perform measurements like transient power or radiated measurement for spurious emission close to carrier. Our chamber is fully booked or the coming weeks so this will remain pending for a while.

I can just see that the CC1312R7 with the settings available from RF studio has significant degraded performance for a narrow band gmsk signal compared to other competitor radios or a keysight signal generator as described earlier.

I tried to override the settings for the loop bw from 20 kHz to 40, 60, 80,150 kHz and 200 kHz. IIR filter does not gives a significant change.

The improvements related to the loop bw are shown below.

The shape is improved close the carrier but the noise level remains 10 dB higher than the competition (i.e. above -40 dBm @ +/- 25 kHz offset. This will spoil the band for other near by devices in adjacent channels