CC1125 in high power application

Please refer to the attached app note. Figure 4.17 on page 14 shows the modulation bandwidth measured as defined in En 300 220 V2.3.1 @200 kbps and +25 dBm output power.

The first question is always "What are your register settings"?  If you could pass those on, there might be some tweaks there that can help a bit.  Does the noise look broadband or is there a spur in there?  What crystal/TCXO frequency are you using. 

Regards,

Chris

Hi, I am a  little confused by the plot there. The fundamental power only looks like +10dBm rather than +25dBm.

My measurments are on the CC1125 eval board only, and give little scope for gain.

Regards.

I am using the CC1125 eval board as purchased from your distributors as the TI CC1125 Cat 1 development kit.  All the components are as provided in that kit.

All of my measurments have been done using the TI Smart RF Studio 7 linked to the eval board, so the register settings are those provided by that application.

The noise is broad-band, there are no real peaks in it.

I have attached the data from Smart Studio below.

Thanks

/***************************************************************  *  SmartRF Studio(tm) Export  *  *  Radio register settings specifed with C-code  *  compatible #define statements.  *  *  RF device: CC1125  *  ***************************************************************/

#ifndef SMARTRF_CC1125_H #define SMARTRF_CC1125_H

#define SMARTRF_RADIO_CC1125 #define SMARTRF_SETTING_IOCFG3           0xB0 #define SMARTRF_SETTING_IOCFG2           0x08 #define SMARTRF_SETTING_IOCFG1           0xB0 #define SMARTRF_SETTING_IOCFG0           0x09 #define SMARTRF_SETTING_SYNC_CFG1        0x0B #define SMARTRF_SETTING_DEVIATION_M      0xA3 #define SMARTRF_SETTING_MODCFG_DEV_E     0x0C #define SMARTRF_SETTING_DCFILT_CFG       0x1C #define SMARTRF_SETTING_PREAMBLE_CFG1    0x00 #define SMARTRF_SETTING_FREQ_IF_CFG      0x33 #define SMARTRF_SETTING_IQIC             0x46 #define SMARTRF_SETTING_CHAN_BW          0x04 #define SMARTRF_SETTING_MDMCFG0          0x05 #define SMARTRF_SETTING_DRATE2           0x89 #define SMARTRF_SETTING_DRATE1           0x99 #define SMARTRF_SETTING_DRATE0           0x9A #define SMARTRF_SETTING_AGC_REF          0x20 #define SMARTRF_SETTING_AGC_CS_THR       0x19 #define SMARTRF_SETTING_AGC_CFG1         0xA9 #define SMARTRF_SETTING_AGC_CFG0         0xCF #define SMARTRF_SETTING_FIFO_CFG         0x00 #define SMARTRF_SETTING_SETTLING_CFG     0x03 #define SMARTRF_SETTING_FS_CFG           0x12 #define SMARTRF_SETTING_PKT_CFG2         0x06 #define SMARTRF_SETTING_PKT_CFG1         0x00 #define SMARTRF_SETTING_PKT_CFG0         0x40 #define SMARTRF_SETTING_PA_CFG2          0x6D #define SMARTRF_SETTING_IF_MIX_CFG       0x00 #define SMARTRF_SETTING_FREQOFF_CFG      0x22 #define SMARTRF_SETTING_FREQ2            0x56 #define SMARTRF_SETTING_FREQ1            0xF2 #define SMARTRF_SETTING_FREQ0            0x8F #define SMARTRF_SETTING_IF_ADC0          0x05 #define SMARTRF_SETTING_FS_DIG1          0x00 #define SMARTRF_SETTING_FS_DIG0          0x5F #define SMARTRF_SETTING_FS_CAL0          0x0E #define SMARTRF_SETTING_FS_CHP           0x30 #define SMARTRF_SETTING_FS_DIVTWO        0x03 #define SMARTRF_SETTING_FS_DSM0          0x33 #define SMARTRF_SETTING_FS_DVC0          0x17 #define SMARTRF_SETTING_FS_PFD           0x50 #define SMARTRF_SETTING_FS_PRE           0x6E #define SMARTRF_SETTING_FS_REG_DIV_CML   0x14 #define SMARTRF_SETTING_FS_SPARE         0xAC #define SMARTRF_SETTING_FS_VCO4          0x1F #define SMARTRF_SETTING_FS_VCO2          0x4A #define SMARTRF_SETTING_FS_VCO1          0xAC #define SMARTRF_SETTING_XOSC5            0x0E #define SMARTRF_SETTING_XOSC3            0xC7 #define SMARTRF_SETTING_XOSC1            0x07 #define SMARTRF_SETTING_XOSC0            0x01 #define SMARTRF_SETTING_LQI_VAL          0x80 #define SMARTRF_SETTING_PARTNUMBER       0x58 #define SMARTRF_SETTING_PARTVERSION      0x21 #define SMARTRF_SETTING_RX_STATUS        0x10 #define SMARTRF_SETTING_XOSC_TEST1       0x00

#endif

Is this evaluaiton board available for purchase?  I can see reference to the design on the CC1120 page but no link to the part number for the evaluaiton board.

The plot at the carrier frequency is with RBW = 1 kHz. To measure output power you need a much higher RBW (this is a 200 kbps modulated signal).

The output power is +25 dBm in the measurements.  

I am surprised that TI published the app note as is because the transmitter in this app note actually fails the EN 300-220-1 regulations.  

The bandwidth specified in the mask in figure 4.17 is bigger then allowed in the EN regulations. You are only allowed to transmit at the levels given in the figure in the band  869.4 - 869.650 Mhz. The bands on either side are reserved for alarm use. You are not allowed to interfere with them. This means the maximum allowed bandwidth for the transmitter is 250Khz.. If we specify this in the mask rather then the 500-600 kHz given,  the spectrum exceeds the -30dBm required for the EN regulations at the edge of the regulatory band. .  

Hi,

Thanks for the clarification.  At 200KBp/s you do fail the band-edge, the allowable band is only 869.4 to 869.65MHz and not the wider mask that the apps note shows.

Are there any plots with a lower data rate (say 38.4KBits or similar) with the correct band mask indicated?

Are there any techniques that could be used to improve the phase noise performance as I am not comfortable with trying to approve a solution with only 1 or 2dB of margin - itys just not practicable.

Please get back to me.

Hi,

Thanks for the new plots.  I do not get that much margin without the PA (just the Cc1125 eval board).  it is possible to get the CC1120 + Cc1190 eval board so that I can make some independent measurements?

regards

Andrew was absolutely correct in his posting and thanks for pointing out the error. The sub band, where +27 dBm is allowed is 250 kHz wide.

The mask in the plot is for the g1 sub band, which is 600 kHz wide and is not relevant for operation in the 8694 - 869.65 MHz band. Attached is the correct measurements for CC1120+CC1190 @50 kbps and +25 and +27 dBm output power. We will correct the app note!

CC1125 uses different synthesizer settings compared to CC1120 to improve the phase noise so the margin to regulatory limits will be higher by approx 2 dB when using CC1125+CC1190. 

Also, setting FS_DIG0 to 0x55 reduces the loop filter BW to 131 kHz from the maximum of 170 kHz. Lowering the BW has the effect of reducing far-out phase noise but increases close-in phase noise. If you operate at 50 kbps you should try changing FS_DIG0 to 0x55. 

HI,

Thanks for the information.  I will give these suggestions a try.

Could someone please get back to me to let me know if the CC1120 + CC1190 Evaluaiton board that is referred to in these applications is available to purchase as I really would like to confirm these results before comitting to a PCB design and production.

Ian,

I sent you a "friend request" in E2E. You might want to respond to this to get your question regarding EB sorted out.

HI, I have accepted the friends request. What do I do now?

Thanks

Hi,

are there any mask measurements for narrow band (i.e. 4.8 kbps 2GFSK) for cc1120 and cc1190 @ 27 dBm? Using 12.5 kHz channels, one can have many channels inside the 500 mW band instead of one wide band channel.

In sub-band 869.4 - 869.65 MHz you can have a channelized narrowband system or 1 wideband channel.. For narrowband systems make sure you meet the ACP requirements

One can have two types of narrowband channelized systems (12.5kHz & 25 kHz) - each with different ACP requirements.

My simple question is: Will the cc1120 + cc1190 fulfill either or both ACP requirements for 27 dBm output power? Has this been tested?

I guess the ACP is mainly dictated by the phase noise...

ACP is set by phase noise and modulation (data rate, frequency deviation, modulation type; e.g GFSK vs FSK).

ACP figures in the data sheet are relative figures, whereas the ETSI requirement is an absolute number (dBm).

For 25 kHz channel spacing the requirement is -37 dBm. CC1125 @1.2 kbps, +/-4 kHz deviation, 2GFSK the ACP is 58 dBc. If the output power is +27 dBm the ACP will be 27 - 58 dBm = -31 dBm. I.e not compliant unless output power is lowered to +21 dBm.

For 12.5 kHz channel spacing the requirement is -20 dBm. Using the same system parameters as in the example above CC1125 will be compliant to ACP requirement @2+27 dBm output power.