TRF370417: TRF370417 Unwanted emission and amplifier intermodulation

The adjacent channel leakage power is required to be 55dB, but the output of the TRF370417 is 62dB, but intermodulation occurs in the amplifier, resulting in a leakage power of 42dB.

Kang-san,

The DAC is integrated with the modem and cannot be changed.
Is there any way to correct this between the DAC and TRF370417 or at the output of TRF370417?
Is it possible to increase the CN ratio over the current one by making a correction?

Hi Okada-san,

The only other way is to add additional RF band pass filter to filter out the image and the LO offset. However, you will need to ensure you have sufficient offset of the signal frequency away from your LO carrier.

-Kang

Kang-san,

If we adjust the input, how far can we lower the unwanted emission of TRF370417?
When using a class C amplifier in an FSK system, is there any way to make the system immune to unwanted emission?

Takanori,

I'm facing similar issues in my design. See if you can adjust the phase of your IQ signals. I did some experiments in my post

e2e.ti.com/.../trf370417-if-harmonics-issue

Mr.Daniel,

The phase of the IQ signal is varied by an RC circuit to measure the output spectrum and input signal.

When the phase difference is 90 degrees

When the phase difference is 94 degrees

When the phase difference is 99 degrees

When the phase difference is 109 degrees

Changing the phase difference changes Sideband Emission, but not Carrer Leak.

Hi,

carrier leakage is adjusted through DC offset. Phase adjustments will not help with carrier leakage.

As Kang already mentioned, you need to adjust DC offset for carriage leakage. Phase helps with sidebands. My post has some information & some simple circuitry you can try

Dear All,
I have concluded that changing the DC offset will not lower the carriage leakage than it is at current levels.
If the DC offset is changed during operation, the Carrer Level immediately drops, but gradually increases.

DC offset was best at 1.74V among those varied from 0.5 to 3V.

Kang-san,
Since FSK varies on both sides of the carrier (+945Hz to -945Hz), the bandwidth of the RF filter must include the carrier.
The GSM in the Applications for TRF370417 is GMSK and similar to this modulation, but do the amplification know how it is done?

Takanori OKada said:

I have concluded that changing the DC offset will not lower the carriage leakage than it is at current levels.
If the DC offset is changed during operation, the Carrer Level immediately drops, but gradually increases.

We expect the DC offset to change over temperature. Please ensure there is a dynamic loop to change the DC offset over temperature and time. We expect DC offset adjustments in the uV ranges. 

Takanori OKada said:

DC offset was best at 1.74V among those varied from 0.5 to 3V.

The range below is too large. You will need to adjust in the uV ranges. 

Takanori-san,

Our basestation customers will use high precision DACs such as our DAC3482 and DAC3283 to adjust both QMC and LO leakage. You have to figure out some way to adjust the DC offset otherwise it will be part of the RF emission. 

Kang-san,
I also have a question about Carrier Feedthrough and Sideband Suppression.
In the TRF370x Quadrature Modulator Evaluation Module User's Guide (www.ti.com/.../slwu062.pdf), Sideband Suppression is 60dBc and Carrier Feedthrough is -50dBm when optimized.
Are these values the limits?

Hi Takanori-san,

This will depend on the resolution of your carrier and LO adjustment. I have measured 70dBc to 80dBc range, and will need to correct based on temeprature.

-Kang

Kang-san,

Is there a reference design or application note for the DC offset correction method?

Kang-san,

Thanks for the reply.

I think this user guide is how to manually adjust Carrier Suppression and Sideband Rejection.

Kang Hsia said:

Please ensure there is a dynamic loop to change the DC offset over temperature and time.

Do you have any references regarding Dynamic loop?