AWR1642: AWR1642 runtime calibration and Tx output power LUT

Hello Namhyung,

The Tx output power variation with LUT gain codes is not linear, unlike Rx gain LUT. We recommend you use internally generated LUT values. Is there  a specific reason you want to overwrite the TX gain LUT values?

regards,

Vivek 

Thanks for your reply Vivek.

Yes, actually we are now trying to use the default Tx/Rx LUT, but we've seen almost 2 dB gain steps during the module operation by increasing the temperature.

As you already explained the last thread, it is caused by the Rx LUT LSB size. 

However the power variation (including signal power and noise floor, affected by Rx gain and also Tx power...) is not that much large among the module temperature changing in our test cases, so i want to set the Rx LUT without change.

But it may cause linear or nonlinear power decreasing, and i think this small slope of changing can be calibrated by changing the Tx LUT.

That is why i'm asking the variation of the Tx output power by changing the Tx gain code.

My summarized question is sorted as below.

1. Rx LUT minimum step is around 2 dB, and this is a quite large value to calibrated our module performance in our case. (sharp change happened)

2. If the Tx LUT minimum size is much smaller then the Rx LUT, i want to use modified Tx LUT for calibrated to maintain our module's performance as flat as possible.

Regards,

Namhyung

Hello Namhyung,

If you are using the max Tx o/p (0 dB backoff) then you would not be able to compensate for the power reduction of Rx gain at higher temperature since the Tx gain already maxed out. If you are OK with a gradual change in Rx gain across temperature then you can disable the run time calibrations and the Rx gain setting will not change with change in temperature.

Typically change in Rx gain is not an issue at system level since it does not impact the SNR significantly. Is there a reason why you need the Rx gain to be constant?

regards,

Vivek 

Hello Vivek,

I understand your explanation. Actually the reason why i'm trying the Rx gain to be constant is to make all of the parameters including SNR, noise floor and the signal power to be constant.

As you said, yes, the SNR is not much affected by calibrating(changing) the Rx gain, but as you know, the noise floor and the signal power affected.

I'll think some more about that.

Can i ask some more details while you are in this thread?

When i control my module with mmWave studio with maximum power setup (setting the backoff as 0), the output power seems to be changed by changing the Tx LUT values.

(for example, when we get the default Tx LUT, the value was 52 at our module's operation temperature. in this case, if we change the value from 52 to 81, the power seems to be increased almost 1 dB. on the other hand, if we set the value from 52 to 44, then the power seems to be decreased almost 2 dB)

If we set the output power as a maximum value, the LUT should not affect the output power, but in our measurement, it seems to be changed.

What do you think this kind of thing happens?

In addition, by setting the backoff value not to be 0 (make some attenuation), the output power seems not to be changed.

I think it would be some bug of the mmWave studio.

I'm using 02.01.01.00 version.

Could you check about this too?

Regard,

Namhyung