Understanding Pulsed signal Demodulation and decimation in GNU Radio

Hello everyone

My name is Isaac. I'm dealing with a curious behaviour that I can't understand...

I'm generating my pulsed signal by multiplying a vector source and a signal source. Then I'm multiplying this pulsed signal with a continuous signal source at the same frequency, passing then for a low pass filter and seeing it in a time sink block.

The result should be a set of rectangles with the same amplitude and period. The curious thing here is that the amplitude varies in time (from 0 to max amplitude and vice versa). Any idea why the amplitude varies in time?

I attach the flowgraph just in case.

Any idea or help will be appreciated.

Thanks

Isaac t.