Please help me about the pulse signal problem

Hi all,
I have a pulsed square wave need to be amplified to 200W, and after the Low Pass filter, its shape changed, not square any more, the shape becomes like below picture showed. Instead of good square shape, it becomes discontinued.
https://obrazki.elektroda.pl/7407682500_1375542572.jpg

Is there anyone know the root cause?
Please give any advice, thank you!

zltogo said:

Hi all,
I have a pulsed square wave need to be amplified to 200W, and after the Low Pass filter, its shape changed, not square any more, the shape becomes like below picture showed. Instead of good square shape, it becomes discontinued.
https://obrazki.elektroda.pl/7407682500_1375542572.jpg

Is there anyone know the root cause?
Please give any advice, thank you!

Click to expand...

I have two questions: 1. can your amplifier process the full baseband bandwidth of the input pulse train including DC component?
2. does your oscilloscope have e.g. a 50-Om input impedance? For 200 W, you should use a coaxial high-power load with 50 Ohm impedance, otherwise the LPF is bAdly mismatched and thus you can see the ringing response.

jiripolivka said:

I have two questions: 1. can your amplifier process the full baseband bandwidth of the input pulse train including DC component?
2. does your oscilloscope have e.g. a 50-Om input impedance? For 200 W, you should use a coaxial high-power load with 50 Ohm impedance, otherwise the LPF is bAdly mismatched and thus you can see the ringing response.

Click to expand...

thanks for your reply,
1,the PA is designed for 1.3-1.4ghz, for 1.3-1.36ghz, there is no problem, only 1.37-1.4ghz, the problem happens. The pulse is 100 us width,10% duty.
2,after the filter, there is a 40db 200W ATT, so it does matched to 50 ohm.

From your schematic I can see you feed a L-band RF amplifier with 100 us pulses?What do you expect to amplify? A 10 000th harmonic?
A 100 us pulse with 10% duty cycle means that the spectrum is centered around 1 KHz with harmonics possibly extending over 1 MHz. What do you expect the L-band amplifier to amplify?

jiripolivka said:

From your schematic I can see you feed a L-band RF amplifier with 100 us pulses?What do you expect to amplify? A 10 000th harmonic?
A 100 us pulse with 10% duty cycle means that the spectrum is centered around 1 KHz with harmonics possibly extending over 1 MHz. What do you expect the L-band amplifier to amplify?

Click to expand...

sorry for my mistake。the signal is modulated by a 1.38ghz cw + high speed pin swith which is controled by 100us，10% pulse。

zltogo said:

sorry for my mistake。the signal is modulated by a 1.38ghz cw + high speed pin swith which is controled by 100us，10% pulse。

Click to expand...

Thanks for clarification, the switch is not in the schematic.
Concerning the strange output waveform, it still looks to me like a mismatch. Have you tried to see the output spectrum without the LPF? The RF output spectrum should look like the spectrum before RF amplifier. Please try to check both. What is LPF purpose? I would expect to see a band-pass filter there.

- - - Updated - - -

Also, your picture shows a screen of a 200 MHz scope... is the signal at 1.38 GHz there? What about the detector used? Is its video output loaded in 50 Ohms?

you are very unclear. Is it a digital pulse of 100 ms duration, or a microwave carrier that is amplitude modulated on for 100 ms, then off? and why would u use a lowpass filter at all if u are concerned about preserving the pulse shape?

u keep saying "the signal is modulated by ghz " what does that mean?