Strange behavior with antenna injection

[fyli16]

Hello developers and the community,
I am trying to use antenna to achieve Alfven wave injection inside a plasma in 1D (e.g. see discussions in #743), but found a strange behavior. I initialized flattop current Jy, Jz to launch a left-hand CP wave. However, the wave is only successfully excited when the current length is 2 di in the wave propagation direction (x). Any small deviation in the length will lead to no waves excited. Below is two examples. The first one with the current initialized within x=[10, 12] di (di is the ion inertial length) is good; the second one shows no waves but only had the current length extended by 0.001 di. In the first good case, the excited wave is of wavelength ~ 7.8 di. The second one is just an extreme case for illustration purpose; actually I've tried many fine tunes (e.g. Gaussian profile instead of flattop current profile, much longer/shorter current length) and none has lead to wave excitation. This is super weird. I've attached my input file
inp.txt. Did I miss anything here regarding the antenna initialization? Thanks!

PS: for the space-time plots on the right, the excited waves (first case) will be reflected on both boundaries, forming standing wave patterns. This is not relevant here, though.

[mccoys]

I have no experience in this, but is the current well added in the box? Does it match your input?

[fyli16]

looks like the current does not show up in the diagnostics. Is this by design? (i.e. external currents are not included in diagnostic output) How do I visualize the current?

For the only good case, the wave is excited as expected, so there must be sth initialized.

[mccoys]

Check the currents using probes or fields diagnostics

[fyli16]

Checked through field diagnostics. there is absolutely no additional currents between antenna on vs off. Is it by design of smilei that the antenna current is not included in field diagnostics?

[mccoys]

I am confused. I thought these currents were visible but I have not checked for a long time. Have you tried probes?

[fyli16]

no antenna currents showing up in probes either.

[fyli16]

while it is still unclear why exactly, the problem is resolved by increasing the number of particle per cell.