Expected oversampling behavior xA4

[cnlohr]

I was hoping to do 2-antenna full 2.4 ISM capture, so I bought a xA4 after seeing this article https://www.nuand.com/2023-02-release-1 ... bandwidth/

I haven't dug into the data too much but wanted to make sure I was on the right track.

Everything seems to work at 80MHz, but if I enable META at 100MHz, using BLADERF_FORMAT_SC8_Q7_META, I get
`Overrun detected. 4088 valid samples were read.`

If I use BLADERF_FORMAT_SC8_Q7, then I see a strange behavior, I get 185MSPS (should be 2X the bandwidth). But I get no errors.

I've tried a variety of buffer sizes, but nothing seems to keep up.

While it looks like I can sort of kind of get 90MHz with MIMO, what is the expected behavior when doing MIMO and oversampling?

[cnlohr]

Ok, I think I understand. The goal is to include the _META, and if it detects a drop it will notify me. If I don't include the _META, I will be unable to discover when there is a drop?

[CoreyLivingston]

This is expected at high rates. At 100 MHz MIMO, you are hitting USB and host limits, especially with META enabled, which adds overhead and causes overruns. The ~185 MSPS you see without META is due to 2× oversampling, but errors are not reported in that mode. In practice, stable MIMO capture is usually limited to ~80–90 MHz depending on the system.

[cnlohr]

I think I'll be fine with 88MHz of bandwidth. I worry a little bit because the thing looks like it gets kinda hairy at the extremes. But, this may be sufficient.

Thanks for the help and explanation!

Mods can feel free to close this issue.