Running TX and RX simultaneously from CLI

[krlawrence26]

I've been encountering some strange behavior when trying to simultaneously input a signal into TX and measure that signal from RX (with a cable connecting the tx port to the rx port) using the CLI. I'm running on Windows 7, and while CLI won't let me tx and rx simultaneously from a single command prompt, if I open two command prompt windows and use one to start/stop tx and the other to start/stop rx, then CLI doesn't complain. But the data I get from rx looks very unlike what I send into tx. For example, when I send a constant value (I=1024, Q=0) into tx, I see the following: plots

buffer_10_constant.png

Top left is I vs t, bottom left is Q vs t, and right is I vs Q.
Here's a zoomed-in view of one of the large-magnitude portions:

buffer_10_constant1.png

Next, I tried sending in a sine wave. When I send a sine wave into rx from a signal generator, I see a nice clean sine wave. But when I send it in from tx, I get:

buffer_9_sin.png

which looks similar to the case for constant input.

I checked my rx and tx gain settings, and I believe that I should be able to see some amount of the tx signal. Has anyone else attempted using tx and rx simultaneously from CLI? I'm not sure what could be causing this choppy behavior, since the conditions are the same over the whole data-taking time (eg no stopping/starting or changing settings). Does sending and receiving at the same time cause problems with PC loading or with USB? My eventual goal is to use the bladeRF as a network analyzer; has anyone else had success using it for that purpose?

[bpadalino]

It's good to hear you can get samples working on Windows 7! That's good news. If you don't mind me asking, with xHC is in your computer?

When we moved to libusb for the interface, we never brought over the async interface to the CLI. We are getting back to bringing the streaming interface over, instead of using the old synchronous style.

What samplerate were you setting? The loading shouldn't be an issue. If you lower the samplerate, you should see a better picture.

We'll work through the rest of this week and the weekend to get the CLI onto the streaming asynchronous interface. Sorry for the inconvenience.

[krlawrence26]

Thanks for your response!
I'm not sure what you mean by xHC--how can I find out which I'm using? My samplerate was at 1,000,000; I've since tried again with 500,000 (the lowest it seems to allow) but still encounter the same problem. Though, the computer I'm using isn't the newest so it may be a PC issue.
The lack of an asynchronous streaming interface for CLI does seem like it might be a plausible explanation. Do you have any idea of when the new version might be available?

[syntroniks]

Perhaps bpadalino means your host controller -- the device providing your usb interface on the PC side. For instance, my Rosewill USB 3.0 card reports as a Renesas Electronics USB 3.0 Controller with VID 1033 and device id 0194.

I checked it out using window's built in 'device manager'.