Two devices in one instance of app

[vanChesco]

Is there any chance to get two traces from two separate devices on one graph? I often use two TinySA scanners at one time. Using two separate instances of app on one screen, totally killls visibility due to size of them.

I've also found one bug: when you're trying to add new line in Bands and Markers and by accident press Enter after writing only a name it hangs up.

It also be nice to have a chance to draw a horizontal line (for reference) at chosen value of dB.

[g4ixt]

It might be possible to control two devices concurrently but it might need a lot of code changes. I will think about it.
It is already possible to choose a device, in 'preferences', but as I don't have two tinySA I can't test if 'swapping' works or if it would hang something - you could try this and let me know. Try scanning with one device, set a trace to 'freeze', stop scanning and then swap to the other device and run with a different trace.

Adding bands and markers uses PyQt Model / View standard features and is hard to get right for a Python beginner (me). I will look to see if I can fix it or error trap it but I might not be able to.

A horizontal line is easy but I don't have much screen space left to add controls. I will add one with a dBm label, and allow it to be dragged into the desired place.

[vanChesco]

As for beginner (as you claim) you've done a pretty nice work. I love your app much more, than the original one.

I'll do the test you mentioned and will let you know.

Much appreciate your efforts.

As for horizontal line: maybe double click on it to change the reference value would be a good compromise? What you think?

[vanChesco]

Try scanning with one device, set a trace to 'freeze', stop scanning and then swap to the other device and run with a different trace.

Just tested it. It works. The first trace stays frozen and the second one is running after changing device to the other one.

The only thing is, that swapping from one device to the other takes some time and once a time hangs the app.

That's what I'm getting in command line window, when it happens:

Database QtTSAprefs.db found at C:\Users\dariu\AppData\Local\QtTinySA
QtTSAprefs.db open: True Connection = "settings"
database version = 110, expected 110
QtTinySA v1.1.0
Serial port COM3 open: True
COM3 test 0 : tinySA4_v1.4-199
measurement thread running
Traceback (most recent call last):
File "pyqtgraph\graphicsItems\ImageItem.py", line 616, in paint
File "pyqtgraph\graphicsItems\ImageItem.py", line 590, in render
File "pyqtgraph\functions_qimage.py", line 289, in try_make_qimage
IndexError: index 25 is out of bounds for axis 0 with size 25
Traceback (most recent call last):
File "pyqtgraph\graphicsItems\ImageItem.py", line 616, in paint
File "pyqtgraph\graphicsItems\ImageItem.py", line 590, in render
File "pyqtgraph\functions_qimage.py", line 289, in try_make_qimage
IndexError: index 25 is out of bounds for axis 0 with size 25
Serial port COM4 open: True
COM4 test 0 : tinySA4_v1.4-199
Traceback (most recent call last):
File "QtTinySA.py", line 1178, in run
File "QtTinySA.py", line 470, in measurement
File "QtTinySA.py", line 296, in serialWrite
File "serial\serialwin32.py", line 325, in write
serial.serialutil.SerialTimeoutException: Write timeout
Serial port COM3 open: True
COM3 test 0 : tinySA4_v1.4-199

[g4ixt]

When you had that error, had you changed one of the scan settings like number of points, frequency, RBW etc?

[vanChesco]

When you had that error, had you changed one of the scan settings like number of points, frequency, RBW etc?

Nope. Just swapped ports for TinySAs.

And I think that instead of swapping source devices for every sweep (if that was the idea), it would be better to just open two ports and read the data simultaneously. Less comm problems and much quicker response.

[g4ixt]

Yes it certainly would. That was just a test to see if it actually worked at all.

[tweeter103]

May I add my support to the concept of seeing the traces from multiple TinySA on the same screen. Maybe four?
Could the existing 4 traces be one from each device?
If not then maybe if four instances of QtTinySA could coexist on a quad-view of a monitor.
Happy to be involved in scoping this.
But of course I respect that you have created what is excellent, and it is your time, and effort, and you may have other priorities in life, than humoring all us who are clamoring for more.

[g4ixt]

It's an interesting challenge. The way it connects to the tinySA and measures ought to make it possible to concurrently use more than one but plotting the results on a single graph is probably going to be tricky, especially if they are scanning different frequency ranges. The database that stores all the settings is another challenge, although I originally designed it to be able to store multiple (user choice) settings but never implemented it.

[tweeter103]

Certainly from my usage perspective, they are all set to the same parameters, frequency start/stop, RBW, points, etc.
And likewise zooming the graph should affect all 4.
The purpose it to look at the same signal from 4 different locations.
Selecting "Run" should start all 4 going, but there is no need for synchronous measurements.
If I wanted to look at different information, I would prefer to open multiple instances of QtTinySA each talking to a TinySA.
Thanks for your interest.

[vanChesco]

of course I respect that you have created what is excellent, and it is your time, and effort, and you may have other priorities in life, than humoring all us who are clamoring for more.

Just wanted to double what's been said here. I'm really grateful for your work as well. The app is brilliant already.

[g4ixt]

I have started splitting the code into more separate modules with the aim of being able to show scans from different devices concurrently. I will probably need to switch from 'threading' to 'multiprocessing' as well and I am hoping that this will also improve responsiveness on low performance devices like the Pi3B.

[vanChesco]

The more I think about that feature, the more I'm convinced that it will be useful in many ways.

If there would be a way to set different ranges to each device, we could scan much wider spectrum with much better accuracy.

Let's say we want to scan a range between 500-700MHz, then we could set first device to a range 500-600 and the second one to 600-700 working simultaneously. Both with 1000 points resolution lets say. The outcome would be (approximately) 2x faster, than scanning the whole range with resolution of 2000 points with only one device.

[g4ixt]

If there would be a way to set different ranges to each device, we could scan much wider spectrum with much better accuracy.

Yes that thought had occurred to me and I plan to code it so that traces can be assigned to each device and each device has its own scan settings. Not sure how I am going to find room in the GUI to allow it but I will think of a way.

I am also intending to see if I can obtain and plot scan data from other devices, e.g. I have a LimeSDR and I think I can use the 'soapy' interface for that. A friend has a Siglent SA and I have already tested some very basic code to obtain scans from it using 'pyvisa'.

[vanChesco]

I was just thinking that it would be a good idea to have three modes:

1. Two separate tracks for each device working in the same range (for monitoring diversity RF system).
2. Multirange working exactly as it already works in TinySA (Range-Multi)
3. One common track for both of them and if we'd going to scan a given range, the app would divide it in half and the first half would be scanned by first tinysa and the second one for the other half.

I thought about the third option, cause that's exactly how pro wireless receivers from WisyCom does their job. The more receivers you are using, the fastest scan you can do, cause the WisyCom app is dividing the range for as many parts, as the count of receivers is.

As for GUI - I think there's no need to split screen or anything like that, since we could use it as it is and just destine one track to the first device and other one for the second one.