Open Research Institute Weekly Reports, 6 May 2022

Previous Meeting Next Meeting

⏯

youtube image

►

From YouTube: OpenRTX MD-380 M17 modulation test

Description

Paul KB5MU demonstrates the M17 transmit modulation from a TYT MD-380 handheld transceiver modified with open source OpenRTX firmware. Versions from April 30 and May 5, 2022, are compared.

Both versions display an anomaly during the first few seconds of the first transmission after powerup. This anomaly is clearly seen on a Rigol RSA5065A spectrum analyzer.

The Rigol instrument also has a Vector Signal Analyzer mode, which displays a one-dimensional constellation diagram for M17's 4FSK modulation. This analysis clearly shows that the April 30 firmware version produces inadequate frequency deviation, about one-third of nominal. The May 5 firmware version corrects this problem.

The VSA analysis also shows that the signal from the May 5 version is not as clean as it might be. A signal from a reference GNU Radio flow graph is shown for comparison. The reference signal has much more precise modulation of the four nominal tones. This may point to a further issue with the OpenRTX firmware, or a limitation of the MD-380 hardware, or perhaps a weakness in the test procedure due to the much stronger signal from the MD-380 as compared to the ADALM Pluto SDR used to transmit the reference signal.

OpenRTX: https://openrtx.org

OpenRTX firmware for M17 on MD-380: https://github.com/OpenRTX/OpenRTX/

Hardware modifications required for OpenRTX on MD-380: https://openrtx.org/#/md380_mods

M17 Project: https://m17project.org

Open Research Institute: https://m17project.org

A

This is a modulation demonstration of the md 380 modified with open rtx firmware right now. We've got it loaded with the april 30th 2022 version, which is known to have some problems.

A

I'm going to transmit the first time after power up and watch what happens on the spectrum.

A

Isn't that interesting.

A

Now we have the spectrum analyzer configured in vector signal, analyzer mode, it's configured for fsk, four tones 2400 hertz maximum deviation, 4 800 symbols per second, just like the m17 parameters, and if you can barely see those four little white circles on the line in the upper window, those are indications of where the tones are supposed to land and the red dots are where this, where the signal is. This is a no signal present and now I'm going to transmit with the handheld using the april 30th version of the software.

A

And you can see that it's grossly under modulated the tones are confined within the 800 or the 1 and -1 symbols and don't go anywhere near the plus 3 and -3 symbols, that's wrong and maybe the new version. The may 5th version of the firmware we'll fix that.

A

And now we have the may 5th version of the firmware loaded turn it on for the first time and transmit.

A

Here's the vector signal analyzer again, I'm transmitting with the may 5th firmware now. We've got red dots that are divided up in about the right, spacing centered on where they're supposed to be centered, but they're really kind of wide.

A

It should not be that wide.

A

By contrast, this is what the gnu radio flow graph looks like transmitting.

A

The four symbols are where they belong, but they're also confined within the little circles, and it's not because the signal is loud. The pluto puts out a very small signal compared to the ht.

A

So, what's wrong with the ht.