CC1101: No transmission on design based on CC1101EM reference design, 915MHz

- How do you know it's no transmission?
- The required matching components are a function of the ground plane size and the casing. Have you tuned the antenna?
- Do you have access to a spectrum analyzer?

I'm assuming there's no transmission because the communication works using external radio modules hooked up to the boards.
Unfortunately I have no access to a spectrum analyzer, which makes debugging and tuning more difficult.

It sounds like you have access to a known good module with the Anaren one. Are you able to control this with SmartRF Studio or program i to measure RSSI? If you place the known good module next to your new module, do you see a difference in RSSI if you start sending with your new module?

You need at least something like this: miniradiosolutions.com/.../ to tune the antenna. To make a successful RF product you need some RF instruments.

Communication works between the Anaren module and the generic module. I haven't measure RSSI, but I'll try that as soon as possible.

One weird thing I noticed is that my generic 915MHz module actually uses the 433MHz reference design balun topology, but, as stated, works fine at 915MHz.

I also tried bypassing the antenna and its removing its matching components and then using a quarter-wave wire antenna connected directly to the 50 ohm feed line. Also no signal. I think it's something related to the balun circuit, but I'm out of ideas.

I managed to get access to a spectrum analyzer.
The radio is actually receiving a very weak signal (around -56dB), and a network analyzer shows almost total reflection (S11 around -0.88dB).
I'll post some screenshots later.

At which point in your design is this measured? Together with screen shoots, could you also post pictures showing how you measure with the spectrum/ network?

I measured using a SMC connector on my board connected to a Signal Hound BB60C. It's connected directly to the 50 ohm feed line.

Unfortunately I don't have pictures of my setup, but I'll try to get access to the spectrum analyzer again to run some more tests.

These are the screenshots:

This is measured on the receiver board.

For transmitting I have used an external CC1101 module which works. Here's the output signal (the registers are configured for 0dBm TX power):

Thanks

It could look like you are missing a cap somewhere and what you are measuring is the signal through some parasitic cap which attenuate the signal. I assume you have checked that all the components in the balun are correctly placed and that the soldering is ok?

Yes, I've double checked all the components and the soldering. Also tried multiple boards.

As a last resort, I'm running some SPICE simulations. I figured that the suggested inductor and capacitor values in the reference design don't really match the impedances for the desired frequency.
Analyzing the LC balun, the calculated values are 12nH and 2.7pF instead of the 18nH and 1.5pF. I don't know if it's actually an issue or the suggested values were found by another method.

Please post pictures of your setup when you get access to the spectrum again. Before that could you post a picture of the balun part of the board?
Based on the details you have given your board should work but based on what you write it sounds like a systematic error but it's a bit difficult to debug cases like this remote.

Here's the balun. I'll try to take a better picture but the layout and the soldering are visible in this one nonetheless.

Thanks.

The picture has a bit low resolution. It's not easy to see how C131 and L122 are connected.

Sorry, here's a better picture:

Thanks.

Everything looks correct. I don't have any ideas on what could be wrong at the moment. Debugging hardware remote is a bit difficult...

As a last resort, I'll try replacing the components on the LC balun for the ones I've calculated. The SPICE simulations gave better results with these new values.

When you measure with the spectrum, do you remove the components marked 1 and 2  and solder on a semi rigid or do you do it differently? What is the tract that goes up after the circle marked 2? 

The components marked 1 are matching components for a PCB antenna as described in http://www.ti.com/lit/an/swra227e/swra227e.pdf

When using the analyzer, I desoldered the matching components and soldered a 0 ohm 0402 resistor (marked 2) and an antenna connector to which I hooked up the analyzer.

Here's a picture of the board:

Thank you, it wasn't clear from your previous picture that you had a SMA on the board. The description of how you do it sounds correct.

Btw, looking at the picture you posted Tue, Aug 1 2017 9:19 AM: Have you added a wire and soldered to the chip?

What is the expected volume for this project?

This is a prototype board, the wire is connecting the CSN pin which was left unconnected due to a project error.

We don't have exact figures for the volume for this project. It's not large scale, at least initially, though. We may produce a first batch of 500 boards, I think.