5G signal sources and simulation inside Spectre / Virtuoso

sidm said:

do we have signal sources for 5G signal simulations inside Spectre / Virtuoso like there is a LTE source in RFLIB ?

Yes, it's called "nr" (New Radio, as mentioned here: https://en.wikipedia.org/wiki/5G_NR )

Andrew

thank you very much Andrew for the reply. Is there any documentation which demonstrates the usage of this NR source ?

I couldn't find one on COS.

This should hopefully be added soon to the Spectre Circuit Simulator RF Analysis Library Reference manual in the SPECTRE documentation via CCR 2521540. In other words this manual: Spectre Circuit Simulator RF Analysis Library Reference -- rfLib Library - Wireless Components Category

It's just not been added yet. In the meantime, you can at least see some information about the parameters using "spectre -h wsource" and looking at the "NR Parameters" section.

Andrew

thank you very much for the confirmation Andrew.

For the LTE signal source which parameter can I use to generate a LTE waveform at 2Ghz ? As of now my CDF form for this source looks like the following -

Below is the snapshot of the schematic testbench with the LTE source -

regards

It's a bit tricky to get exactly 2GHz (I didn't try to figure it out entirely because there's a limit to how much of the ETSI standard I wanted to read for fun - https://www.etsi.org/deliver/etsi_ts/136100_136199/136101/15.03.00_60/ts_136101v150300p.pdf ). Anyway, if you pick Operating Band 23  and Channel Number 25550 then you get 2.005GHz (with a 10M channel bandwidth). There may be other ways of getting the frequency you want - but the source is controlled by the operating band and channel numbers in the standard.

By the way, I also realised that the NR source is in the documentation - unfortunately it's just missing the heading and so is not shown in the contents page or the index. If you follow the link I gave in my previous reply, and then scroll down from the LTE source, you'll see a bit where it has a heading "SYMBOL" and then below that it's describing the NR Modulation Source (here's a screen shot of part of it):

Andrew

Thank you very much Andrew for the details.

I will dig deeper into the insights you have provided.

For nr and LTE sources is the wave probe the main method to perform measurements even for signal waveform and power ? For example to measure the output power in dbm?

I have been performing envelope simulation as per LTE RAK and getting mainly the psd output in db as of now.

In envelope analysis I see a variable W0_wfreq under the spectrum analysis that has default frequency value of 1GHz. Can this be changed to set the LTE frequency to 2GHz ?

Regards

Thank you very much Andrew for the details.

I will dig deeper into the insights you have provided.

For nr and LTE sources is the wave probe the main method to perform measurements even for signal waveform and power ? For example to measure the output power in dbm?

I have been performing envelope simulation as per LTE RAK and getting mainly the psd output in db as of now.

In envelope analysis I see a variable W0_wfreq under the spectrum analysis that has default frequency value of 1GHz. Can this be changed to set the LTE frequency to 2GHz ?

Regards

The wireless options on the Direct Plot form give you a variety of things you can measure (it's not entirely clear what you would mean by the "output power in dBm" because of course you have a spectrum of output - that's what the Direct Plot->Main Form->Wireless->Spectrum give you (you get the average main channel power, and ACPR for the two sides). You can also (with the CCDF choice on the wireless form) get the average power and peak power as an output measurement in ADE. There are other measurements like BER and EVM as well as constellation plots.

It doesn't make sense to set W0_wfreq - the whole point is that the analysis sets that to the frequency for the channel. This is mainly intended for when you're using the baseband source, so that you can set your LO frequency to match the expected channel specified on the baseband wsource. So during simulation, W0_wfreq will be updated to match the channel - ADE just has an arbitrary default value just to keep ADE happy, but that value isn't really used. The reason why it does it this way is that you can then specify the channel on the source, the LO frequency is set appropriately, and the frequency for the wireless analysis will also be automatically set so that it skips the right frequency periods.

Andrew

 Thank you very much Andrew for the details. In my case I don’t see the CCDF option , not sure what can be the reason for that ?

Snapshot below for reference – (in the Wireless RAK the CCDF option is shown where the cursor is in the below snapshot)

thanks

Looking through when this was added, I think you need to be using at least SPECTRE18.1 ISR8 and IC6.1.8/ICADVM18.1 ISR6 (so with .500.6 on the end). My guess is you're not meeting one or the other of these.

There's a bit more about it in this article: What is the CCDF function in the Direct Plot Form after running wireless analysis?

Andrew