Menu
▾
▴
Re: [Qucs-help] How to simulate a RF circuit with differential signals ?
|
From: Andrés M. M. <and...@gm...> - 2017-03-24 05:03:39
|
2017-03-23 14:48 GMT+01:00 Cristiano Gavião <cvg...@gm...>: > > > Looking at your examples I saw that my attempt is not right. seems that I will need to use a transformer in order to deal with the differential signal... am I right? could you explain this a bit ? That transformer I included is an ideal balun. Its purpose is to convert a balanced signal into an unbalance one. Notice that the app circuit you've posted has a LC balun ( http://analog.intgckts.com/impedance-matching/lumped-lc-balun/). The main difference between that circuit and the ideal transformer is matched for all frequencies whereas the LC balun isn't I noticed that you're working at 900 MHz. At that frequency, you'll need to include the transmission lines (microstrip, for example) and include the RF models of the capacitors and inductors. You can get that models (s2p) from the Murata webpage ( http://ds.murata.co.jp/software/simsurfing/en-us/index.html). Even if Murata is not your provider, that models will be more realistic than the ideal inductor/capacitor. > > > what is specified in the datasheet is that the balun is seeing an differential impedance of 44 +j15 Ohm. According to the datasheet, you device should see 44+j15 Ohm for optima performance. Since Qucs cannot deal with power sources of complex impedance, you can do the following as a workaround: Place a 50Ohm port between RFp and RFn and see the S11. Then compare that with the S11 of a 44+j15 load in a 50 Ohm system (You can use the s1p data component for synthesizing a 44+j15 impedance). At this point you can tune you network so as its S11 is similar to the 44+j15 response. I'll try to elaborate further at night. To be honest, I've never done a differential impedance matching. But that's the way I would do that... > I notice that you have divided the total impedance of 50 Ohm for two (25 to each port), but how to deal with imaginary part in my case (44 +j15) ? > > best regards, > > Cristiano > > > On 22/03/2017 16:21, Andrés Martínez Mera wrote: >> >> Hello Cristiano, >> >> There were no images attached in the previous message, probably because the mail server rejects attachments > 100kB Could you post the link to the datasheet? >> >> If you want to see how you network performs in real conditions (modulated signal), you need to capture/simulate the signal induced in your antenna and use the file-based voltage source component. Anyway, I reckon that having a single tone (as you already did) is quite meaningful to see the relationship between the different signals. In my opinion, the magnitude of the signal is not really important here since it will depend on the signal strength (I assume that you network is passive). >> >> Regarding the impedance matching problem, you need to know the input impedance of the RF_P and RF_N ports. Once having this, you'll need to place them into the circuit (using the 1 port s-parameter component) and see the S11 at your antenna. >> >> Below you can find two example schematics to illustrate the transient and the S-parameter analysis in a circuit with a balun and a lowpass filter. >> >> Transient analysis: http://pastebin.com/nPDfBirA >> S-par analysis: http://pastebin.com/mmZ9fMt5 >> >> Hope this helps, >> >> Andrés >> >> >> 2017-03-20 23:31 GMT+01:00 Cristiano Gavião <cvg...@gm...>: >>> >>> Hello, >>> >>> I'm a programmer with some knowledge on electronic and learning RF and QUCS. >>> >>> I've have a RF IC and a reference design that I took from its datasheet >>> (image attached). >>> >>> First I would like to do setup a transient simulation just to >>> investigate its behavior and then a simulation to help me calculate and >>> tune the impedance of the board and matching with the antenna. >>> >>> This IC has two differential ports (RF_N and RF_P), plus a RXTX(Optional >>> bias pin for the RF LNA). it will be connected to a 50ohm antenna. >>> >>> >>> >>> enter image description here >>> >>> I start to play a bit with QUCS in order to have a transient >>> simulation. But I got a bit confused about how to deal with the >>> differential signals. :-\ >>> >>> Initially I created two AC voltage sources and inverted the phase for >>> one of them. >>> >>> Well, seems that it worked, but not sure yet if the amplitude voltage is >>> right since it is not informed in datasheet. will discover that soon. >>> >>> could someone tell me if the path I took is right? >>> >>> if so, what component should I use for RXTX port then? >>> >>> >>> to end, as soon I do that transient simulation what path should I take >>> in order to calculate and tune the impedance matching for my board? is >>> the S-Parameter the simulation that I need ? >>> >>> thanks for any tip, suggestion or link to documents/articles/tutorials. >>> >>> best regards, >>> >>> Cristiano >>> >>> >>> >>> >>> ------------------------------------------------------------------------------ >>> Check out the vibrant tech community on one of the world's most >>> engaging tech sites, Slashdot.org! http://sdm.link/slashdot >>> _______________________________________________ >>> Qucs-help mailing list >>> Quc...@li... >>> https://lists.sourceforge.net/lists/listinfo/qucs-help >> >> >> >> >> -- >> Andrés Martínez Mera > > -- Andrés Martínez Mera Hello Andrés, First of all, many thanks for your time and for the example files. ;-) I notice that mail server reject my first post but somehow the reject message was to my junk folder. I checked the site and didn't see my post there, then I tried again... but now seems there are two copies of it... sorry guys. :-[ I've uploaded the schematic image from the datasheet here: https://postimg.org/image/84tvk5mph/ , the complete datasheet is here: http://www.ti.com/lit/ds/symlink/cc1310.pdf Also, I've uploaded the image of my first attempt here: https://postimg.org/image/mpfjsq71x/ Looking at your examples I saw that my attempt is not right. seems that I will need to use a transformer in order to deal with the differential signal... am I right? could you explain this a bit ? what is specified in the datasheet is that the balun is seeing an differential impedance of 44 +j15 Ohm. I notice that you have divided the total impedance of 50 Ohm for two (25 to each port), but how to deal with imaginary part in my case (44 +j15) ? best regards, Cristiano On 22/03/2017 16:21, Andrés Martínez Mera wrote: Hello Cristiano, There were no images attached in the previous message, probably because the mail server rejects attachments > 100kB Could you post the link to the datasheet? If you want to see how you network performs in real conditions (modulated signal), you need to capture/simulate the signal induced in your antenna and use the file-based voltage source component. Anyway, I reckon that having a single tone (as you already did) is quite meaningful to see the relationship between the different signals. In my opinion, the magnitude of the signal is not really important here since it will depend on the signal strength (I assume that you network is passive). Regarding the impedance matching problem, you need to know the input impedance of the RF_P and RF_N ports. Once having this, you'll need to place them into the circuit (using the 1 port s-parameter component) and see the S11 at your antenna. Below you can find two example schematics to illustrate the transient and the S-parameter analysis in a circuit with a balun and a lowpass filter. Transient analysis: http://pastebin.com/nPDfBirA S-par analysis: http://pastebin.com/mmZ9fMt5 Hope this helps, Andrés 2017-03-20 23:31 GMT+01:00 Cristiano Gavião <cvg...@gm...>: > Hello, > > I'm a programmer with some knowledge on electronic and learning RF and > QUCS. > > I've have a RF IC and a reference design that I took from its datasheet > (image attached). > > First I would like to do setup a transient simulation just to > investigate its behavior and then a simulation to help me calculate and > tune the impedance of the board and matching with the antenna. > > This IC has two differential ports (RF_N and RF_P), plus a RXTX(Optional > bias pin for the RF LNA). it will be connected to a 50ohm antenna. > > > > enter image description here > > I start to play a bit with QUCS in order to have a transient > simulation. But I got a bit confused about how to deal with the > differential signals. :-\ > > Initially I created two AC voltage sources and inverted the phase for > one of them. > > Well, seems that it worked, but not sure yet if the amplitude voltage is > right since it is not informed in datasheet. will discover that soon. > > could someone tell me if the path I took is right? > > if so, what component should I use for RXTX port then? > > > to end, as soon I do that transient simulation what path should I take > in order to calculate and tune the impedance matching for my board? is > the S-Parameter the simulation that I need ? > > thanks for any tip, suggestion or link to documents/articles/tutorials. > > best regards, > > Cristiano > > > > > ------------------------------------------------------------ > ------------------ > Check out the vibrant tech community on one of the world's most > engaging tech sites, Slashdot.org! http://sdm.link/slashdot > _______________________________________________ > Qucs-help mailing list > Quc...@li... > https://lists.sourceforge.net/lists/listinfo/qucs-help > -- Andrés Martínez Mera |