anybody know about the MOD$$T simulation of e1do rf?

EDAboard.com Forum Index -> Software Problems, Hints and Reviews -> anybody know about the MOD$$T simulation of e1do rf?

Author

Message

cqmyg5

Joined: 21 Oct 2003
Posts: 142
Helped: 1

14 Jun 2007 1:22 anybody know about the MOD$$T simulation of e1do rf?

Hi,

I'm studying the MOD$$T simulaiton of e1do rf, there's a question,

Is it possible to run co-sim with eldo and eldo rf? That is to co-sim some RF block(in spice) running in SST simulaiton and some ANALOG block(spice) running in TRAN?

Last edited by cqmyg5 on 21 Jun 2007 18:36; edited 1 time in total

bozzo

Joined: 01 Sep 2005
Posts: 34
Helped: 4
Location: Europe

14 Jun 2007 10:07 Re: anybody know about the MODSST simulation of eldo rf?

Hello,
Yes you can co-simulate with EldoRF differents blocks running either in .TRAN
or in .MODSST
the syntax to specify how to run in .MODSST the desired blocks is :
.part modsst inst=(name, ...)
so the rest of the circuit will be analyzed in .TRAN

feel free for more questions !
Bozzo.

cqmyg5

Joined: 21 Oct 2003
Posts: 142
Helped: 1

14 Jun 2007 19:59 anybody know about the MODSST simulation of eldo rf?

I tried the ".part modsst inst=(name, ...) ", but the simulation result is not correct.

Actually my circuit contains a lna, a mixer and a bpf, if used SST, it can give me right result. But using ".part modsst inst=(Xlna, Xmixer) " for lna and mixer, ant MOD$$T along with two tones SST for whole ciruit, it gave me wrong result.

Last edited by cqmyg5 on 22 Jun 2007 19:27; edited 2 times in total

Google
AdSense
Google Adsense

14 Jun 2007 19:59 Ads

bozzo

Joined: 01 Sep 2005
Posts: 34
Helped: 4
Location: Europe

15 Jun 2007 11:18 Re: anybody know about the MODSST simulation of eldo rf?

Hello,

MODSST works for any Ntone simulation but the right way to simulate a 2tone circuit
is to specify a 1tone SST along the LO in order to have the second fundamental frequency
as a modulation of the LO (or one of its harmonics).

for instance if your mixer has LO=910meg and RF=900meg
the correct setup for a MODSST simulation is :

.SST FUND1=910meg NHARM1=10
.param valfund2=900meg
and where you have sources along RF put valfund2 in place of FUND2
So the excitation of the RF sources will be considered as a modulation of FUND1.

Last thing : there are no limitations concerning the type of the blocks in .part modsst

Bozzo

cqmyg5

Joined: 21 Oct 2003
Posts: 142
Helped: 1

15 Jun 2007 23:45 Re: anybody know about the MODSST simulation of eldo rf?

Thank Bozzo,

I have tried your solution of MOD$$T with one tone, but it doesn't work.

I am so sad, because I have studied it for 5 days, and still don't know what's the problem. The IF should be 1MHz, two tones SST is OK, but MODSST(.part modsst inst=(XLNA,XMIX)) with one tones SST is failed.

Last edited by cqmyg5 on 22 Jun 2007 19:28; edited 3 times in total

bozzo

Joined: 01 Sep 2005
Posts: 34
Helped: 4
Location: Europe

18 Jun 2007 10:08 Re: anybody know about the MODSST simulation of eldo rf?

Hello,

I am not sure you use correct syntax to display waveforms at the outputs.
If you want to display nodes in .TRAN block you have to use the following syntax :
.plot tran v(V1I) v(V1Q)

concerning the nodes at the interface between the mixers and the BPFs, you
have to use .plot tmodsst v(IFI) v(IFQ), and in order to have the best representation
for the waveforms, please use .option modsst_full_display=2 (the simulation will be slower).

please post the new .JPG. Which release do you use (ams2006.1, 2, 2a, 2b) ?

Bozzo.