AnsweredAssumed Answered

Frequency Dividers?

Question asked by zheshmat on Jan 29, 2004
Latest reply on Apr 7, 2004 by zheshmat

Like • Show 0 Likes0
Comment • 6

Question: Has anyone successfully simulated analogue frequency dividers in ADS2002?

Background: I am trying to simulate a simple analogue frequency divider circuit using nonlinear elements (in this case a varactor) and some resonant circuits at the input and output frequencies. However I cannot see the subharmonic generation. I have tried both HB and transient, but don't see where the problem is! Any ideas?

No one else has this question

Outcomes

6 Replies

nilesh1 Jan 30, 2004 9:51 AM
Mark Correct
Correct Answer
I have seen frequency divider ckts in the past. They were successfully simulated in HB. I would recommend that you send in your circuit to Local Tech Support to have them try it out.

If that is NOT possible, then please share more information in this forum. Let us know, all the parameters you have set in the simulation controller (HB) and how the sources are setup and what sort of components you are using to create the varactor?

Thanks
nilesh 8)
Like • Show 0 Likes0
Actions
zheshmat Feb 2, 2004 7:37 PM
Mark Correct
Correct Answer
Thanx for your reply Nilesh!

With regards to the HB simulation: so far I have found that time-based simulation is the more natural solution to study analogue frequency dividers because it allows direct analysis. The HB technique is unable to spontaneously generate the subharmonic frequency which appears in frequency division. However, if I can find a way to use HB it will be more efficient.

The circuit I am trying to simulate consists of a source at the fundamental freq (1 GHz) and a sweeped power from 0-40 dBm. Then a series resonant LC network tuned at the fundamental. The varactor which is connected in shunt is the 'cbVar' model in ADS for a biased varactor. The output circuit also has a resonant LC network tuned at the half freq (0.5 GHz) and the load. The circuit is fairly symmetrical and simple.

Simulation parameters for Transient can vary but it is best if run until steady state is reached (0-500nsec). For HB, I even tried to make the expected tone the fundamental frequency, and use a source 2*fund instead. Then it would have in the list 0.5GHz and 1GHz, but

The one thing which makes this simulation slightly difficult is the dependancy of the output on several parameters, such as: input power level (there is a min threshold level), bias level of the vaarctor, input and output impedances, input freq etc.

I would be very happy if you can provide any help, to enable me to simulate freq division. Thanx!
Like • Show 0 Likes0
Actions
nilesh1 Feb 3, 2004 11:53 AM
Mark Correct
Correct Answer
Thank you for the extra information.

Well lets try a simple test first. can you get HB to simulate at just ONE POWER value. So you should NOT sweep the power level and just see if the freq divider works and if you get the outputs you want.

I would recommend that you download this example and check the HB setup used to run the divider simulation. See fract_divider_test.dsn. Your own test bench shld be similar to this.

If that still doesnt help, maybe you can send the files to your local tech support and that may give us a better clue on solving this issue.

Let me know how this goes.
Thanks
Nilesh 8)
Like • Show 0 Likes0
Actions
nilesh1 Mar 8, 2004 11:05 AM
Mark Correct
Correct Answer
I found out some more about Frequency Divider examples:

There are two examples on the "latest examples" web page that were developed.

http://eesof.tm.agilent.com/applications/latest.html#ads_latest

=================================================================

DivBy8_wTAHB_prj (1.0 MB)

This project shows how to simulate a 1/8 frequency divider, including phase noise, using transient-assisted harmonic balance (TAHB). The project runs only on ADS 2002, although the techniques can be used with versions of ADS as early as ADS 1.3.

This file will unarchive as the project DivBy8_wTAHB_prj.

NOTE. This example runs only on ADS 2002. It has been placed here in advance of the release of ADS 2002 as a convenience for Beta testing and field demos. For more information, please contact your Agilent EEsof EDA representative.

=================================================================

FreqDivider_prj (755 KB)

This example shows how to simulate a frequency divider or prescaler circuit. It includes both time-domain and frequency-domain simulations, as well as sweeps of the input clock signal frequency. Phase noise is also simulated, using the Transient-Assisted Harmonic Balance (TAHB) technique. The input and output signals are differential-mode.

This example requires the use of ADS 2001, although the same simulation techniques are also available in ADS 1.5.

This file will unarchive as the project FreqDivider_prj.

=================================================================
Like • Show 0 Likes0
Actions
zheshmat Apr 4, 2004 8:32 PM
Mark Correct
Correct Answer
I've been working on a parametric frequency divider design, and so far no luck with HB ot TaHB. I did try to test a few other frequency divider circuits such as regenerative and injection-locked frequency dividers, which I was sure of the design and circuits. Transient analysis worked fine and showed the frequency division in the simulations, but the same circuits didn't work under the HB analysis.???

Open to suggestions and comments ... Thanx!
Like • Show 0 Likes0
Actions
zheshmat Apr 7, 2004 2:16 AM
Mark Correct
Correct Answer
Just a quick addition/correction to the last message:

I had another go at using the TaHB with the regenerative and injection-locked frequency divider designs, and it worked! It shows frequency division provided you set the fundamental frequency in both Transient and HB setups equal to the smallest frequency present in the circuit, i.e. the output frequency and use an appropriate order value. Obviously the transient analysis has to be run first in the steady state region and then HB. This provides the HB simulator with a good initial guess, which is why TaHB works!
Like • Show 0 Likes0
Actions

Frequency Dividers?

Outcomes

Related Content

Recommended Content