CC1110 433Mhz Simulation Help

Tony Siu

Intellectual 610 points

Dear Sir,

I have tried to simulate the schematic of the 433MHZ Reference Design CC1110EM 2.0

the link below show the drawing in Agilent's ADS schematic.

http://picasaweb.google.com/siuchuntung/Rf?authkey=Gv1sRgCJipoK7J1viVXQ#5557449382230300978

the link below show the simulation result.

http://picasaweb.google.com/siuchuntung/Rf?authkey=Gv1sRgCJipoK7J1viVXQ#5557449383528708194

It seems that the matching is not done at the frequency of 433MHZ when I try to follow the schematic.

Would any one tell me if I have done anything wrong while I am doing the simulation?

Another question:

When I am doing a new design with the same PCB thickness (0.8mm) with slightly different dielectric constant (4.2 vs 4.5 used in the reference design), can I just copy the layout from the 433MHZ CC1110EM 2.0? Do I have to change the width or height (or both) to make sure it is 50ohms and have the same angle as the reference design?

Thanks a ton!

Tony

over 14 years ago

0 Suyash Jain over 14 years ago

TI__Genius 10650 points

Hello Tony,

I see that impedance number for port 2 is Z = 116 +j*41, this is the value chip wants to see as the optimum load. When looking into the chip i.e. the port impedance, it is the conjugate i.e. use Z for CC1110 as 116-j*41. Let me know the simulation result after this change. Thanks

Changing dielectric constant from 4.5 to 4.2 should not change the performance a lot at 433MHz and the bandwidth of the design should allow for that as well. I will suggest checking your simulation results for confirmation.

Regards,

Suyash Jain

0 Tony Siu over 14 years ago in reply to Suyash Jain

Intellectual 610 points

Dear Suyash,

Greeting for the new year!

Thank you for your help! It seems that the simulation is closer to 50ohms but there seems to be some thing needs to be proved, would you please help?

I have changed the impedance to -j41 and adjusted some short traces betweens components to some more realistic lengths:

http://picasaweb.google.com/siuchuntung/Rf?authkey=Gv1sRgCJipoK7J1viVXQ#5558330984530506338

Here is the simulation result:

http://picasaweb.google.com/siuchuntung/Rf?authkey=Gv1sRgCJipoK7J1viVXQ#5558331301983235538

It seems that it is still not quite at 50 ohms. and the central frequency is not quite at 433mhz. Do you think if it is already an acceptable result?

I then tried to do a simulation in Momentum as follows:

http://picasaweb.google.com/siuchuntung/Rf?authkey=Gv1sRgCJipoK7J1viVXQ#5558330993585701762

http://picasaweb.google.com/siuchuntung/Rf?authkey=Gv1sRgCJipoK7J1viVXQ#5558331001805956002

The central frequency seems to be even further from 433 mhz

http://picasaweb.google.com/siuchuntung/Rf?authkey=Gv1sRgCJipoK7J1viVXQ#5558331006350038258

I would be great if you could give me more direction.

Best regards,

Tony Siu

0 H Stewart over 14 years ago in reply to Tony Siu

Mastermind 8170 points

The optimal load must be determined using "load pull" techniques where an adjustable load is connected to the device, adjusted for best performance, and then measured with a vector analyzer. A matching network that presents this load to the device is then designed. The load impedance at the fundamental and out several harmonics (especially odd harmonics) is important. We are not given this load information and a simple 50 ohm match at the fundamental is not the answer, the odd harmonic energy must be reflected back into the device at the proper phase for optimal efficiency. The best approach is to model the Ti reference design at the device terminals (without the device) looking toward the load through the 5^th harmonic. You will see series Ls and even small shunt Cs close in to the device have their effect at the higher harmonics. As you make changes to the design compare the new simulation results with the baseline Ti reference design results.

0 Suyash Jain over 14 years ago in reply to H Stewart

TI__Genius 10650 points

Hello Tony,

Your return loss is high and insertion loss is low and looks acceptable at desired frequency. I will also agree with H Stewart that best way to go will be to compare your design with TI reference design. Also a tip for the simulations, ADS allows you to import gerber files and then you can create momentum model for the imported design. Importing that momentum model in the schmatic along with component values will give you much closer modelling of the actual circuit and will be easier than drawing each T-line.

Regards,

Suyash Jain

0 Urikaka over 9 years ago

Intellectual 320 points

hello tony
i start learn about CC1110. i want to simulate CC1110 with altium but i didn't find the library CC1110 for altium.
if you use altium for simulating CC1110, please give me library CC1110. if not, please talk with me which simulation u used.
thank you!

0 TER over 9 years ago in reply to Urikaka

TI__Guru**** 317180 points

user4653634: What do you want to simulate? This thread is about simulating the match for the CC1110 or in other words the PCB and the passive components. You need a simulator like ADS to do this. You don't need a library for CC1110 for Altium.

0 Urikaka over 9 years ago in reply to TER

Intellectual 320 points

i want to simulate Application Circuit for CC1110Fx 315 MHz. it is figure 10, page 38 of cc1110 datasheet.
this is link to datasheet of TI:
www.ti.com/.../cc1110-cc1111.pdf

0 TER over 9 years ago in reply to Urikaka

TI__Guru**** 317180 points

You should do simulations on the reference design and not on the circuits in the datasheet.
- The circuits shown in the datasheet are simplified and are not showing all components
- The layout have an impact on the frequency response of the filters. Thus you should use a EM simulator on the layout and use a model library for the passive components.