Modelling transmission lines based on DXF-layout in HF-rectifier, ADS?
时间:04-06
整理:3721RD
点击:
Hi,
Im working on a high sensitivity rectifier design to use for a energy harvesting application at 2.45GHz.
The approach Ive used is a 5-stage cascaded Villard voltage doubler circuit, with the Skyworks SMS7630 Diode in a SC-79 package.
Freq= 2.45GHz
Pin = 0dBm to -20dBm
Ive simulated the circuit with S-parameter and transient analysis in ADS.
So far Ive managed to get an effeciency of around 34% at Pin = -10dBm, and an output voltage of 1.15. In this example the impedance is matched to 50 ohms, with SMA-model and transmission lines neglected.
In the application Im working on, I intend to use a standard omnidirectional antenna connected to a straight SMA-connector on the PCB, and not a rectenna approach.
So over to the question(s):
1.Linecalc suggested W = 24mil L= 555mil for my substrate while using stripline. Tracewidth of 24 will be possible, but the length might be a problem. Would you just compensate for that factor in the matching network to the antenna?
2. For this purpose, would you do the layout in ADS or in for example Altium and add transmission lines manually? Have anyone got experience with importing DXF into ADS and do the matching based on the layout? In this case, could pads be neglected?
3. How would you model the SMA-connector in the circuit? Just a series resistance and cap to ground or are there specific models for this in ADS? (Im using 2011)
Im still a novice with simulating RF, so if Ive got something wrong. Please shed some light!
Any comments or suggestions are appreciated!
(Example simulation results: 5stage_villard_sim_dblank.pdf)
Thanks,
Dave
Im working on a high sensitivity rectifier design to use for a energy harvesting application at 2.45GHz.
The approach Ive used is a 5-stage cascaded Villard voltage doubler circuit, with the Skyworks SMS7630 Diode in a SC-79 package.
Freq= 2.45GHz
Pin = 0dBm to -20dBm
Ive simulated the circuit with S-parameter and transient analysis in ADS.
So far Ive managed to get an effeciency of around 34% at Pin = -10dBm, and an output voltage of 1.15. In this example the impedance is matched to 50 ohms, with SMA-model and transmission lines neglected.
In the application Im working on, I intend to use a standard omnidirectional antenna connected to a straight SMA-connector on the PCB, and not a rectenna approach.
So over to the question(s):
1.Linecalc suggested W = 24mil L= 555mil for my substrate while using stripline. Tracewidth of 24 will be possible, but the length might be a problem. Would you just compensate for that factor in the matching network to the antenna?
2. For this purpose, would you do the layout in ADS or in for example Altium and add transmission lines manually? Have anyone got experience with importing DXF into ADS and do the matching based on the layout? In this case, could pads be neglected?
3. How would you model the SMA-connector in the circuit? Just a series resistance and cap to ground or are there specific models for this in ADS? (Im using 2011)
Im still a novice with simulating RF, so if Ive got something wrong. Please shed some light!
Any comments or suggestions are appreciated!
(Example simulation results: 5stage_villard_sim_dblank.pdf)
Thanks,
Dave
I think that is the length for 1/4 wavelength !? If you need a 50 ohm transmission line to connect to a 50 ohm antenna, impedance matching will NOT change with line length. Just use any length that you need for layout purposes.
The connector is designed to maintain the 50 ohm line impedance. It adds some length/phase, but does not change matching. Thus, there is no need to include it in simulation.
- ABCD Matrix of transmission lines connected in cascade
- are 50 ohm transmission lines required for 2.4GHz amplifier?
- Guidelines to design a custom RF T-pad attenuator
- How to avoid the warning:"One or more port lines were lost due to geometry modificati
- Microstrip coupled lines
- Guidelines about RF/IF test point