Change in capacitance/resistance is not proportional to area

capacitance
bandwidth

#1

Dear sir,
According to the formula:capacitance=dielectric constant*area/length, as device area increases, the capacitance will increase while the resistivity will decrease, then the frequency, which equals to 1/RC, will maintain the same value.
However, I changed the device area and then found that the frequencies were not the same for different case.
In the dropbox (https://www.dropbox.com/sh/ailvkwtm40h4kjs/AACddf4EwKZa0t8dwT8jP63_a?dl=0) are my simulation files, hope you can help me solve this problem.
Thank you very much.

Best regards,
Meng-Chin Lee 李孟津
Graduate Institute of Photonics and Optoelectronics, NTU
Mobile: +886-989580210
E-mail: r05941061@ntu.edu.tw


How to enable mobility model
#3

Hi @r05941061, The reason you do not have identical bandwidth in these different structures is because the bandwidth of this particular type of photodetector is not restricted by the RC value but rather the transit time of the photo-generated electrons and holes. In the narrow devices the photo-generated electron-hole pairs are under a strong electric field between the top and bottom contact and they have a smaller path to travel which makes the bandwidth larger. In the wider devices come of the photo-generated electron-hole pairs are under the influence of a weaker electric field (away from the top contact) and also these carriers have a longer path to travel which makes the photodetector slower and the bandwidth smaller.

As a side note I also wanted to point out that it looks like you have used the same generation rate for all the structures. Please note that this will give you inaccurate results as the distribution of the photo-generated careers will effect their transit time and hence the bandwidth.


#4

Dear sir,
I’ve checked the electric field of the device, which is in the attachment. It seems to be the same in the devices with different width. As for the optical generation, it is under-building now. Thank you.

Bestregards,
Meng-Chin Lee 李孟津
Graduate Institute of Photonics and Optoelectronics, NTU
Mobile: +886-989580210
E-mail: r05941061@ntu.edu.tw


#5

Hi @r05941061, you are right. I stand corrected. Since you have the entire top surface of the germanium heavily doped you should have a constant electric field over the width of the device. However the different patterns of the generation rate may still affect the bandwidth since the bandwidth is dependent on transit time rather than the RC value. This is because for one case you may have electrons and holes getting generated closer to the electrodes and in another case farther away from the electrodes which will affect the transit time. i will be very interested to see your results with the correct generation rates once you have them. Cheers!