Continue to Site

Welcome to EDAboard.com

Welcome to our site! EDAboard.com is an international Electronics Discussion Forum focused on EDA software, circuits, schematics, books, theory, papers, asic, pld, 8051, DSP, Network, RF, Analog Design, PCB, Service Manuals... and a whole lot more! To participate you need to register. Registration is free. Click here to register now.

Register Log in

How to calc value for R and C for low pass filter

Status
Not open for further replies.
K

kaixiang

Newbie level 4
Joined
Apr 2, 2009
Messages
6
Helped
0
Reputation
0
Reaction score
0
Trophy points
1,281
Activity points
1,337
Hi guys,
I have Hamamatsu S9648-100 photodiode.

The photodiode must be reverse-biased so that a positive potential is applied to the cathode.
To eliminate the hgh-frequency components, it is recommend to place a load capacitance CL in parallel with load resistance RL as a low-pass filter.

Cut-off frequency fc=1/(2*pi*CL*RL)

my question is how do we get the CL and RL?
attached below is my circuit!
 

Hi,
I have a suggestion that first of all you should know about the cut of frequency for a specific application. for example if you design it for a 300 Khz frequency then place 300Khz instead of fc. take any value of resistance for example 1Kohm's. solve the equation you will be able to get the capaciatance. thats how you can calculate the values.
Cl = 1/(2*pi*300Khz*1Kohm)
 

I have forget the reference but it is assumed that time constant of the low pass filter circuit should be atleast 3 times of the period of input pulse.
 

hi keith, i have question,he is using dc power supply, no frequency component there, then how can he choose it. correct me if i am wrong..
 

24+ said:
hi keith, i have question,he is using dc power supply, no frequency component there, then how can he choose it. correct me if i am wrong..
Click to expand...

Even for a "static" measurement you need to define a response speed. For example, I have worked on system where the response speed is deliberately designed to be tens of seconds. That may be too slow for this application. You cannot define the required filtering without a bit more information.

Keith.
 

hi keith, actually i'm using the photodiode to calculate the speed up to 80km/hour which is 22.22meter/second so does that mean my response speed is 22.22?
 

Are you breaking two through beams? The sort of response speed will depend on the accuracy you want. So, if you want 1m/s accuracy/resolution I would suggest you need a signal rise/fall time less than the error of 1m/s in 22m/s. However, what that translates to in time depends on how far apart your sensors are (assuming you are breaking two beams in succession).

Keith.
 

Hi keith,
The photodiode I am using will be exposed to the sunlight, so when a vehicle pass by it will block the light into the photodiode, i assuming the maximum speed of the vehicle is 80km/h. Currently I am doing a prototype so the distance between the photodiode will only be around 8cm
Btw, how do I find the signal rise/fall time less than the error of 1m/s in 22m/s

[photodiode1]<-- 8cm -->[photodiode2]
^
|
8cm
|
[photodiode3]
 

At only 8cms, at 22m/s you will cover that in 3.636ms. At 23m/s you will cover it in 3.478ms, so for 1m/s resolution you need to resolve 158us - the difference between the two. That gives you an idea of the sort of timer you need as well - 10kHz or faster, I would suggest.

I would suggest you want your rise time faster than the tiem you want to discriminate, although eventually there will be a trade off between extra noise and not getting any more precision. You would get 158us rise time from single pole filter with a cut off of around 70us.

Keith.
 

Status
Not open for further replies.

Similar threads

Part and Inventory Search

Welcome to EDABoard.com

Sponsor

Back
Top