Phototransistor vs Photodiode

[mbu]

Phototransistor vs Photodiode: Which is better to use in a tachometer circuit for the sensing of a rotating shaft? In other words, which will react faster when used as a switch?

Thanks…

 

[dknguyen]

A photo transistor has higher sensitivity with lower response time.

 

[gramo]

So for "speed" sensitive designs, you should use photo-diodes?

There are thousands on digikey and futurlec, anyone recommend a popular dip package one? Or should I just go lucky dip (no pun intended )

 

[mbu]

dknguyen,

Thanks for the response!

Since the sensor is going to be used in a tachometer, then I guess the best one to use would be the photodiode.

By the way... is there any site on the web that would tell me the response time of a phototransistor and a photodiode? I didn't see anything in the datasheets for the items.

Thanks...

 

[dknguyen]

No idea. My guess is way faster than what you need. They are solid-state devices after all and can probably switch at at least a hundreds of khz with no problems. Just a guess. I am not sure. Regular diodes and transistors of such small sizes switch on the order of 10s of nanoseconds (this is affected by the drive voltages, so the intensity of light also has an effect on how fast the photodiode/transistor will switch).

Cadmium sulphide cells (I think) respond slowly enough so that they ignore the flashes of light from fluorescent lights (around 60Hz). But those aren't semiconductors so...

Here are some examples from Googling "phototransistor response time". Didn;t look over them carefully though:
**broken link removed**
**broken link removed**

Oh, BTW, my first response I got from Wikipedia under Photodiodes.

 

[Roff]

I don't normally think of a tachometer as a demanding application for a phototransistor. Even 100,000 RPM is only 1.67kHz. even if your sensor is only active for 10 degrees out of 360, the pulse width would be 16.7 usec. I doubt you have something turning at 100,000 RPM.

 

[Sceadwian]

I just did a quick Google and Cadmium sulfide has a responce time under bright light of about 250rise to 300fall micro seconds. That's about 2khz. The 101 electronics project my stepson has has a circuit that uses an amplifier on a Cadmium sulfide cell to listen to modulated light sources and it says you can pick up the flicker from even an incandescent bulb. Cadmium selenide is supposed to have a rise time of 17usecs and a decay of 8usecs that's in the 40khz range.

 

[mbu]

Thanks for the information everyone!

 

[Ayne]

In tachometer(Handheld) we use infrared receiver modules for receiving reflected signle..
infrared receiver modules mean, infrared receivers which we use in TVs for receiving data coming from remotes..
I mean we send countinous signal or 38Khz modulated signal...

 

[Hero999]

Cadmium sulphide cells (I think) respond slowly enough so that they ignore the flashes of light from fluorescent lights (around 60Hz). But those aren't semiconductors so...

Cadmium sulphide is a semiconductor, if it wasn't then it wouldn't work. It's for useful some applications because as well as being slow it has a similar spectral response to the human eye.

I've heard about some people using cadmium sulphide as tachometers at school so phototransistors shouldn't have a problem.

 

[AllVol]

How would a Hall effect sensor stack up for this application?

 

[Hero999]

Pretty well, there often used for this sort of thing. Open up any old flopy drive and you'll see hall effect sensors mounted near the magnetic rotor.

 

[Ayne]

In tachometer(Handheld) we use infrared receiver modules for receiving reflected signle?????????????????????
infrared receiver modules mean, infrared receivers which we use in TVs for receiving data coming from remotes..
I mean we send countinous signal or 38Khz modulated signal?????

 

[ikalogic]

In tachometer(Handheld) we use infrared receiver modules for receiving reflected signle?????????????????????
infrared receiver modules mean, infrared receivers which we use in TVs for receiving data coming from remotes..
I mean we send countinous signal or 38Khz modulated signal?????

no never modulated signals, as it will leave time gaps where pulses may be missed at very high speed..

check my projects page,here is how to build a contact less tachometer here:
**broken link removed**

with all the details you need to know.. it also shows how to build a cheap IR sensor... that worked very fine for this application.. i could measure shapts rotating at up to 12000 Rpm very easily.. (i don't have any equipment that turns any faster.... but theoreticaly.. you can measure much more using this method..

good luck

 

[mbu]

ikalogic,

Great! I'm going to go thru your information but will take some time since I need to study it for awhile. Will you be around here so if I have some questions I can get in touch with you? Right now I'm mainly interested in your tachcometer and related devices. I would like to make a tachometer for a lathe and need fairly accurate readings for rpm's between 5 and 2000.

Thanks...

Mike

 

[mneary]

no never modulated signals, as it will leave time gaps where pulses may be missed at very high speed

If you're using a fairly common 38 KHz detector, then you should expect it to respond faster than a kilohertz.

Do the math on your actual rotational speed. A 38 KHz modulation can make a big difference in noise immunity.

 

[ikalogic]

ikalogic,

Great! I'm going to go thru your information but will take some time since I need to study it for awhile. Will you be around here so if I have some questions I can get in touch with you? Right now I'm mainly interested in your tachcometer and related devices. I would like to make a tachometer for a lathe and need fairly accurate readings for rpm's between 5 and 2000.

Thanks...

Mike

i'll try to be here for a couple hours a day...

now, for accurate readings between 5 and 2000, you may only need to make some changes to the program, but the hardware will be the same... i mean for the low range of the the tachometer reading like100 or below, it's better to measure the time between each 2 pulses instead of couting the number of pulses every X seconds, where X is usually less than 1..

good luck, and i shall be with you if you need anything..

 

[mbu]

ikalogic,

Thanks!

I'm going to have to make this tachometer using a MicroChip microcontroller since the programming software for it is what I have on hand. I have never worked with a LCD display so will a PIC 16F628A chip be able to run your circuit?

Mike

 

[ikalogic]

ikalogic,

Thanks!

I'm going to have to make this tachometer using a MicroChip microcontroller since the programming software for it is what I have on hand. I have never worked with a LCD display so will a PIC 16F628A chip be able to run your circuit?

Mike

Sure it will... Using an LCD is very easy.. and a PIC is even faster than 8051.. and, Nigel is also here to help you anything realted to PIC programming

 

[Ayne]

When light strike on Photodiode it's resistance increase or decrease??