Maker Pro
Maker Pro

Choosing between photodiodes and phototransistor designs for a amplification circuit.

JPPoulin

Sep 21, 2012
8
Joined
Sep 21, 2012
Messages
8
Hello everyone,

We need to construct a very low cost circuit that need to sense a scene about 4-7 meters away the maximum amount of information available through photonic sensors at various bands (visible light, infrared, ultraviolet)

All of these photon sensors are available as photodiodes or phototransistors... Given that we need to amplify each of these with an OpAmp in logarithmic amplification and that we wish to keep that part of the circuit under $3 for 8 channels, do you think we should opt for a design based on photodiodes or phototransistors? (The circuit needs to be physically small)

Because of our cost constraint, my intuition tells me we’re better off with phototransistors in order to use the lower-cost OpAmps, but perhaps they greatly reduce the signal-to-noise ratio and we’re better off with photodiodes? (We need to amplify the signal for a standard ADC attached to a voltage rail of 1.8V and obtain the best possible dynamic range and signal-to-noise ratio)

Any recommendation on a decent type of OpAmp and the amplification topology to use would be greatly appreciated!

Thank you!

Jean-Pierre
 

davenn

Moderator
Sep 5, 2009
14,254
Joined
Sep 5, 2009
Messages
14,254
Hi Jean-Pierre
welcome to Electronics Point

what do you mean by "sense a scene" and "maximum amount of information" thats very vague ?
what are you trying to sense ? just the light level ?
Any sensor is going to have to be very close to the "scene" cuz if its at some distance .... several metres ... then all the ambient light inbetween is going to be picked up by the sensors aka photo diodes/transistors etc
And even if the sensors are close to the "scene" the "scene" is still going to have to be shielded from any ambient light

Dave
 

JPPoulin

Sep 21, 2012
8
Joined
Sep 21, 2012
Messages
8
Hi Dave, thanks for taking the time.

It is for a low-cost fire prevention product our team is developing. The general idea of the product is to place several low-cost sensors (all below $0.10) such as IR, visible light, UV, PIR and temperature (by convection) so hopefully the on-board CPU can determine the presence of fire from the different sensor permutations.

(so it's ok if ambient light is detected by IR & UV... hopefully minimized by compensating with visible light sensor)

Do you think a photodiode-based design is superior to a phototransistor-based one if the total cost of that part of the circuit is to be under $3? (8 channels of amplification & 8 low-cost sensors)

Many thanks!!

JP
 

(*steve*)

¡sǝpodᴉʇuɐ ǝɥʇ ɹɐǝɥd
Moderator
Jan 21, 2010
25,510
Joined
Jan 21, 2010
Messages
25,510
Personally, I'd look at what the cheapest sensors are, and see if I could work with them. If they require a lot of support circuitry, then perhaps look at other sensors that may be slightly more expensive but easier to read.

You'll need more gain with photodiodes than with phototransistors, but that may not be a significant factor. What are the cost differences?
 

JPPoulin

Sep 21, 2012
8
Joined
Sep 21, 2012
Messages
8
Hi Steve, thanks for taking the time...

We're hoping to keep the costs per photon sensor under $0.15 each, hopefully less than $0.10 with a $1.50 or so OpAmp for the 8 channels...

If it wasn't for the cost sensitivity I'd just go with a photodiode-based design but my feeling tells me I can possibly get a lower-cost design that does the job if the currents/voltages fed into the OpAmp are not so tiny like in the case of photodiodes...

A tricky job to optimize performance & cost in this case! :)

Thanks again,
 

davenn

Moderator
Sep 5, 2009
14,254
Joined
Sep 5, 2009
Messages
14,254
we have an almost identical thread here on the same topic within the last week

I suspect you will find as was told to the other person that this is going to be very difficult to achieve in the way you are expecting it to be done
especially for the low costs you are stating

You need to consider the low sensitivity of a standard IR photodiode or transistor, My immediate thought is that the diode or transistor isnt going to pick up enough IR energy till the fire has become significant in size and by that time is well on its way to causing serious damage to the building etc

The 2 common uses for those style of IR sensors are short range ... less than 1cm for opto-choppers, or over a few metres with the well aimed and narrow frequency and beam of a remote control.

They are not going to pick up say a burning match or candle at 3 - 5 metres range. And from my past experience in the fire service a small flame like that can become a whole room engulfed in fire within a few minutes.

Passive IR sensors are going to be the most sensitive and most likely to work, but implimenting their use is going to well and truely blow your budget

Dave
 
Last edited:

JPPoulin

Sep 21, 2012
8
Joined
Sep 21, 2012
Messages
8
Hi Davenn, thanks very much for these useful observations.

I knew that the pickup would drop by the square of the distance but was not aware that they generally are made to pickup over a few centimeters. Henceforth we will use only the sensors with the narrowest beams!

Obviously setting up a quality logarithmic amplification to these is critical and much attention has to be placed on the signal-to-noise ratio of either the photodiode of phototransistor... I wonder what the relative SNR of both are...

However as you stated, infrared remote controls can work over quite a range (even 10 meters or so)... so there is hope here somewhere!

Perhaps using photosensors made for IR remotes would give better results for this app? What do you think? (Obviously we'll study their datasheets and pore over the details)

Other than photodiodes, phototransistors and IR remote control sensors, are there any low-cost sensors (less than $0.25) that you think we should study carefully for this application?

- Should the PIR sensor of motion detectors be included in our study? They seem able to detect small changes in temperature up to 10 meters away...
- Are there low-cost thermopiles that your are aware of? (A quick look reveals they are in the $1.25 range)

Many thanks again for sharing your experience with this!!!

Jean-Pierre
 

BobK

Jan 5, 2010
7,682
Joined
Jan 5, 2010
Messages
7,682
An IR remote works by modulating the signal at the sender. The receiver can then check for the specific modulation frequency and ignore other background that is orders of magnitude higher. Unless you can convice your fire to oscillate at 40KHz, this method is not going to work for you.

The problem is not so much the signal dropping off as other signals swamping it. Amplifying does not help becuase the noise is amplified as well.

Bob
 

JPPoulin

Sep 21, 2012
8
Joined
Sep 21, 2012
Messages
8
Hi BobK, thanks very much for that input.

My thinking there however is that fire is an enormous output of energy and would probably dominate that scene (mostly in the IR band a 4.3nm for carbohydrate-based fires but hopefully enough IR energy exists in the bands that cheap IR sensors excel at).

My thinking is that a low-cost IR photodiode amplified logarithmically by a well-chosen low-noise OpAmp could pickup the difference between a hot stove element 5 meters away versus a hot stove element with a pan on fire on top.

(Please note that the focus here is not necessarily using photodiodes or phototransistors... it is using the best possible sensor that costs below $0.30 amplified by an OpAmp below $1.50 and trying to detect as best as possible this scene given a BOM below $2 for that part of the circuit.

Sensors such as PIR elements, remote control IR sensors, IR photodiodes & phototransistors are all being studied for suitability to this app...

Do you think it can be done? (picking up a candle 10 meters away is probably wishful thinking but we're trying to get there as best as we can under $2 BOM)

Jean-Pierre
 

BobK

Jan 5, 2010
7,682
Joined
Jan 5, 2010
Messages
7,682
Are you sure that a fire is putting out a lot more IR energy than a hot element? After all, you can ignite paper of the hot element easily, so its temp is higher than the flaming paper. I do not know how much IR at what wavelengths either of these would put out, but I would bet that it would take some sophisticated processing to determine the difference.

Bob
 

davenn

Moderator
Sep 5, 2009
14,254
Joined
Sep 5, 2009
Messages
14,254
Hi BobK, thanks very much for that input.

My thinking there however is that fire is an enormous output of energy and would probably dominate that scene (mostly in the IR band a 4.3nm for carbohydrate-based fires but hopefully enough IR energy exists in the bands that cheap IR sensors excel at).

I think you missed the point I was making in my previous post by the time you have a fire is that intense its starting to get to the stage of out of control

My thinking is that a low-cost IR photodiode amplified logarithmically by a well-chosen low-noise OpAmp could pickup the difference between a hot stove element 5 meters away versus a hot stove element with a pan on fire on top.

do some experiments and try it but I think you are going to need PIR's not photo diodes/transistors, they are just way too insensitive

(Please note that the focus here is not necessarily using photodiodes or phototransistors... it is using the best possible sensor that costs below $0.30 amplified by an OpAmp below $1.50 and trying to detect as best as possible this scene given a BOM below $2 for that part of the circuit.

Sensors such as PIR elements, remote control IR sensors, IR photodiodes & phototransistors are all being studied for suitability to this app...

I have spent the last hour going through dozens of www sites to do with fire and flame detection for you. You are going to have to increase your budget / detector substantially
At least $10 -20 / completed detector unit at minimum. Just the case to house the sensor in is likely to cost you $5 - 10 for a start, that is before you do PCB,s and components to put inside. The more you try to skimp on cost, the more likely it is not to produce the results you desire

Do you think it can be done? (picking up a candle 10 meters away is probably wishful thinking but we're trying to get there as best as we can under $2 BOM)
Jean-Pierre

the way envision probably not. Sorry to sound so negative, but you need to be realistic
there is going to be significant signal processing both at the detector and at the receiver end where all the detector signals are sent to.
This is why it costs many many $1000's for commercial systems for flame and smoke detection. All the flame detection units I saw used 2 PIR's per sensor with electronics to do comparisons between the 2 sensors signals

And speaking of smoke detection, this is probably your easiest and cheapest way to detect a fire. Home smoke detectors are ~ $10 each and it would be dead simple to do a cable from the output back to a central control computer system.

cheers
Dave
 
Last edited:

JPPoulin

Sep 21, 2012
8
Joined
Sep 21, 2012
Messages
8
Hi Bob, no I'm not sure but as our product has sensors of different bands the hope is that we'll be able to tell the difference by repeated observance of both situations.

They key problem we're trying to solve at this point in time is how to connect <some type of IR sensor below $0.30> with the best possible OpAmp under $2 using the best possible logarithmic amplification that is calibrated such that a hot stove element 1 meter away from the sensor = 1.8V at our CPU.

Lots to study!! :)
 

JPPoulin

Sep 21, 2012
8
Joined
Sep 21, 2012
Messages
8
I have spent the last hour going through dozens of www sites to do with fire and flame detection for you. You are going to have to increase your budget / detector substantially
At least $10 -20 / completed detector unit at minimum. Just the case to house the sensor in is likely to cost you $5 - 10 for a start, that is before you do PCB,s and components to put inside. The more you try to skimp on cost, the more likely it is not to produce the results you desire

the way envision probably not. Sorry to sound so negative, but you need to be realistic
there is going to be significant signal processing both at the detector and at the receiver end where all the detector signals are sent to.
This is why it costs many many $1000's for commercial systems for flame and smoke detection. All the flame detection units I saw used 2 PIR's per sensor with electronics to do comparisons between the 2 sensors signals

Hi Davenn, thanks very much for sharing that info with us! Very useful observations..

I really like the idea of switching to PIR instead of photodiodes.... these are getting dirt cheap now... I research this direction now and try to get a bearing on their input vs. output ratio in the IR bands.

This really is a challenging circuit but we're committed to figuring this stuff out at the lowest possible cost. (We'll sell a higher cost design with the right sensors soon after...)

I am intrigued at your suggestion that a cage would be needed. If I use the 2 PIR approach, do you think the cage can be omitted? (in other words, what is it shielding against? Photons coming from areas of non-interest?)

Thanks again!!

Jean-Pierre
 
Top