IR LED and Phototransistor circuit: I can't get any range
Sarah Kamal
I am designing object avoidance on a self-controlled model car for a project
course. I am using IR LED's and phototransistors to sense up coming objects.
Currently I am only getting a range of 2 inches (the distance of the object
that the IR bounces off of). I need the distance to be at least 2 feet so
that the car has time to slow down and stop before hitting the object. I was
trying to boost the 'brightness' of the LED by pulsing higher current through
it (according to specs).
Can anyone recommend a specific brand of LED and phototransistor that might
work well for this purpose.
Can anyone recommend a different approach to object sensing if this one seems
out to lunch.
Any input would be appreciated.
Thanks
Sarah Kamal
Gareth Harris
Sarah Kamal (ska...@pop.srv.ualberta.ca) wrote:
: I am designing object avoidance on a self-controlled model car for a project
: Currently I am only getting a range of 2 inches (the distance of the object
: that the IR bounces off of). I need the distance to be at least 2 feet so
: that the car has time to slow down and stop before hitting the object. I was
: trying to boost the 'brightness' of the LED by pulsing higher current through
: it (according to specs).
: Can anyone recommend a different approach to object sensing.
I think the problem is that there are many sources of IR (sun, lights, ect.)
other than your LEDs which make it difficult to detect the signal you
want. The best way around this is to modulate the IR output (40KHz is
the usual frequency for this I think) and filter the received signal so
that everything except 40KHz is ingnored. This will allow you to have a
much more sensitive detector circuit, which could give you a range
similar to that of IR remote controls.
Steve J. Noll
>I think the problem is that there are many sources of IR (sun, lights, ect.)
>other than your LEDs which make it difficult to detect the signal you
>want. The best way around this is to modulate the IR output (40KHz is
>the usual frequency for this I think) and filter the received signal so
>that everything except 40KHz is ingnored. This will allow you to have a
>much more sensitive detector circuit, which could give you a range
>similar to that of IR remote controls.
And you can get a TV remote-control receiver module to do this with
from Radio Shack.
Steve J. Noll | Ventura CA | WA6EJO | sjn...@ix.netcom.com
| http://www.netcom.com/~sjnoll/usedequip.html
| http://www.netcom.com/~sjnoll/peltier.html
Richard Torrens
In article <5g2tbq$p...@pulp.ucs.ualberta.ca>, ska...@pop.srv.ualberta.ca
(Sarah Kamal) wrote:
>
> I am designing object avoidance on a self-controlled model car for a
> project
> course. I am using IR LED's and phototransistors to sense up coming
objects.
> Currently I am only getting a range of 2 inches (the distance of the
> object that the IR bounces off of). I need the distance to be at least
> 2 feet so that the car has time to slow down and stop before hitting
> the object. I was trying to boost the 'brightness' of the LED by pulsing
> higher current through it (according to specs).
> Can anyone recommend a specific brand of LED and phototransistor that
> might work well for this purpose.
Long range photo-sensors also need good optical systems: you need to focus
both transmitter and receiver into a tight beam.
This probably means a single good (compound) lens wth a half silvered mirror
to split transmitted and received signals.
--
/| Richard Torrens
/ |
/ | 4 Q D - 4...@argonet.co.uk
/ | |
/ /| | We manufacture
/ / | | MOSFET controllers for battery operated motors
/_/__| |____
/_____ ____\ See us on http://www.argonet.co.uk/users/4qd
/ _ \| | _ \ our www site also contains FAQ sheet on motors & controllers
| | | | | | | | and a selection of interesting circuit diagrams
| |_| | | |_| |
\__\_\ |____/ We use an Acorn StrongARM RISC-PC 32 bit computer.
Ali Mehdizadeh
Sarah Kamal (ska...@pop.srv.ualberta.ca) wrote:
: I am designing object avoidance on a self-controlled model car for a project
: course. I am using IR LED's and phototransistors to sense up coming objects.
On a side note, would it be possible to post the circuit you are using?
I am trying to build a similar circuit using IR led/phototransistor
pair. The sensor pair comes in a prepackaged form with each element
installed in a angle. It is supposed to have a range of about 5mm, which
is fine for what I am looking for.
I know the phototransistor is supposed to induce an Ib current through
the base when it rececives IR emmsions from the LED. I just can't figure
out how to hook it up to give me a voltage out put depending on the
amount of IR received.
Thanks,
-Ali M.
Tom & Mary
Sarah Kamal wrote:
>
> I am designing object avoidance on a self-controlled model car for a project
> course. I am using IR LED's and phototransistors to sense up coming objects.
> that the IR bounces off of). I need the distance to be at least 2 feet so
> that the car has time to slow down and stop before hitting the object. I was
> trying to boost the 'brightness' of the LED by pulsing higher current through
> it (according to specs).
>
> Can anyone recommend a specific brand of LED and phototransistor that might
> work well for this purpose.
>
> out to lunch.
>
> Any input would be appreciated.
>
> Thanks
> Sarah Kamal
Beleive me.. You really want to use Ultrasonic methods for this
application. IR can be very tricky when trying to get it to "bounce"
off other objects and then detect it. Just the simple nature of IR
limits the use specifically for such a application as yours.
(interference for one, and tight ir beam-forming for two...)
Paul Mathews
Generally pretty good advice from Luminax1. I've added a few comments
interleaved:
Luminax1 wrote:
>
> Sarah:
> The first thing you want to do is get hold of Siemens Opto Data Book.
Siemens Opto, Cupertino, CA USA phone 408 257 7910
Plenty of other good sources of literature, including: HP, Quality
Technologies, Hamamatsu, Sharp, IPI Centronic, etc.
> Their manual has all sorts of app. notes in the back with suggested
> circuitry. I read an earlier posting of yours regarding operating IR leds
> at high peak pulse power. In most cases the led is dissipation limited by
> the size of the die and the thermal resistance of the leads. If a
> particular led is rated for 50 mA cont. (d.c.) then at a duty cycle of 1%
> you should be able to pulse the device at 5A. Of course nothing is that
> easy ... at high drive currents the forward voltage drop of the led
> increases, so that the dissipation of the device also increases. The
> limit in peak drive current is also a function of the very fine gold wire
> that is used to bond to the die. This wire is only a few hundred microns
> in dia. and can be fused at higher currents.
Too true. Most LEDs don't like to be pulsed at any duty factor with
peak currents higher than about 1 Amp.
The other reason you want
> to pulse the led(s) is because when operated at a few KHz the reflected
> light can more easily detected. Instead of a d.c. coupled amplifier you
> can use a tuned a.c. detector with much better gain and stability ....
> that will ignore ambient light.
Many sensors use some sort of matched filtering instead of tuning. One
crude way to achieve this: sample amplifier output immediate before LED
pulse. Sample amp output during LED pulse. Subtract sample pair.
Difference estimates pulse amplitude, independent of ambient light.
This is easy to do with either analog or digital circuitry.
To help eliminate the saturation effect
> of bright ambient light on the i.r. detector, use one of the filtered
> devices ... they come in a "black" plastic housing.
Yes, the black housing helps considerably. However, it also helps to
bias the detector such that it is not easily saturated. Also, use the
narrowest practical field of view (included angle). Unfortunately, as
you narrow the FOV, you'll have to be more careful about aligning the
beam produced by the LED with the FOV of the detector.
Also note that there
> are two common wavelengths for i.r. leds and detectors ... 880 and 940 nm.
> be sure you use matched pairs.
This is seldom of much importance. Virtually all Si photodetectors have
peak sensitivity around 880 nm, which is also where the highest
efficiency emitters are found. So, use an 880nm emitter and choose the
detector based on package and FOV. For photodiodes, lens area pretty
much determines available signal. For phototransistors, transistor
current gain comes into play. Have a look at Jerald Graeme's book
'Photodiode Amplifiers' for details of preamplification.
Generally, several i.r. leds are
> connected in series and are pulsed by a reasonably capable transistor.
> The series connection eliminates the wasted energy associated with large
> ballast resistors, and it puts more light on the subject ...pun intended.
This is an easy way to increase range a bit. However, range increases
will only go as the square root of the number of LEDs.
> You should have no trouble getting at least 1 meter range for your
> obstacle detector ... unless the obstacle is black!
If the sensor need only operate indoors, I agree. Outdoor operation is
a bit more of a challenge. Also, beware of susceptibility to newer
fluorescent lighting fixtures, which use switching power supplies
operating near 40kHz! (another reason to limit detector FOV).
>
>
> Good Luck!
>
>
> Peter
--
Paul Mathews, consulting engineer
AEngineering Co.
3300 S Fox Spit Rd.
Langley, WA 98260-8010
email: opt...@whidbey.com
fax or phone: 360 730 2058
non-contact sensing and optoelectronics specialists
Luminax1
Sarah:
The first thing you want to do is get hold of Siemens Opto Data Book.
circuitry. I read an earlier posting of yours regarding operating IR leds
at high peak pulse power. In most cases the led is dissipation limited by
the size of the die and the thermal resistance of the leads. If a
particular led is rated for 50 mA cont. (d.c.) then at a duty cycle of 1%
you should be able to pulse the device at 5A. Of course nothing is that
easy ... at high drive currents the forward voltage drop of the led
increases, so that the dissipation of the device also increases. The
limit in peak drive current is also a function of the very fine gold wire
that is used to bond to the die. This wire is only a few hundred microns
to pulse the led(s) is because when operated at a few KHz the reflected
light can more easily detected. Instead of a d.c. coupled amplifier you
can use a tuned a.c. detector with much better gain and stability ....
of bright ambient light on the i.r. detector, use one of the filtered
are two common wavelengths for i.r. leds and detectors ... 880 and 940 nm.
connected in series and are pulsed by a reasonably capable transistor.
The series connection eliminates the wasted energy associated with large
ballast resistors, and it puts more light on the subject ...pun intended.
obstacle detector ... unless the obstacle is black!
Peter
John Woodgate
In article <5g2tbq$p...@pulp.ucs.ualberta.ca>, Sarah Kamal
<ska...@pop.srv.ualberta.ca> writes
>I am designing object avoidance on a self-controlled model car for a project
>course. I am using IR LED's and phototransistors to sense up coming objects.
>Currently I am only getting a range of 2 inches (the distance of the object
>that the IR bounces off of). I need the distance to be at least 2 feet so
>that the car has time to slow down and stop before hitting the object. I was
>trying to boost the 'brightness' of the LED by pulsing higher current through
>it (according to specs).
>
>Can anyone recommend a specific brand of LED and phototransistor that might
>work well for this purpose.
>
>Can anyone recommend a different approach to object sensing if this one seems
>out to lunch.
>
>Any input would be appreciated.
>
>Thanks
>Sarah Kamal
>
what current you need to get it. But phototransistor specs are not all
written in terms of the same units, and there can be *great* differences in
sensitivity. For example, TIL81 seems to be much more sensitive than
MEL12. Many phototransistors respond also to visible light, so an IR filter
may help. Also, a small parabolic reflector around the LED, or a conveex
lens in front, can help. TV remote controls use two to four LEDS and
cordless headphone transmiters use six or eight. You can connect them in
series to save current drain on your battery.
--
Regards, John Woodgate Tel. +44 (0)1268 747839
Fax +44 (0)1268 777124. OOO - Own Opinions Only
Marty Vona
> I am using IR LED's and phototransistors to sense up coming objects.
> Currently I am only getting a range of 2 inches (the distance of the object
> that the IR bounces off of). I need the distance to be at least 2 feet
I don't have much expericence with these things, but I've heard that
several inches is pretty much max range for IR. Please, someone
correct me if they know differently.
So, if you want several feet, why not get into ultrasound? I work in a
lab where they use Polaroid sonar modules all the time for this sort of
thing. I think they get them direct from Polaroid. I have also seen
many projects for building these sorts of ultrasound devices - look at
"Robotics Builder's Bonanza" by Gordon McComb (TAB Books).
Hope this helps,
Marty Vona
Timothy A. Seufert
>I am designing object avoidance on a self-controlled model car for a project
>course. I am using IR LED's and phototransistors to sense up coming objects.
>Currently I am only getting a range of 2 inches (the distance of the object
>that the IR bounces off of). I need the distance to be at least 2 feet so
>that the car has time to slow down and stop before hitting the object. I was
>trying to boost the 'brightness' of the LED by pulsing higher current through
>it (according to specs).
>
>Can anyone recommend a specific brand of LED and phototransistor that might
>work well for this purpose.
>
>Can anyone recommend a different approach to object sensing if this one seems
>out to lunch.
Try an IR remote control receiver, the kind with a photodiode and complete
amplifier in a small square metal can, designed for use in VCRs and the
like. The ones I've used are from Jameco. They require you to pulse the
IR emitter at some frequency (for filtering of ambient IR noise); for
mine, that was 32.768 KHz. Although the normal signal output is digital,
you can disassemble the bottom of the can and get at the IC on the tiny
circuit board. One of the pins will be an analog output from an amplifier
stage. For mine, this output a voltage between ~1.5 and ~2.5 V, depending
on the signal strength.
Put a 10K trimpot in series with your emitter and play with the setting
until you get the desired range. These remote control receivers are VERY
sensitive and will not need much brightness from the LED to register a
signal, so you don't even have to worry about overdriving them. (I had to
use a 4.7K resistor to reduce the effective range to a few inches, because
my application was a wall-following robot.)
--
-Tim Seufert, bwa...@cats.uc*sc.edu
The * is to fool automated email address grabbers. Remove it if you
wish to send me email. No unsolicited commercial junk mail!
John Woodgate
In article <332750...@dartmouth.edu>, Marty Vona
<marty...@dartmouth.edu> writes
>> I am using IR LED's and phototransistors to sense up coming objects.
>> Currently I am only getting a range of 2 inches (the distance of the object
>> that the IR bounces off of). I need the distance to be at least 2 feet
>
>several inches is pretty much max range for IR. Please, someone
>correct me if they know differently.
>
>So, if you want several feet, why not get into ultrasound? I work in a
>lab where they use Polaroid sonar modules all the time for this sort of
>thing. I think they get them direct from Polaroid. I have also seen
>many projects for building these sorts of ultrasound devices - look at
>"Robotics Builder's Bonanza" by Gordon McComb (TAB Books).
>
>Hope this helps,
>Marty Vona
Infra-red TV and VCR remote controls have a range far exceeding 'several
inches' and they are designed with cost as a priority.
Bruce Nepple
Marty Vona <marty...@dartmouth.edu> wrote in article
<332750...@dartmouth.edu>...
> > I am using IR LED's and phototransistors to sense up coming objects.
> > Currently I am only getting a range of 2 inches (the distance of the
object
> > that the IR bounces off of). I need the distance to be at least 2 feet
>
Microchip has an application note using the Polaroid Ultrasonic Sensor
module as a rangefinder. Check their website (http://www.microchip.com).
You might get better results than with IR, plus knowing something about the
distance may be usefull.
For IR sensing, I'm real impressed with the Sharp monolithic IR sensor
(ISU601 or something close to that) which is design for TV remote control
systems. It senses 38KHz pulsed (modulated) IR and is very cheap. I've
found it to be very sensitive, especially to leading edges. It contains
photo detector, AGC, 38Khz filter, and slicer in one 3 lead package. (IR
in, Data out) I doubt if you could get *much* better performance with a
design from scratch. The 38 KHz passband filter is critical. The filter
is what really gives you the sensitivity. By limiting the noise bandwidth,
you can detect very weak signals. With a DC based (unmodulated) system you
have to tweak out all of the DC offsets in order to get high enough gain,
and then you're detecting all the IR in the room. I don't expect there is
much IR junk around 38 KHz.
An added advantage (disadvantage) is that you can use a remote control to
affect the vehicle.
Bruce Nepple
Remove nospam to email
/***************************************************************************
**********/
!!!!FREE E-MAIL TESTING!!!
Please send Usolicited Commercial email in order to receive a free 1000
message email test consisting of 1000 random length messages from 1000
different addresses.
/***************************************************************************
**********/
Ann
Tim Seufert wrote:
> Try an IR remote control receiver, the kind with a photodiode and complete
> amplifier in a small square metal can, designed for use in VCRs and the
> like.
Have IR devices replaced ultrasound devices for these types of
applications?
If so why?
brian whatcott
In article <1XaOveAH...@jmwa.demon.co.uk>,
j...@jmwa.demon.co.uk says...
>
>Infra-red TV and VCR remote controls have a range far exceeding 'several
>inches' and they are designed with cost as a priority.
>--
>Regards, John Woodgate
The radar equation underlines the large increase in power required to reflect
enough signal off a random diffusing target to detect at the source vs that
required for a detectible signal at the target.
A sonar module isn't such a bad idea.
brian whatcott <in...@intellisys.net>
Altus OK
Sylvan Butler
IR has been used by ALL of the mass market consumer items. A few early
or niche products used ultrasound, but it just didn't work well and the
transducers are way too expensive.
sdb