chk my drawing please

hi would the following circuit work? the idea is my robot will fire a RED 5mW Red laser at say a window
the reflected light will hit a RED LED array inside half a ping pong ball (or something simillar) i want the circuit to take this signal and amplify it so the sound can be heard through the speaker its a spy mic for my robot so the question is will this circuit do what i want??
many thanks LG
View attachment 63933


sorry some the lines not exactly straight but i had trouble with the snap to grid function i would also like comments on how i could have drawn it better (tips)
forgot to add the output will also be going to the ADC of one the pics as the outside of the bot will have SMD Leds the idea is when the mic is in use the leds will flash according to output of the mic and hopefuly this will get me extra points in the style catagory

I'm not that familiar with op-amps, so some things that look strange to me may be perfectly normal, but here goes:
C3 and C4 - why do you have 2 caps here? You only need one.
Given that you are amplifying quite a small signal, I wonder if you could find more appropriate coupling caps than electrolytics. Do they need to be 10μF or can you use something smaller?
Someone here could probably tell you how to work out what size caps you need - you need to be looking at the impedance at the frequencies you are interested in, compared to the impedance of what each one is connected to. It's significant because given your application, you should probably limit the bandwidth of this amp, possibly to something lower than a telephone's bandwidth. The laser will pick up all kinds of sounds you don't want to hear, so you need to limit these as much as possible.
You should probably have more stages, and feedback over the entire amp to improve the noise ratio.
Again, I'm not familiar with op-amps, but should you have a single bias network supplying both op-amps? (R3 R4) I thought they would each need their own.
What does R6 do?
I would anticipate a lot of distortion and some non-linear response in the system you describe. I would anticipate need for some kind of filter and adjusting the response of the amp. The only way you are going to find out the response curve you need is by experimentation.
When you say LED array - does this mean photo-diodes, which wouldn't be LED's at all? Only, the way you describe, LED doesn't make any sense, but photo-diodes do.

I would guess you need some kind of impedance matching between the array and the input - or is this the purpose of the first op-amp? You also need to take account of the biasing needs of your sensor. Photo-transistors would be better since they have gain.

Have you considered the angle at which the beam hits the window? Anything other than dead straight on and the beam is going to be reflected at an angle, and you are going to have to position the sensor to take account of this.

Use snap-to-grid to get the basic layout, then turn it off to tidy up the drawing. Looks like you've got some auto-routing there - see if you can turn it off when it gets a nuisance. I'd draw either RV1 under U1:B, or put the + - symbols on U1:B the other way up - you avoid the line crossing over that way.

Who are you spying on?

I think you will need much greater gain than your present circuit provides, especially if you are using LEDs as photo-sensors. The first opamp could have higher gain than unity.
I agree with the above re the coupling caps. 1uF, non-polarised, would be better. C3/C4 could be a single 1uF cap.
The LM358N won't be able to drive a speaker directly. You need a power amp stage.
The LEDs are going to be swamped with ambient light. Have you considered using pulse-modulated laser light to overcome this?

I never heard of using LED's that way. I though the OP made a typo!

The 10uF capacitor C2 feeding the 100k input resistor R2 of opamp U1A has a cutoff frequency of only 0.16Hz!
A non-polarized film capacitor with a value of 0.1uF (100nF) is more suitable and produces a reasonable cutoff frequency of 16Hz.

Why are the opamps inverting? Then they have a low input impedance (the first opamp will have a low input impedance when its inverting gain is increased by reducing the value of R2) which requires huge capacitor values.
The two 10uF polarized capacitors feeding the second opamp could be a small non-polarized film capacitor if the second opamp is non-inverting.

I agree than an opamp cannot drive a speaker. A speaker is almost a dead short to the output of an opamp.

hi
wow plenty of advice!! the drawing was taken from a velleman preamp kit and is not my own well at least i now know it isnt suitable as it stands. i will go away and do the changes sugested and hopefuly that will make it work
thank you all for the help
lg

View attachment 63950sorry to trouble you all again velleman do another kit that might be better or could be used with the modified first one could you please take a look at this one and see what you think many thanks

Look at the datasheet for the TDA2003 power amplifier IC which is what Velleman copied.
You did not post the parts values that determine its gain but the circuit in the datasheet has a gain of 100 which will not be enough.

Your laser is shining on a mirror. Its reflection might or might not shine on your LED array. Voices in the room are vibrating the mirror slightly which varies the amplitude of the reflection slightly. If the pickup is good then the signal will be tiny and will need a lot of amplification.

thank you very much for that mr audioguru i will have to read up more on opamps etc can you please give me an idea of how much gain i will need? so i can try and draw a suitable circuit many thanks logan

sorry i just did a screen shot of the last drawing i should have looked harder! i have added the values to this one i had a look at the ST datasheet and the values seem different so hopefuly its better

View attachment 63961

The gain of the Velleman amplifier is 1+ (R1/R2)= 101 times.

The required gain is determined by many things:
Power output of your laser. Reflection ratio of the mirror. Stiffness of the mirror and calculations showing how sounds modulate it and produce amplitude modulation of a carrier wave that you are missing. Sensitivity of your LEDs used as photo-diodes.
Loudness of the speech. Required loudness at the speaker.

thank you i better do some trying out!!! many thanks for the information lg

I don't think you should be aiming to build an "all in one" amplifier for this. You'd be better to concentrate your design effort on a pre-amp stage where you can control noise, bandwidth, and so forth, since that's the hard part, which will give you a more robust, albeit still small, signal which you can plug into a more general purpose amp where you can add lots of gain and control the volume. That said, it doesn't mean the two (or possibly more) stages can't go in the same box and share a psu.

It looks like you are also designing blind here - you don't know how strong or clean the signal is that you are wanting to amplify, so probably no-one is going to venture a gain figure for you. This is something you will only find out by experimentation.

Suggest you look at other pre-amps for small signals such as those used for coil microphones, phono pickups, tape players etc, to get some ideas. Tape and phono both have filters to shape the response, which might make the circuits look a bit strange to you, but they are designed to pick up tiny signals.

ok thank you for the tips!! i am trying to get the laser working so i can get an idea of the signal. i realy do appreciate all the help
lg

hmmm... I already developped some tests for our group's magazine, based on a single LED tied to an AVR microcontroller, to be used as an ambient light sensor. The same LED that iluminate the corridor as a "night light" also serve to measure the ambient light. The LED is connected in reverse polarity for few milliseconds, it will charge its "internal capacitor", then this "capacitor" is read by the microcontroller. If after few milisecond the charged voltage is there, then it is dark enough. Upon light over the LED junction, this "capacitor" discharges. The curve of discharge is proportional to the incident light over the LED. If it is dark the uC will reverse the LED polarity and feed current through it for some time, half a second or more.

As far as I can remember, I was never able to "read" lumens hitting a LED when it is directly polarized, as in the schematic posted, katode to ground...

For the project in subject I would NOT use a LED as sensor, but a very sensible cadmio cell. Even photo-diodes or photo-transistores are not the best choice for such subtle light variation gathered from hundreds of yards of a reflected laser upon a non-mirrored window glass or any other reflective object close to a conversation.

To eliminate ambient noise, mostly street noise, that will vibrate much more the window glass than the conversation inside, you will need to sample another window glass, a little bit away (few meters) from the conversation room. If possible, the sample window room should be empty. The idea is that both "laser microphones" will capture the same street noise, but just one will capture the inside conversation. The sample signal could be subtracted from the conversation signal, and you might have a clear voice with street noise reduced. If the sample signal is captured from a far away window, you will have different street noises waveforms, different times and reflections, and will not have a good noise subtraction result.

If you can read Portuguese, or use a translation, read my article about LED as a sensor at our Dec/2009 magazine:

View attachment 64000

The following circuit was found somewhere in the net, serves as a reference if wanting to use transistores as the pre-amplifier. It should work, not sure for the faint laser reception you are trying to do.

View attachment 64001

ok thank you very much i will use translate and look at your page i will also build the circuit and try it out

i had a look at the magazine but couldnt get google translate working with it looks like a very very good magazine i will keep trying to get it translated