Sorry your browser is not supported!

Recently I have been getting pretty good at circuit design and implementation (I'm a sophomore majoring in EE).

So, if anyone has some device (preferably analog so that I don't have to spend too much time designing some enormous circuit) that you have wanted to have designed in the past but never knew how to or never got around to designing, let me know what it is ,and I will test my skills.

It will probably push me to look for and learn something that I wouldn't have.

Analog you say? Here's a challenge:

Challenge

Build a simple infrared transmitter and receiver. You don't have to be able to transmit any data, all the receiver has to do is be able to tell you if the transmitter is turned on or off.

Guidelines

-Transmitter range in sunlight : 2 meters
-Transmitter range indoors : 4 meters
-Transmitter range in complete darkness : 8 meters.
-Both circuits must be powered by a 9V battery so they are portable.
-If the receiver detects the transmitter, an LED should glow. If the transmitter isn't detected, the LED should turn off.

Additional challenges

-Receiver should not react to any TV remote controls.
-Receiver should not react to neon lights.
-Transmitter should have a range of 12 meters in normal indoor conditions.

Good luck and have fun

TheComet

That's a good one. I'll need to find the right part that will generate the right frequency of light and spec it out along with a light sensing part.

I have a little experience with light sensing, but I used an ardunio which I don't intend to use this time.

Well we'll see what I use. Thanks!

Edit:
So I'm thinking, its kind of hard to make a transmitter if you don't know what your transmitting. If you have something in mind, let me know. If you don't care I'll just pick music.

That's a nice idea Initially I said it didn't have to transmit anything except for the carrier signal, but if you feel you can extend it, go for it

TheComet

Well, I have some plans down, but I need a few parts like an op amp and a diode and phototransistor that work in the voltage range I need.

I'm searching for those now.

You have me confused with this statement. Why do you think that digital would mean building a larger circuit??? To me it doesn't have that impact, deciding between analog and digital IC's would depend on the application.... I can see not wanting to deal with A/D or D/A converters... or perhaps not wanting to get involved in a project that is best suited for a microcontroller (if you don't have them) to save on cost and size (if you only have analog components, taking on such a project would be a heck of a task),, but it's not the digital aspect that makes the circuit enormous in and of itself. Most digital IC's are fairly cheap,, and simple microcontroller like arduinos are aslo relatively cheap in comparison to analog components.

So why limit it, and rule out Digital? Unless of course you only have a multimeter, and not even a single trace o-scope? And if the latter is the case... your project should be to build a simple single-trace virtual oscilloscope/interface with your computer. Especially if you have a desktop with a serial port or parralel port. Yeah you won't get insane timing... but you can at least build something that can get you into digital.

The programming aspect is not to rough especially interfacing with the serial port and C/C++. And there are plenty of plans and access to source code on-line, even one to turn a gamneboy into simple o-scope. Don't know what your experience programming wise is though

These days it's cheaper to do everything with a micro controller, and most of the time it's also simpler. Here's one for just 3 bucks:

https://www.distrelec.ch/mikrocontroller-8-bit-so-14/microchip/pic16f1825-e-sl/660241

If you were to use that for coding/decoding for the transmitter and receiver, you'd be much more flexible in changing the behavior of your circuit. Just a thought.

TheComet

Yes zenassem, you're exactly right. I could do a digital design. However, it does usually lead to buying a relatively expensive ($50) board to program the chip. And if you are going to do something that isn't totally useless, you probably will need either a large circuit or a good amount of A/D conversion.

Edit: Oh, and yes I have some experience programming those. Please read what I put at the end.

So, ya I could do some digital, and I like digital, but I just didn't want it to get too big and consume too much time.

Still, go ahead and let me know if you have an idea.

@ TheComet I have used a similar chip to do my A/D conversions and to do various tasks but it won't really challenge me in the way I wanted it to (another reason for saying AC). You can do anything using micro controllers, but I am more of testing myself to be more technical and to thing more old school.

Digital is fine, but I'd rather not use a microcontroller. I'd rather it be something I can do with logic gates and/or multiplexers if it is to be digital.

How about a drum machine that listens to the gaps between you playing say a keyboard, or guitar, and then the drum machine attempts to play along with you after it has timed your rhythm?

@Pincho Paxton well that seems to be better done by a microcontroller, and besides, I know little of drumming. However, I think I could do something like that, but it would be pretty limited and much better and easier with a microcontroller, which I'd prefer not to use because then I will spend most of my time programming and not making circuits, and I don't doubt my programming skills.

Make a computer, powerful enough to take over the world??

Nah just kidding. How about something that takes an audio signal and changes different aspects of the sound wave such as frequency, amplitude, and wave form? Well amplitude would be easy but seeing a frequency or especially a waveform change would be interesting! Not sure how a waveform change would work!

Just a heads-up on driving the transmitter IR LEDs. My own experiments and research conclude you don't need a resistor in series with the LED. TV remote controls use the same technique. This is what I used:



TheComet

Oh that's why they do that! Saves money to have one less unique component on the pick and place machine apparently!

Well, actually I think they put those in to limit current and to obtain a more ideal performance from their power supply.

My internet has been on and off so I haven't been able to post.

Anyway, I simulated my circuits, and they work fine.

I have not included that LED that lets you know if the devices are connected yet, but I can, and I know how I am going to add it to the circuit.

If anyone cares for an explanation of the circuits, let me know.
The top one is the receiver. The bottom one is the sender with the 50k acting as the diode and resistor.

now you can attach any speaker or audio altering device to the output. So, next I will design what dbd79 suggested.

Of course, what the point if I don't build it? So, I will, but I need parts. I'm currently trying to find some parts which operate in a reasonable voltage range. (0-4.5V)

Once I do that, I just need to purchase the parts and build the thing.

Hmm, I wouldn't mind an explanation.

TheComet

Precisely! That's what i meant, that way the pick and place machine doesnt have to put a resistor and LED, thus saving the number of unique components.

Awesome! Im excited to see how that works!

Internet is up for a short time. I'll make the most of it.

So, I've been thinking. altering frequency isn't possible with an analog device like the ones I've been making. You would need to store the signal in order to manipulate it in anyway that involves changing frequency.

That doesn't mean other things cannot be done.

For one thing, I can try to make an equalizer. I've done that once already, but I could make it better.

I'm thinking about things I could do to the waveform. I could make the signal closer to a square wave (possibly a broken or robotic sound) by ramming the signal into the rails of an opamp.

I could use a capacitor to put a limit on the slew rate of a circuits output therefore making a sort of triangle wave, but that would probably make the audio signal totally unrecognizable. However, maybe if I use a diode or 2 I could make different slew rates for different changes in voltage (positive verses negative). Maybe that would make the audio recognizable.

I don't know, but I'll simulate it, and when I get the parts I'll build it.

I could also make a garbled sound by changing the amplitude of the output rapidly, I think.

Ok, I'm going to post and then edit this.

Edit: An Explanation of the circuits:

Starting with the bottom circuit, V2 is intended to represent an audio input signal to the circuit coming from an mp3 player.

The voltage (or potential) at the right of the opamp will be -2 times the voltage at the positive node of V2.

So, the voltage at the positive end of V2 will vary between + and -1 volt. One might think that you can simply amplify this signal and send it through the LED but that won't work. First of all, the receiver would not be able to tell the difference between a positive voltage and a negative one. Also, LED's are one way; they don't work with negative voltages. So, ya, we need a positive signal at all times.

So, I know that the max voltage will be - or + 4.5 V so I set + 4.5 to be the voltage at one end of the diode ( the 50k resistor). This way the voltage across the diode will always be positive.

The power supply is the 9 volt battery V3 connected to those 2 resistors which divide the voltage so that I can have + and - voltages relative to some node ( the ground node). Ideally I want those resistances to be zero but the need to be something, so its not a perfect supply, but hey, what is? Seriously, what is? Nothing!

As long as most of the other resistors are a lot more than 1k (or whatever I decide to make them) the circuit will work fine.


The receiver is the more interesting circuit. It needs to take a signal which will come through as a purely positive voltage and turn it into a positive and negative voltage. The interesting part is choosing an ideal 0 voltage, because this voltage will change depending on how far away the receiver is from the sender.

So what I did is simply amplify the signal the same way I did before except I now use the large capacitor C1 and large resistor R11 to slowly take on the same voltage as the output. Because it is so slow at doing this, it will not change when there is a high voltage or change when there is a low voltage. It will instead slowly find the middle between the high and low voltages.

Because the capacitor holds our "average voltage" our output is the voltage difference between our amplified input and our C11 voltage.

Note, when I said slowly I mean like 1 10th or 1 60th of a second which is slow compared to the period of an audio signal in the audible range.

BTW, TheComet, what is that picture supposed to be? I don't know the device near the bottom there.

I thought you might mean something like this from your message.


Not be, well, a smart@$$ but I figured I'd mention that the diode there must is probably a silicon diode.

I've just heard a lot about this stuff so I notice it impulsively.

One can guess by seeing the strange choice in resistance which indicates that they want a clean current number like 1mA which causes the voltage across the diode to be .7V which is typical for a silicon diode when forward biased.

Thanks for the explanation I've studied your circuit and will give you feedback.

You have a good grasp on the basics (Resistors, capacitors, transistors, diodes, LEDs etc.), I see you have some extensive knowledge about OPAMPs, and you've also had some experience with alternating currents.

Theoretically your circuit does work, however if you were to build it physically, it wouldn't work at all. There are a few reasons as to why, here they are:

1) You represent your IR LED as a 50k resistor, however the LED has a dropoff voltage of ~2V. This causes the bottom 2V of the signal that comes out of the OPAMP to disappear.

2) Let's say you fixed the problem above by shifting the signal up by another 2V. The signal you transmit is no longer cut off, but it's going to be so weak, it will probably not even be able to travel a millimeter.

Something I also don't fully understand is how your receiver works. You need some kind of sensor to receive the signal (most likely a phototransistor), and you'd need to filter out the daylight and other disturbances.

I'll disregard that however, as you've represented that part as a frequency generator. Even so, I don't quite see how you then convert that received signal to something that one can hear?

So now to solve the problems.

As mentioned before, the sending distance isn't going to be far. The way to fix this is to create some kind of modulated signal so we can pulse the LED. LEDs are cool in that way. Let's say the LED has a maximum continuous current of 100 mA. It's actually possible to burst a very high current through the LED for a short period of time. For example, the above LED can handle 1000 mA, but you must keep it turned off for a tenth of the time for it to "cool down". This means you are exchanging turn on time for brightness. This is where my little schematic comes in (note the green signal on the Gate. That's the signal I'm talking about when I say you turn the LED on for only a 10th of the time):



That symbol is an n-Channel MOSFET. See more about it here : http://en.wikipedia.org/wiki/Mosfet

MOSFETs are a form of transistor. They have a very high input impedance, and can handle currents up to 1000 A (and I'm not even exaggerating). I chose a MOSFET because Bipolar transistors just can't deliver that amount of power.

I mentioned earlier that when you operate infrared LEDs in this pulsed mode, and the frequency is high enough, you don't need a resistor to limit the current because the LED and MOSFET have such long delay times. This allows you to operate the LED at maximum efficiency.

So what we've done now is we've created a carrier signal for the actual data that we want to transmit. Because we burst the LED with that amount of power, the transmission range is of course much larger. Now the question arises, how do we get our music data on to that carrier signal? The answer is : Frequency modulation.

The simplest way is to just use the well known IC timer, the good ol' 555. (NE555N, LM555N, SE555N... They all pretty much do the same thing). In their datasheet they even have an application for this kind of thing (page 8 section 5): http://www.fairchildsemi.com/ds/LM/LM555.pdf

So I've modified that so we have a pulse width of 1/10. This would be my untested solution for a transmitter:



From here on it shouldn't be too difficult to create a receiver. Unfortunately I've run out of time. If you want to see me go through the receiver part I'd be happy to

TheComet

Ok, so I understand that circuit and what it is meant to do, but I'm not sure that the circuit will transmit enough information to have a good sounding signal reproduced at the receiving end.

Also, the 2V you mentioned, that will be the amplitude of the signal not the voltage across the LED. The minimum voltage across the LED would be 2.5 volts and the max would be 6.5 volts the way it is now.

I'm going to look at your circuit a bit more and see if I can come up with a receiver for it.