I am seeking advice concerning construction of what I assume to be fairly simple circuit designs. Excuse my ignorance as my background is in computer science, not electrical engineering. There are two projects I wish to build.

The first is a modification of this device:

http://www.eskimo.com/~billb/emotor/chargdet.html

Specifically modifying the simplest configuration:

http://amasci.com/emotor/fetdet1.gif

The circuit there activates an LED when the FET is switched on. What I want to do is make a version of this design where the LED is always on, until the FET is activated, which will then switch the LED off. Afterwards I want the LED to stay off, until the circuit is reset via a button. If someone could tell me the simplest way to make such a circuit, I would be very thankful.

The other project I would like to build is the simplest white noise generator circuit possible (including a speaker). I do not want to use an induction coil or antenna to do this via radiowaves. I want to produce white noise using circuit components only. What is the simplest way to produce white noise from a non-radio source?

I appreciate any and all advice concerning these projects.

 

go to google IMAGES and search "white noise generator". You will get hundreds of simple circuits imaged on the page.

 

I have done that Kermit2, the trouble is I don't know which of those hundreds of circuit designs are the optimal one to use. I'm looking for advice concerning the overall best and simplest method for that.

 

What I want to do is make a version of this design where the LED is always on, until the FET is activated, which will then switch the LED off.

You could try placing the LED, with an appropriate current-limiting resistor (e.g.. 1kΩ), in parallel with the FET. The LED will be lit while the FET is not conducting. When the FET conducts, the ∆V across the LED should become too low for it to light.

For white noise, I think you could just use any op-amp at maximum gain, maybe using carbon resistors (which are noisy) on the inputs. That should generate noise at line-out impedance. You'll need a small power amp if your speaker has a low impedance.

 

Thank you for the information Wayneh. The white noise circuit you described sounds like a very good solution. I will give this a shot.

I am having trouble visualizing the circuit you described for the LED solution. What you described probably works but without a visual schematic I'm at a bit of a loss. I'm just not that knowledgeable of these things yet. Maybe I am approaching this application in a more complicated manner then it has to be; perhaps I don't need a MOSFET for this. So to explain my idea as simply as possible:

There is an LED that is lit. It has a small antenna near it. When you disturb the electromagnetic field near the antenna, the LED is turned off, and it stays off until the circuit is reset (button to cycle power).

I am planning to make an array of these little LED circuits which function in this manner. So to make each LED circuit as simply and cheaply as possible is paramount. I am 100% open to any ideas anyone has to accomplish this tiny piece of tech. Like I said, a MOSFET might be overkill, I don't know.

I appreciate any advice, thanks.

 

It's not a MOSFET, it's a FET.

 

I am planning to make an array of these little LED circuits which function in this manner. So to make each LED circuit as simply and cheaply as possible is paramount. I am 100% open to any ideas anyone has to accomplish this tiny piece of tech. Like I said, a MOSFET might be overkill, I don't know.

Things like this tend to go much better when you clearly define what you want a circuit to do. Circuits like the one you linked to that detect stray voltage in the air are pretty simple in forms like the link. Unfortunately, they don't work very well and can be viewed more as a novelty. Having a circuit latch and require a reset isn't difficult but the latch or trip point needs defined. Hanging an antenna off the gate of a Jfet or any transistor isn't difficult but again let's say you want a LED on and it turns off when detecting a field. What magnitude field?

Ron

 

Things like this tend to go much better when you clearly define what you want a circuit to do.

What I am trying to construct is a variation of something like this in mass multitude:

http://www.eskimo.com/~billb/emotor/chargdet.html
http://amasci.com/emotor/fetdet1.gif

If I understand things correctly, in that design the LED lights up when the FET is switched by EMF disturbance. Let us consider this device "1 pixel". I want to create a grid like array of "pixels" that consist of LEDs individually triggered by nearby EMF disturbance. I would want the EMF detection to be sensitive enough that when someone places their finger near a respective LED "pixel"'s micro-antenna it would turn off that individual LED. So in a way you could say I'm making a drawing tablet of these LED pixels to be drawn on by using one's finger. The complexity lies in the fact that the LEDs need to be all lit initially, and turned off as they are triggered independently, and then stay off until power is cycled on the entire circuit.

I do not know if a FET such as the one described in the article I cited is powerful enough to detect a single finger's nearby presence to it, in enough magnitude to trigger the transistor switch. Then there is the issue of individual EMF antenna placement in such a way that the LED itself does not trigger its own self off. Obviously my electrical engineering knowledge is only rudimentary. That's why I'm seeking expert advice on this project. There may very well be a far simpler way to construct this circuit application than I am currently realizing. I just need to create an array of lit LED pixels that can be independently turned off by the presence of a nearby finger's EMF disturbance.

Thank you for your time.

 

Among the many white noise generator circuits, the most basic is an intentionally noisy semiconductor junction and an audio amp chip. The semiconductor can be a zener diode or a reverse-biased transistor base-emitter junction, and the amp chip can be just about anything. For a small speaker, the LM386 is a favorite.

https://secondhandsynth.wordpress.com/2010/04/08/random-white-noise-sample-and-hold-circuit-part-1/
https://www.flickr.com/photos/48381360@N02/4501573713/

True white noise tends to sound to tinny or bright, like the treble is turned up. That's because true white noise has constant energy per frequency. Pink noise is what most people think white noise should sound like. Again, lotsa circuits out there. The general plan is the diode noise source, an active filter opamp circuit to shape the noise spectrum changing it from white to pink, and then the audio amp chip.

ak

 

Sorry I missed your reply for a day.

As to the little circuit, it's pretty basic and while I never built that version it likely would work even as crude as it is. Note the Jfet used in the circuit is obsolete but just about any replacement would work. Since you want an array of these things that respond to touch and latch the LEDs off what you want is a very basic touch sensor circuit. In some versions a finger bridges a small split surface and in others a finger acts like an antenna and noise triggers it. You want to start with all the LEDs on and as a finger goes over the sensors in a matrix the corresponding LEDs shut off. That what you are looking to do? A Google of Touch Sensor Circuits will explain in more detail what I am talking about. The finger used to bridge is the more reliable I think.

Ron

 

... white noise has constant energy per frequency. Pink noise is what most people think white noise should sound like.

I always wondered what pink noise was. (Not enough to look it up, apparently.) Learn something every day here.

 

The semiconductor can be a zener diode or a reverse-biased transistor base-emitter junction, and the amp chip can be just about anything. For a small speaker, the LM386 is a favorite.

Thank you for telling me the actual components I will need, this is very useful information.

in others a finger acts like an antenna and noise triggers it

This is the type of sensor that I need. I thought a FET could do that. I may be entirely wrong about that.

You want to start with all the LEDs on and as a finger goes over the sensors in a matrix the corresponding LEDs shut off. That what you are looking to do?

That is exactly what I am trying to build. The sensors need to be triggerable from a human finger. Imagine a grid array of LEDs that react this way independently.

I know this an unusual concept, and that's also why I need help. I do not know what components I need to buy and in what configuration to wire them. If I knew that aspect for the actual LED control side of things, building the array would be no problem for me. I can handle the bus power setup, potentiometer implementation, etc. It's just the circuit logic and component configuration I need expert advice for.

Thank you for your help so far.

 

Tomorrow I'll toss out a circuit that you can try. I have an idea or two.

Ron

 

Tomorrow I'll toss out a circuit that you can try. I have an idea or two.

Ron

Fantastic, looking forward to any ideas you have.

 

I haven't forgotten you. Been playing with a touch circuit I think will work. I need to get it latching and get a reset in.

Ron

 

Been playing with a touch circuit I think will work. I need to get it latching and get a reset in.

Sounds promising, I appreciate your effort.

 

Here is what I came up with. There would be one circuit for each LED, the reset would be wired common to all the CD4013 D Flip Flops. Each CD4013 is a dual D flip flop so two flip flops per package. Unfortunately I can't say for sure this will work or how well it will work. Touch plates like this depend on several variables including how much free energy is floating around in the air and is the circuit sensitive enough. R4 actual value will depend on the LED but for example a Red LED with a forward current of 20 mA and forward voltage of 1.2 volts would work with the 9 volt supply. The circuit would also work well likely on 6 volts supply. Again, no guarantee on this as I just can't be sure how well it will work. Things like this are more a novelty than anything else.

Maybe another member has some ideas, thoughts or another circuit?

Ron

 

Reloadron, I want you to know that I deeply appreciate the effort you put into that circuit design. I will attempt a construction of it in due time. Indeed this is a novelty concept but not without potential merit. I would explain why but I'm under an NDA. Again, thank you so much for your work.

 

A possible alternative to the fet, bjt, and resistors would be a capacitive touch IC like this one. Each ic can handle 5 inputs, so the cost spreads out pretty well and it would be simpler wiring. You'd still want the output to feed into Ron's CD4013 portion of the circuit for latch and reset functions.
I'm really interested in these ICs, but I haven't actually played with them yet, so I can't vouch for them personally.

 

A possible alternative to the fet, bjt, and resistors would be a capacitive touch IC like this one. Each ic can handle 5 inputs, so the cost spreads out pretty well and it would be simpler wiring. You'd still want the output to feed into Ron's CD4013 portion of the circuit for latch and reset functions.
I'm really interested in these ICs, but I haven't actually played with them yet, so I can't vouch for them personally.

Interesting chip. Thanks for sharing that as I never saw one.

Ron