How to shield a live wire at 240V ?

[marconi]

Hello Mr Inventor. Good to hear from you again. The solar project has been decided on to go ahead but we need permission to connect it to the mains supply: this we have applied for. This means I have time to pick up where I left off with my version of your proximity switch.

l have an idea to control my ruler of lights based on whether the voltage from the sensor is increasing, decreasing or constant. This is to say that the actual voltage from the sensor is not important but rather whether it is going up, down or steady over two samples of it spaced out in time. It is being constructed at the moment. If you like a sensor of hand motion rather than hand proximity.

Are you in good form and making progress?

 

[_q12x_]

I'll probably have to draw the circuit - I didn't draw it yet and a movie to actually show you how things are standing at this point.
I will update soon.

 

[_q12x_]

 

[marconi]

Once again a nice clear video which shows your progress and current challenges. I liked that you have started to prototype circuits on the breadboard before moving them to the larger wings board. I'd like to see a ruler used to confirm your distances as I have done - we don't want any 'fisherman's tales'! ?

I will plough on with my project today and over the weekend.

Regards

Marconi

 

[marconi]

My breadboard so far.

 

[_q12x_]

"Once again a nice clear video which shows your progress and current challenges."
Thank you that you like it. And good job to you as well with the breadboard.

 

[marconi]

These are the stages I am working through with my version of your project:

1. Movement sensor to control the ruler of LEDs - my work at the moment, about 80% complete and then thorough testing;
2. Build pulsed LED array with 940nm torches* and associated pulsed LED receiver to produce Vs - pulsed sensor to distinguish torch reflections from steady (albeit slowly changing) ambient IR ;
3. Ambient IR monitor and controller to adjust for ambient IR - circuit built but yet to be fully tested and developed further as required.
4. Beep/click sound signal each time a ruler LED turns on/off.

* My current sensor arrays are 850nm.

 

[marconi]

My breadboard a day later.

 

[marconi]

This is a video demonstrating my version 2 of the ir proximity switch which includes the bleep/click sound as each ruler led turns on or off. As I have run out of space I will do version 3 the pulsed ir on a new breadboard. Quite a lot of time was spent on eliminating oscillation of an led if the hand stopped at a threshold or went through a threshold slowly.

 

[DPG]

This is a video demonstrating my version 2 of the ir proximity switch which includes the bleep/click sound as each ruler led turns on or off. As I have run out of space I will do version 3 the pulsed ir on a new breadboard. Quite a lot of time was spent on eliminating oscillation of an led if the hand stopped at a threshold or went through a threshold slowly.

Very nice Marconi

 

[marconi]

Very nice Marconi

I am a digital baby. So I aim to turn any problem into 1s And 0s as early as practical. Thereafter one has so many more options for signal processing. And it then becomes logical and mathematical which suits me. Analogue electronics is really an art which I have not had the opportunity or inclination to master. ?

 

[static zap]

I loved my 555 timer , but DIY ing it from op amps/comparators brings an understanding of Hysteresis and Long cycles for triangular waveform ... (I played with home brew car alarms !)

Analogue electronics is really an art which I have not had the opportunity or inclination to master.

 

[_q12x_]

Thank you for the video. You got good results. It's nice to hear the clicks.
It looks like it is finished. Can you make a circuit for me, with everything you have there? Thank you.
And this is my progress so far. Not that good.

 

[marconi]

q12 Good morning. If viewers were awarding marks you would receive high ones for the neatness of your construction whereas I most likely be told 'could do better'! I tend to be impatient and rush to see if something works and if it does I then leave it as is and move on to the next part. You are methodical and take the time to accurately record your circuits. I spent a good 2 hours trying to work out why my latest version did not toggle the lamp reliably. It turned out I had failed to connect the JK inputs to +5V ie logic 1 even though I had connected J and K together.

I am going to design next my 940nm pulsed sensor array and see how it performs. The major effort of version 3 is to maintain performance in spite of background IR from things inside the room and sunlight. I will keep you posted once I have something worthwhile to show. For this version I am using a better comparator LM311 which has the nice feature it can be 'strobed' ie told when in time to compare its inputs and produce an output - clearly necessary for a pulsed IR sensing system.

https://www.ti.com/lit/ds/symlink/lm311-n.pdf?ts=1617540813359&ref_url=https%3A%2F%2Fwww.google.com%2F

My versions 1 and 2 have used the LM339:

https://www.ti.com/lit/ds/symlink/l...rl=https%3A%2F%2Fwww.ti.com%2Fproduct%2FLM339

Happy Easter.

 

[_q12x_]

Happy Easter.
Thanks that you liked my "failure". Is kind of working as I said, but not that efficient (1 led always On), and only in night, when is out of day light. Also it is sensitive to my desk lamp as well, it is probably way too much amplified. But it works, in its limited way. It was more as an exercise with opamps for me, a practical one. The most used way of an opamp as I could seen it so far, is as a comparator. Then the other wonderful exceptions ways. At least is how I perceive it, now in the beginning. I must play more with it in the future.
Right now, I am planning on tweaking the amplification of the SCM - 2n7000, the one with the npn tr T3 and its led, somehow to make it work with the lm3914 pin 5. Currently, as a reminder, is only working directly from the Source of 2n7000 to pin5, because it is grounded by a 10k resistance R4.

My good american friend has sent me already an IR detector that has a chip in it, very advanced component and very miniaturized, and it will receive and convert automatically the signal from any remote control or signal generator. It is quite common but is a bit more expensive than a simple direct IR receiver like the black 3mm ones I have. The package is on the way, and we all wait for it's arrival.
Until then, I will tweak what else can be tweak.
Thats why I asked for your circuit, to see where and how did you linked the buzzer. I can not understand from your real picture, since its full of wires and little "black boxes".

 

[marconi]

The click is produced by comparing the number of LEDs illuminated at one instant - let us call it time t - with the number illuminated a short time T earlier in other words at t-T. My version 2 is using synchronous logic circuitry driven by a 'clock' running at several kHz which produces my clock ticks. It is bottom left with a single faint red LED next to it. I can number the ticks since the power up using the letter n, so a sample taken at the nth tick of the clock is Sn, and n is increasing from 1 to infinity.

As there are eight LEDs on my ruler, I sample at regular time intervals of T, to see which ones are on or off. This produces an eight bit sample word viz Sn= ( L1, L2, L3, L4, L5, L6, L7, L8). If an LED is lit Lx = logical 1 otherwise Lx = 0.

I then load Sn into an 8 bit register called P. Several clock periods later - T - the contents of P are loaded into another 8 bit register Q.

After another interval of T another sample, Sn+1, is taken of the LED ruler state, Sn+1 and loaded into the register P. The content of P is now Sn+1 - the latest sample - and the content of Q is Sn - the previous sample. After a further interval of T the content of P is loaded into Q. And this sequence repeats over and over again.

Registers P and Q therefore hold samples taken at Sn+1 and Sn respectively as n counts upwards. I could write this as Sn and Sn-1 which makes it clearer with is the latest sample Sn and which was the previous sample Sn-1.

An 8 bit logical comparator compares the contents of P and Q by doing the subtraction X = P - Q. If X = 0 then Sn = Sn-1 which means there has been no change of the state of illumination of the LEDs on the ruler over the interval T. If X does not equal 0 then more or less LEDs are illuminated now than there were before.

https://www.ti.com/lit/ds/symlink/sn54ls684.pdf?ts=1617648996931&ref_url=https%3A%2F%2Fwww.google.com%2F

The logical state of X then is a signal on whether the number of LEDs which are lit is steady (X = 0) or changed over successive time intervals T.

I use the X signal to gate a 1kHz square wave to the sounder to make the beep. Actually, because X is such a short duration pulse since T is actually very short in time I send X to a pulse stretcher (a monostable) to produce a longer gate pulse. This creates the click sound every time there is a single change of state - on to off or off to on - of the most significant LED.

 

[marconi]

q12 Some time ago you did not have in front of you a design of IR continuous wave active passive sensor which would gradually light your LEDs in response to the proximity of your hand. And now you do. To my mind then, describing your latest model of led wings as a ‘failure’ is unwarranted.

 

[marconi]

q12x Good morning. Just a thought on your latest IR sensor. The photodiode is operating in the photovoltaic mode which is ideal for low light levels because the dark current noise the photodiode produces is minimised. Any diode current flow is then mainly due to photons hitting the photodiode.

It may be that at the moment the single IR LED is not providing sufficiently intense illumination of your hand in front of the sensor which is compensated by using a high gain cascade of op amps (1) and (2) - (A = 100 x 10 = 1000). Of course this also means that ambient light from other sources will be detected and similarly amplified by a 1000.

What about brightening up the IR light you shine by having 2, 3 or 4 IR LEDs placed around the single IR photodiode? One might then reduce the gain of op amps (1) and (2) which would make your sensor less responsive to ambient light. The sensor would then mainly be reacting to the reflected light from the IR LEDs and tend to ignore ambient IR light. Worth a try? The second image shows my 850nm array of 4 ir LEDs and 4 pds.

 

[_q12x_]

Hello mister @marconi and thank you for your answers.
I kind of understand.... what you did there with the buzzer circuit, but PLEASE, make for me a simple circuit diagram. All I need is the output signal (pin1 to 10 of LM3914, or in other words the leds negative pins) and the input signal, (from what I understood, you used a comparator IC and a clock (quartz? or RC?)). But I still dont understand when the click sound is activated and by who. Explain in very simple words, not math words. Im an artist, I deal better with visual explanation than weird math algorithms explanations. Sorry but im not a math guy. Make me a short diagram please, draw it as simple and explicit as possible on paper and scan/photo it. Thank you ! I really appreciate your effort so far.
The other subject, the IR leds, YES, I will have to test more of them as you just did on your board. But again, this was more an educative circuit for me, since im quite new to both IR and opamps. Now im a bit more familiar with them. I will make it in the near future.

 

[marconi]

I hesitate to start designing circuits for you because I then become 'responsible' for them by working with you until they perform as desired when connected to your other modules. There is no point in copying the way I produced my 'buzzer' circuitry because my versions of your project are mainly using digital ics and techniques and as you can see from my video you do not have room. If you want me to guide you through the design process for something to work with your wings then that would be preferable to me. I spent quite some time making sure the LEDS on my ruler - so your wings LEDs - cleanly turned on or off without any fading or oscillation to ensure a single click each time there was a single LED change. As a fact your LM3914 fades the turning on and off of adjacent LEDs - it says so in the datasheet; I suspect - forgive I may be wrong - that you still have oscillations. Oscillations will be seen if you very slowly move your hand towards or away from the sensor or if you hold your hand in a position where adjacent LEDs turn on and off/off and on. Perhaps you could investigate.

In outline you would need to build a tone generator to drive the sounder. I suggest with a 555 chip. Is your sounder a piezo electric one and if so is it passive or active? If it an active one then it has an internal tone generator so you would not need to build a separate 555 tone generator. Mine is a passive one. Assuming a passive sounder driven by a 555 tone generator, we could use the trigger pin 4 to turn the tone on and off easily with a negative going pulse.

Next step would be to generate a negative pulse for each of the LED driver outputs of the LM3914 and then to combine them as logical OR (ie any driver output) to be used to turn the tone generator on briefly and thereby make the click sound. The sort of circuit I envisage for the negative pulse generator is a simple CR differentiator across each LM3914 LED driver transistor. Something like I have drawn in the attachment.

How does this sound to you?