This article actually represents two. I did not separate them because they complement each other.
One concerns the power supply of an LED, the other that of the safety measurement in the laboratory.
It is often useful to have a mains presence signaling LED, when building a laboratory power supply for example. Recently, I built a device consisting of an isolation transformer for my safety in the lab but not only. I mainly use it to be able to take measurements with my devices, which are mostly earthed, and the oscilloscope is one of them. For example to measure the voltage at the primary of a switching power supply or to visualize the signal of the chopper transistor ( that of a PC why not ... ), it remains impossible to achieve this without taking considerable risks both for oneself and for the material, in test or measurement.
As the oscilloscope is almost always equipped with an earth connection, the mass of its chassis as well as that of its measurement probes are also connected to it, and it is very well like that!
If by any chance you took the urge to try this kind of manipulation, it would only last ... until your differential circuit breaker trips!
Eh yes !!! because when connecting the ground of your probe to the primary ground of the power supply that you wish to auscultate, you establish nothing more or less than a nice bridging of one of the poles of the AC 230 V sector and ... the earth of your electrical installation. The circuit breaker will immediately trip because the pole thus grounded ( the Phase or the Neutral, you have one chance in 2 ... ) will initiate a leak, hence immediate disruption.
But then how to do?
Some smart kids will immediately say to themselves ( yes, I noticed it in the repair shop of an after-sales service, that's to say to you ... ) ... and if I isolate my oscilloscope of the earth ... I will no longer have this problem of disjunction, and suddenly I will be able to take my measurements ... Eureka!
Well no, lost!
Do not play with your life, it is better to buy a new power supply for your PC, it will be less cruel ...
By isolating your oscilloscope, and therefore connecting the ground of your probe to the primary ground of the power supply to be repaired, you will end up with what is called a floating mass, i.e. without any reference to the earth, earth to which you are most likely connected without really knowing it too much ( even with good leather shoes and rubber sole ...). And if you try to operate this or that setting on your oscilloscope, I wish you luck and ... a lot of courage, because its chassis may very well end up with a potential difference with respect to the earth of several hundred volts, or even more ...
DO NOT PLAY WITH YOUR HEALTH !!!
And as personally it often happens to me to be confronted with a difficulty of this kind, I decided to build myself a protective device, and for that I was inspired by the one found on the Internet here .
- the first, which is white in color, is connected to the mains by means of a 2-circuit switch so as to cut the 2 poles ( Phase and Neutral )
- the second which is red ( to differentiate it ) is connected to the secondary of the isolation transformer (this is how it is called ...) via a 2-circuit switch so to cut the 2 poles (pseudo Phase and Neutral)
With this I wanted to be able to measure the current absorbed by the device on which I am carrying out my manipulations, and so quite naturally I added 2 "banana" sockets which are short-circuited by a single-pole switch.
Add to that a power socket attached to the chassis of the same model as that of the PCs ( I have extra cables, so it's an excellent solution ), a time-delayed type fuse in the primary circuit of the transformer ( due to mainly from the current inrush during commissioning ... ) and a second fast type in the secondary of the same transformer.
Everything is there ... including the two famous signaling LEDs. They are high luminosity models emitting a superb very powerful blue light although only consuming approximately 7mA!
There for once, we can not ignore the fact that the machine is under tension ...
Usually when I design a power supply, I always plan to insert a Led on the DC 5 V in order to visualize that it is present.
But here, none of that! I only have AC 230 V at both primary and secondary ... how do I do this?
Not wanting to add an additional transformer with the diode bridge for rectifying, filtering and then regulation via a zener diode or a 7805 for example, all this just for one Led? in fact no, I need two, one to let me know if there is mains power at the primary, and a second to let me know that AC 230 V is present at the secondary.
Do you always follow me ?
With these considerations in mind, I am going to override and ignore the transformer + rectification + filtering + regulation ... too expensive and too bulky. In addition, the additional transformer would consume energy just for the LEDs ...
It would be really wasteful!
The solution is to feed the LEDs directly from the mains ... well directly ... not quite as you will see below ...
The idea is to say that it must be possible to make a sort of voltage divider bridge so as to be able to power the LEDs ...
yeah ... and this is what I will explain to you:
I would also like to thank Rémy MALLARD for the valuable and relevant information (from which I was inspired ... ) that he regularly posts on his SONELEC-MUSIQUE website .
So let's take a look at the following diagram:
Apart from the capacitor C11 ( 100 nF / 400 V X2 ), we are in the presence of a voltage step-down circuit with for R11 ( 330 Ω ) a current limitation, and a voltage regulation with the zener diode ZD12. Diode D11 ( 1N4004 ) is used to protect the Led by suppressing negative alternations. This therefore only sees the positive half-waves, ie one per period as in the case of a single-half cycle rectification. This is more than enough to power our Led.
At the terminals of this diode we have a voltage directly linked to the zener voltage. Having chosen in my assembly a model BZX85C-3V9 ( which I had quite simply in stock ...), the voltage at these terminals can never be greater than 3,9 V. It is this voltage which will feed our Led via a series resistor ( which I have voluntarily defined at 33 Ω ) which again serves to limit the current flowing this time in the Led.
Yes but ... indeed, here no transformer so no diode bridge and even less filter capacitor.
In fact the filtering is not in the present case necessary since we will not feed an assembly based on microcontroller for example, and for which one would obviously need a rigorous filtering.
The diode bridge is not useful either, a half-wave rectification will be largely sufficient for our use.
There is still the capacitor ... what is it for?
In fact, in sinusoidal regime, its reactance which depends on the frequency of the voltage which is applied to it as well as the value of its own capacitance cause it to behave a bit like a resistance, and form a divider bridge with the set R11 + (D11 // ZD12).
The formula used to define the reactance of a capacitor responds to the following:
|Xc = 1 / ωC with ->||Xc being the reactance of the capacitor expressed in Ω|
|ω being the pulsation of a sinusoidal signal at frequency f(Hz) with ->||ω = 2 * π * f|
|C being the capacitance of the capacitor expressed in Farad|
|Pi being rounded to 3,14 to simplify calculations|
|We thus obtain for Xc the following equivalent formula:|
|Xc = 1 / (2 * π * f * C)||-> Xc = 1 / (314 * 100 E-9 ) -> Xc = 31,847 KΩ|
|by applying Ohm's Law U = R x I we can deduce the current delivered by the assembly:|
|i = (230-3,9) / Xc -> i =||(230-3,9) / 31847 -> i = 7,099 mA|
With this low current, I still manage to power a high-brightness blue LED, and believe me ... it illuminates!
And all this with almost only recovery components:
- the Led comes from an 8 socket strip with a light switch as well as 2 blue Led which I did not need ... no need for signaling since those of my lab ( I have several ... ) are constantly on ... I kept the switches, but recovered the LEDs, in case ...
- the 100 nF / 400 V X2 capacitor comes from the switching power supply of an old destroyed VCR ... there were 2 which were placed directly at the mains supply, and which were used for filtering high frequencies ( about 50 KHz ... ) from the chopper circuit. This device is in fact used to suppress these high frequencies which propagate on the AC 230 V network of your home, and therefore to clean up the sector.
Yes, a poorly filtered switching power supply pollutes!
- the 3,9V zener diode I had in stock, as well as the resistors elsewhere ...
So for me the cost is negligible ... compared to the purchase of a very low voltage transformer + everything you need to add around to make a power supply, the calculation is quickly done!
And the result is quite amazing, see below:
BE CAREFUL, HOWEVER!
The assembly is powered by the mains, and therefore remains dangerous during direct contact with all or part of the human body!
Be careful above all, this kind of circuit should be handled with care!
And I cannot recommend the use of an isolation transformer enough to avoid any risk of electrocution between one of the poles and the earth !
Of course, this has a price ... the cost of my equipment is around 150 €, but at what price do you rate your safety?