I have always wanted to equip my lab with a device to measure the values of various chokes. The cost of such equipment was previously quite prohibitive, and so I still do not have any. First of all, know that this small device can obviously not compete with the more complex measuring devices of the trade, it is only a modest indicator allowing to get rid of the value of an unknown self, quite simply. .
However, I recently discovered a site ( http://kudelsko.free.fr/inductance_usb/sommaire.htm ) in which is described the realization DIY ( Do It Yourself ) a small device allowing the measurement of unknown chokes on a range of a few tens of nH to 10 mH approximately ... In addition, the author has developed a small interface running under Windows which allows to recover via USB the value of the analyzed choke and display it on your computer screen.
His approach greatly interested me, but I was looking more for an autonomous device ...
I would like to thank him here for his sharing which was the basis of my study.
Functional description and software
|Component Types: Through|
|Level: ◆◆ ◇◇◇|
|Construction time: 2 to 3 hours|
|Cost: approx. € 57|
There are several possibilities to design an oscillator capable of providing a sinusoidal signal, for example ( source Wikipedia ):
- Colpitts oscillator ( chosen configuration ... )
- Clapp Oscillator
- Phase shift oscillator
- Pierce Oscillator
- Hartley Oscillator
- State variable oscillator
The operating principle is simple, it uses an LC circuit tuned on the basis of a Colpitts oscillator. The value of the capacitors being known ( I measured their value with my capacitor for more precision in the calculations ), it only remains (...) to apply the formula of THOMSON which goes well to calculate the value of Lx:
When the reactances of the inductor in series with the capacitor are of almost identical value ( condition difficult to obtain taking into account the losses and parasitic capacitances ... ), then the oscillation frequency of the system is called the resonance frequency f0 .
These calculations are entirely carried out by the microcontroller, which is responsible in the first place to measure the frequency of the oscillator via its RC0 port. In order to be able to carry out this measurement, it is essential to bring it back to acceptable values for the µC, and this is why it is divided beforehand via a binary counter of the HEF4040 or similar type. The division ratio used here is 1/32. Thus, the microcontroller only acquires a known fraction of the frequency to be measured, thanks to its Timer1 configured in such a way that it triggers the counting from the first rising edge on RC0, the time scale being set to one second. Therefore, the value of Timer1 represents the exact reflection of the frequency of the signal present on its input. All that remains for the processor isto perform all the calculations necessary to determine the value of our unknown self Lx.
I modified somewhat the original diagram which accommodated a microcontroller PIC18F2550 by replacing it with a PIC18F252, no need for the on-board USB interface, and added a 2x16 character LCD display.
I then developed a small software under mikroC allowing, as in its first version, the measurement of any self, and the formatted display ( in nH, µH and mH ) of its value on the LCD, the range change being automatic. I voluntarily opted for the PIC18F series because there are some floating point calculations as well as squared elevations and the PIC16F series in these specific cases is problematic ...
A little word about the Reed relay REL1:
In order to obtain an oscillation which starts frankly whatever the value of the unknown choke Lx, a first choke L1 is taken in series with this one. During the measurement, the resulting frequency of the oscillating circuit is dependent on these two chokes. Apart from this we only want to know the value of Lx , and to do this, the software performs a first measurement by short-circuiting Lx so as to be able to memorize and then subtract this initial value f , it is the [ ZERO adjustment] for which a particular message is displayed on the LCD. Then the second measurement will take into account the two chokes, and the software will calculate the frequency f0 of our unknown choke. All this is fully automatic, the only maneuver is to momentarily press the push button ( connected to RC5 ) to possibly restart the procedure. If the choke to be measured is disconnected, then the message [ no coil detected] will be displayed and the red LED (LED3) will light up. In this case, it will suffice to check the correct connection of the choke, and to restart the measurement sequence.
My model is built on the basis of a Ready for PIC development board from mikroelektronika . In addition to the micro-controller, it features an 8 MHz quartz, has an LCD display and a push button that must be actuated for each new measurement. The PICkit3 programmervisible here is only for development and is not necessary for the use of the inductance meter.
Note that it is of course advisable to make connections as short as possible between the measuring head and the choke to be analyzed because these can generate errors due to stray capacitances ...
those I use ( visible on the photo of my model ) measure 12 cm and are equipped with gripping plugs.
I again changed the schematic I published in the first place somewhat again for several reasons:
- the relay first of all ... the author had planned a Meder model (DIP05-2A72-21D) with 2 work contacts (NO) connected in //, but I have a Meder (DIP05-1A72-11D) with 1 single contact (NO).
However, I could not find the Meder library in Eagle v7.0, and had to use the one from Hamlin (HE722A0500) replacing. It turns out that the 2 models are perfectly compatible in terms of their pinout, of course. My PCB takes this specificity into account.
- the LCD display ... on my model this is connected to the PortC via a HE10 type connector. On the final PCB it will be directly mounted in "imperial" by means of a 16 way plug-in connector.
In fact, I have provided fixing holes on the PCB to be able to fix it securely and reliably using nylon spacers.
- the push button originally wired to RC3 ... it is now wired to RC5, this for PCB routing convenience.
Design and packaging
This is what this card should look like ( double-sided with metallic holes, silkscreen printing and sparing varnish ).
I specially designed it to integrate it into a Multicomp MCRH3135 box...
Your trained eye will not have failed to notice the very small size of the pellets ... they measure for the most part Ø 1.4 mm for a drilling of 0.8 mm ...
I cannot recommend enough to you to use an excellent iron soldering ( I use a Weller WECP-20 stationwith a 25W MLR-21 ironequipped with a Ø 0.25 mm Micropoint type tip ) to carry out the welding operations, or failing that, you will have to route your card again differently ...
[Edit of 11/15/2017]:
Two "small" details, however, about routing ...
- I forgot to check the dimensions as well as the shape of the J1 power connector pads ...
annoying only if you use the same model as me, for which you will then have to "work" the board a bit with a tool like Dremel ( see Image 2 ) ... sorry, that escaped me ...
- it is also sorely lacking the ICSP connector which allows programming of the processor in-situ ...
inconvenient to say the least, it is however possible to program it easily using a Breadboard ( see Image 3 and 4 ), a pelletized plate, etc. ...
Here is what it gives after the "modification" of the circuit ... the J1 connector not being any more correctly maintained by its soldered connections, I stuck it with a contact glue ...
Photos of the card being assembled :
The programming phase
The finished device
Here, as promised, I give you the last photos of my finished inductance meter.
Details and finishes
For the moment, the connections with the self to be measured are made via 2 wires of about 12 cm equipped with gripping plugs as on the prototype, I do not yet know what type of connector to use ...
I designed the screen printing of the front face using the superb free Front Panel Designer software , then printed on a 3M self-adhesive aluminum support ( bought a few years ago from Selectronic ...) with my laser printer.
The result is clean, the most delicate being the cutting and the correct positioning of the aluminum support ...
On the last photo you can read the value of a commercial self of 220 µH ( marked "221" ), the analyzed value is 226.21 µH is an accuracy of ± 3% considering that there is the tolerance of the component as well as that of the measurement ... the precision for this type of device is largely sufficient ...
The power supply part uses an MC34063, a specialized circuit that takes care of the regulation to deliver a perfectly filtered voltage of +5V.
The voltage applied to the input of the assembly is first rectified and then filtered a first time via the 330 µF capacitor C10 before entering the Buck converter IC3 which is the MC34063. Its working frequency is approximately 30 KHz for a total measured consumption of approximately 35 mA. At the output of this circuit, it is again filtered through the 220 µH choke L2 as well as via the 330 µF capacitor C12. This results in a voltage of +5V which is perfectly filtered and stripped of any high frequency residue due to the switching frequency. This power supply is capable of delivering a maximum current of 1.2A.
I come back to the rectification part ... the main interest of this type of assembly is to be able to use power supplies delivering both AC and DC voltage, and for the latter to ignore any polarization problem. So no need to worry about where the positive pole is, it works perfectly whatever the direction.
The power connector is 2 mm in diameter, which allows the use of many standard materials.
Practical and comfortable therefore ...
It is obviously possible to use a conventional 78L05 type regulation associated with its filter capacitors, but the output signal will not be as perfect as in the present case.
It is a proven power supply which is also embedded on the Ready for PIC card ... hence my choice.
The analysis of the choke to be measured is carried out in 2 stages:
- system calibration sequence with display of a first message
- measuring sequence with display of the calculated value
The use is really very simple and comfortable to use.
Schematic diagram I drew under Eagle:
Note : the values of some components differ from the original diagram by F. KUDELSKO ...
This is mainly due to the fact that I used those I had in my drawers ...
The values noted in blue are the values actually measured with the capacitance meter.
Diagram of the power section:
Packages comprenant : ◊
Software written in mikroC language ( source and executable )
◊ Schematic diagram ( without PCB ) in Eagle v7.7.0
◊ Drawing of screen printing on the front panel
◊ list of components BOM
A buck converter, or serial chopper, is a switching power supply that converts a DC voltage into another DC voltage of lower value. A well-designed buck converter has a high efficiency and offers the possibility of regulating the output voltage. Wikipedia