1_KHz_Inductance

Downloads
Download the complete WINAVR project with Peter Fluery's LCD Library: 1kHz_Inductance_meter_c_files.zip
Download just AVRStudio hex file: 1klm080813B.hex
Downlad the schematic 1klm18aug08.gif

Find updates at www.projects.cappels.org

HOME (More Projects)

Introduction

I get a lot of use from my Even Better LC Meter (http://www.cappels.org/dproj/EvenBetterLCMeter/Even_Better_LC_Meter.html) but as satisfying as the meter is to use -its quick and pretty accurate, but when measuring inductors with low Q, there are significant errors, and there was the ever-present nagging issue of the test frequency being a function of the value - it does not test the inductor at a standard frequency, so with some inductors, in particular those with iron of ferrite cores, the results might not correlate well with some of the better made inductance meters.

1 kHz is a commonly used test frequency, and I resolved to see if I could make an inductance meter that uses a 1 kHz test signal at a low level. I also wanted to use a low level test signal so that the test current would have a minimal effect on the measurement. Magnetic cores, and soft ferrites in particular, change in permeability as a function of the current through the winding. I chose 1.0 milliamps peak-to-peak as a test current. Given that I wanted to be able to measure inductance down to 100 uH, my attention turned to the problem of how to accurately measure the amplitude and the phase of as little as 63 millivolts across a 100 uH inductor.

Searching the web, I saw that people were having pretty good success using synchronous demodulators, and this appealed to me since my exploration of synchronous demodulators a few years ago (Experimental 1 kHz Synchronous Demodulator and Digital Lock-in Milliohmmeter) lead me to appreciate how well these demodulators can separate a signal from the noise.

Looking around on the web a little more, found the user's manual for an M³ Electgronix Digital LCRZ Meter kit, which, aside from spelling out the meter's impressive capabilities, also explained a few short paragraphs explained how the meter worked. That was enough to convince me to give this technique a try. By the way, if case you are starting to get the idea that this project is a quick and dirty knock-off of the M³ design, its not. The M³ kit has much more range, measures lot more than just inductance, and seems to be quite a bit more accurate.

My project is just an exploration of circuit and firmware techniques, mostly to see what it would take to get such an instrument operating. I think that if you have the time to build the kit, the M³ kit is an excellent way to get a full function LCRZ measurement capability that rivals that of some of the most expensive factory built instruments, but at fraction of the cost. You might want to have a look at the M³ Electgronix webiste, and in particular at the web page for their meter kit. You can even download the manual for their kit on the page at this link: http://www.m3electronix.com/lcr.html.

The result of this investigation is some circuit concepts, some firmware techniques, and some learning. Notably, the internal reference impedances are switched on and off without the use of electromechanical relays or switches, and a comparatively low parts count version can be made with some sacrifice in measurement linearity. The ATTINY861 that I chose to use in this investigation seemed to have way more program space that I thought I would need - 8 k Bytes, in all, but as the design progressed, the space filled up rapidly. Soon, I was at the point of scrimping to save memory. I think an experienced embedded C programmer would be able to find plenty of places in my code where space can be saved, but this was all I could cram into that chip.

My meter has two ranges: From about 0.6 to 33 Ohms and a higher, overlapping range from about 30 Ohms to about 330 Ohms. Having only two ranges, made it easy to switch ranges automatically.

In designing the meter, I assumed the most useful model to me: a pure inductor in series with a pure resistance. I ignore capacitance all together. The meter calculates and displays the impedance of the inductor under test, the angle of the voltage across the inductor with respect to the driving current, and displays the actual inductance and resistance. The resistance display can be switched between resistance and Q, where Q is defined as the impedance divided by the resistance.

Taking a hint from the M³ Electgronix manual, the instrument is first calibrated with the use of a precision resistor, which is measured and compared against an internal reference, thus calibrating the internal reference. The internal reference is then used for future calibration. Calibration parameters, for both the original calibration and subsequent calibrations using the internal reference, are stored in the EEPROM on the ATTIN618 micro controller.

As a result of running out of program memory, I was not able to get the data display to work to my liking. Below 100 uH, the decimal point disappears, and can be a problem. A 90 uH inductor, for example, looks like 9 mH on the display.

This is a good time to mention patents. Don't assume that anything you read on this web site is not covered by somebody's patent(s) somewhere in the world. I have seen methods similar to this one mentioned in issues U.S. utility patents, and given the breadth of technologies and techniques covered in this project, I think the only safe assumption is that commercial use of this circuit would infringe on some patent. As mentioned previously, the intent of this project is to learn, not to devise a design for a product.

Block diagram of the measurement system.

The method is very simple: Drive the inductor with a 1 kHz current source and then measure the amplitude and the angle with respect to the drive current, of the voltage across the inductor under test. The results are displayed on a liquid crystal display. An internal reference is used for automatic calibration, and push buttons allow the user to initiate calibration modes and switch the display between Q and Resistance.

The voltage across the inductor under test is measured over two areas, 90 degrees apart. One of the two areas is arbitrarily regarded as being the real part, that in phase with the 1 milliamp sine wave being used to drive the inductor, and the other being arbitrarily regarded as the imaginary part, that which is 90 degrees out of phase from the 1 milliamp sine wave. Within the measurement, a total of four measurements are taken. Two for the real part, and two for the imaginary part. The real and imaginary parts measured, then measurements shifted by half a cycle are taken, and after correcting the sign, the two are averaged to remove some offsets. This approach is similar to that mentioned in the M³ Electgronix manual. Offsets are removed from each measurement and five measurements cycles are averaged. The resulting magnitudes are used to calculate the raw vector.

The major blocks in this section of the circuit are the 2 line x 16 charter Liquid Crystal Display, the push buttons, a latch to create two additional output signals, the pulse width modulation output and first low pass filter stage, and the micro controller itself.

The 2 line x 16 character Liquid Crystal Display uses the Hitachi HD44780 controller. Displays that use the KS0073 are also supported by the drivers.
The excellent drivers provided by Peter Fleury http://homepage.hispeed.ch/peterfleury/ Mr. Fleury's cod is used here and included in the zipped file with his kind permission. Mr. Fleury's code supports both of the popular controllers for this display. The 10k pot, labeled "R7" is used to adjust the display to account for the viewing angle.

The ATTINY816 was not as good a choice for this project as I imagined when I started. besides running out of program memory space, I also ran short on I/O. To solve this problem, I multiplexed three pushbuttons ("Reset", "CAL", and the mis-labeled "D/Q") on some of the LED control and data lines. A latch, the 74HC573 is used to latch data from two of the display's connections, in order to get two additional output lines.

The microcontroller is an Atmel ATTINY861. It runs at 16 Mhz. I had started out thinking I would run it at 20 MHz, but 16 MHz worked out much better in terms of getting a 1 kHz sine wave.

The sine wave is made up of 64 samples per cycle, normalized to a top value of 249 in order to obtain a frequency of 1.000 kHz. The PWM output is on pin 2 of the controller. The 10K pot in series with pin 2 is used to adjust the amplitude of the sine wave. The 10k pot, along with the 10k fixed resistor in series with it, in conjunction with the .047 uf capacitor is the first pole of the series of low pass filters that the PWM sine wave passes through on its way to the inductor under test.

Reference Voltage Buffer

The 1 kHz sine wave from the 7 poles of filtering is used to drive a current source, U2D, and the current source in turn, drives the inductor under test. I used 1% resistors so I could get a suitably high impedance with which to drive the inductor under test without needing to trim the circuit with a pot. The value, 84.5 k just happens to be a value that I had available that was fairly high resistance, but not high enough to invite an high-resistance related problems, like excessive offsets due to opamp leakage currents, or noise pickup.

The 1 K resistor between the current source and the inductor under test, the 1k resistor just to the left of the inductor under test, doesn't really do anything. I had put it in the circuit when trying a 1 k resistor instead of a current source. I think it can be safely replaced with a short.

A buffer, U1A, buffers the voltage across the inductor under test so it can be used to drive the integrator in the next stage. I think the maximum calibration error that would result if this buffer were omitted would be less than 1% because the load is 30k and the impedance the meter measures ranges from near zero to 300 Ohms.

I used a Texas Instruments TLV2774 quad op-amp because it gave the best results of the available op-amps, but I obtained results nearly as good when using an LM324 for U2. when using an LM324, I found that adding a 1K resistor from the output of U2D to ground reduced the harmonic distortion. I first tried this resistor with the expectation that some of the harmonic distortion I was seeing was the result of crossover distortion, and the success of this suggests that crossover distortion was the cause of the harmonic content with an LM324. The TLV2774 is an excellent opamp but it is quite a bit more expensive than the LM324.

U1 needs to be an excellent opamp, such as the TLV2774. The most critical specs are input offset voltage and gain-bandwidth product.

This circuit block samples the sine wave across the inductor under test. The amplitude of the sine wave is proportional to the total impedance of the inductor under test, and the amount the sine wave shifted with respect to a resistive load is used to determine how much of that impedance is from resistance and how much is from inductance.

To determine the phase, the sine wave is sampled at two different phases, each 90 degrees apart and then the angle computed from the measured real and imaginary components. After that, calibration data are used to correct the angle.

A synchronous demodulator is used to sample the sine wave. The operation of synchronous demodulators is pretty well covered on these two web pages:

http://cappels.org/dproj/syncdet/syncdet.html

http://cappels.org/dproj/dlmom/dlmom.html

I opted for a half-cycle phase comparator, sort of half wave rectification of the incoming signal, in order to save parts. With "full wave" sampling, the measurement time could be cut in half. The two web pages referenced above use "full wave" sampling.

Prior to sampling, the integrator is zeroed by taking the Ramp Up/Ramp Down pin high or low, depending upon the state of the output of the comparator, U6, and keeping it high or low until the state of the comparator changes, thus indicating that the output of integrator U1B has crossed the reference voltage. At this point, the Ramp Up/Ramp Down pin is left to float so it doesn't interfere with the measurement cycle that follows. During measurement, 250 cycles of the 1 kHz sine wave are measured in each of four phases. The four phases are the real axis, averaged with the inverted measurement from the negative of the real axis, and the imaginary axis averaged with the inverted measurement of the imaginary axis. Taking a cue from the M³ Electgronix manual, this is done to cancel some of the offsets in the circuit. It was confirmed experimentally that using 250 cycles, which is a harmonic of the 50 Hz power line frequency, made the measurements more repeatable.

To find out how much the average voltage applied to the integrator during each measurement, the integrator is again driven to zero with the Ramp Up/Ramp Down control signal, and the time required to zero the integrator is measured using the on-chip 16 bit timer in the ATINY861. This 16 bit value in the is averaged with several other measurements in a boxcar averager to further reduce the noise in the measurement before being used to compute the resistance and inductance values.

I used an external comparator in the A-to-D converter because the measurements showed less noise than when I used the comparator on the ATTINY861. In a lower cost version, the internal comparator can be used. I the case of the external comparator, the 2.5 volt reference supply is used as the reference, and when using the comparator on the ATINY861, the ATTINY861's on-chip 1.2 volt reference is used.

The Shield
In my experiments, I found that placing the circuit board on a conductive shield, connected to the 2.5 volt reference supply, reduce noise pickup considerably. I had visions of a completely shielded enclosure for the circuit along with a partially shielded pair of test clips, but since this is an experimental circuit rather than a complete instrument, I have not mounted it in a box. Any circuit such as this does need to be carefully shielded, especially from the power lines.

Description of Controls

The instrument only has a few controls:

Reset button
This button resets the micro controller and is also used to start the fundamental calibration sequence. It can also be used to escape a calibration sequence.

Calibration
This button starts the automatic calibration sequence and if pressed immediately after reset, starts the fundamental calibration sequence.

D/Q
This button switches the display mode between displaying resistance and displaying Q. Where Q is defined as reactance/resistance.

Calibration

The instrument needs to be calibrated initially by adjusting one pot and going through an initial calibration cycle with a reference resistor. After that, periodic calibration can be performed by merely pressing the CAL button and letting the instrument calibrate itself using the internal references. While performing the calibrations, it is a good idea to keep your hands away from the inductor under test terminals. If the circuit is not shielded, its a good idea to not move your body around much either, as the inducted changes in the electric field around the can affect calibration.

Initial Calibration

Initial calibration has two parts: The first part is the characterization of the values of the internal reference short and the internal reference resistance, and the second part is the calibration of the instrument by the use of the internal references.

The first thing to do is to set the current waveform to the inductor to about 1 milliamp peak-to-peak. Short the inductor under test terminals, and adjust R7 for 1.0 volt peak-to-peak on the output of U2D. The precise value is not important because this value will be measured in the next part of the calibration.

The next step is the fundamental calibration of the internal calibration, followed by automatic calibration of the instrument. You will need a 30.1 Ohm resistor. A precision, low drift resister is preferred, but lacking that, you can a 30 Ohm 5% resistor, preferably one that you selected to be as close to 30.1 Ohms as possible. You will also need to use two fingers, a finger and a thumb, or to at least be very agile with one of them.

Hold down the CAL button with one finger then press and release the RESET button. The firmware name and revision level should be displayed, followed by this message:             Short pins, then
                            press cal button

Release the CAL button and short the inductor under test pins, then press and release the CAL button.

A display with the following message will be displayed for about 15 seconds.

                           Short.
                       Please wait.

During the time the message above is displayed, the instrument is characterizing the internal short reference. An asterisk (*) will appear part way into this process to indicate that the process is still running. When the characterization is complete, the following message will be displayed.

                          30.1 Ohm on pins,
                        press cal button

When the message above appears, remove the short and connect a 30.1 Ohm resistor across the inductor under test terminals, then press and release the CAL button.

                        Resistance.
                          Please wait.

Again, an asterisk will flash during the process to show that the process is running. After about 15 seconds, the message below will appear.

                            Remove device
                          press cal button

Remove the 30.1 reference resistor, and press and release the CAL button and the message below will appear.

                        Meas. Int. Ref.

Keep your hands away from the inductance under test terminals during this operation, and wait about 30 seconds for the internal reference to be measured and The following periodic calibration cycle to complete. Asterisks will be displayed to indicate that the process is running. When the process is complete the meter will start taking measurements and the "Overflow" message will be displayed.

Now, the internal calibration references have been measured, their parameters stored in EEPROM in the ATTINY861, the meter has been calibrated, and the calibration data stored in EEPROM in the ATTINY861, and the meter is ready to use. After the power is cycled, calibration data in the EEPROM will be used so that measurement can be taken immediately, thought its a good idea to give the instrument some time to settle before either calibrating or taking measurements.

Periodic Calibration

Make sure there is nothing connected to the inductance under measurement terminals and press then release the CAL button. This will start the automatic calibration cycle.

Usage Notes

This inductance meter measured impedance and angle. The maximum dynamic range of the analog circuitry is a little more than 33 Ohms in the low range scale, and a little more than 330 Ohms in the high range scale. Depending upon the phase angle of the voltage waveform with respect to the sampling pulses, higher impedances can be measured. But this is mostly a matter of luck, so don't depend upon it.

Use a low resistance and low inductance connection for the inductor under test terminals. Stray resistance in the hundredths of an ohm can affect the readings and maybe more importantly, calibration.

Low resistances in series with high inductances are not accurate. Some of this is because of an error in correction of the phase angle, but not all of it, and I haven't been able to track this down yet. If you have any insight, please share it with me at projects(at)cappels.org.

Return to HOME (more projects)

Contents ©2008 Richard Cappels All Rights Reserved. Find updates at http://projects.cappels.org.

Liability Disclaimer and intellectual property notice
(Summary: No warranties, use these pages at your own risk. You may use the information provided here for personal and educational purposes but you may not republish or use this information for any commercial purpose without explicit permission.) I neither express nor imply any warranty for the quality, fitness for any particular purpose or user, or freedom from patents or other restrictions on the rights of use of any software, firmware, hardware, design, service,information, or advice provided, mentioned,or made reference to in these pages. By utilizing or relying on software, firmware, hardware, design, service,information, or advice provided, mentioned, or made reference to in these pages, the user takes responsibility to assume all risk and associated with said activity and hold Richard Cappels harmless in the event of any loss or expense associated with said activity. The contents of this web site, unless otherwise noted, is copyrighted by Richard Cappels. Use of information presented on this site for personal, nonprofit educational and noncommercial use is encouraged, but unless explicitly stated with respect to particular material, the material itself may not be republished or used directly for commercial purposes. For the purposes of this notice, copying binary data resulting from program files, including assembly source code and object (hex) files into semiconductor memories for personal, nonprofit educational or other noncommercial use is not considered republishing. Entities desiring to use any material published in this pages for commercial purposes should contact the respective copyright holder(s).