Part 2 - Owens Bridge Inductance Measurement Device
Last time we talked about the basics of an Owens bridge inductance bridge. In this installment, we will describe a real instrument that you can build. It is self contained, except that it does require an external AC voltmeter as the null indication device. This version places a fixed AC voltage across the inductor; namely, the "mains" or "line" frequency through a step down transformer. Also, no DC is inserted into the inductor, but provisions are made to be able to add this feature in part 3.
The circuit is very much like the idealized bridge as described in part 1. R1 of the idealized bridge is the S1/R1-R3 portion of the circuit. This serves as the L*0.1, L*1, L*10 selector in the instrument. C2 of the idealized bridge is C14 in the instrument. That is why C14 should be selected close to the proper value. (See the Calibration section below). R3 of the idealized bridge is SW2, SW3, S3 and R4-R27 in the instrument. C3 of the idealized instrument is SW1 and C1-C12 of the instrument.
The inductance measurement consists of looking for a NULL (minimum reading) on the AC meter by turning S1, SW2, SW3 and S3 (to find the inductance) and also SW1. The SW1 gives you an indication of the resistance of your inductor, but is not intended to be an accurate resistance measurement. Rather, adjust SW1 for a null to help you get a better null of the other switches. Then, the inductance value is read from the switches.
Since Lx=R1*R3*C2 of the idealized bridge, if "R1" is set for 100 ohms, and "R3" is set for 10k, and "C3" is 1uF, the inductance would be 1Hy. Thus, the inductance switch positions are calibrated so that SW2 (1k ohms per step) represents 100s of mH when S1 is set to the 100 ohm position. Similarly, SW3 (10k ohms per step) represents 1Hy per step, and S3 represents 10Hy per step.
The resistance portion simply allows better nulls. In the L*0.1 position, the resistance range to provide complete null is therefore 1 to 147 ohms. In the L*1 position, this would be 10 to 1470 ohms and in the L*10 position, this would be 100 to 14.7k. The capacitors associated with this part of the circuit are not needed to be highly accurate. An ohmmeter will provide a better indication of the resistance of your unknown inductor without a lot more (expensive) accurate capacitors.
In part 3, we'll add DC capability to the unit. When we do this, a fair ammount of power is dissipated in R1/R2/R3 of the schematic. That's why the parts list calls out for power resistors there. If you don't need to measure inductance as a function of DC current, then 1/4 watt resistors will work fine, saving some expense. Similarly, in part 4, we'll be placing larger AC across the inductor, and that translates to more AC signal across R4-R25. That's why those are called out as power resistors. If you won't be doing that upgrade, additional expense can be saved by using 1/4 watt resistors there too. The parts list shown below spells that out.
Operation is easy. Set S2 to the "null" position (S2 serves no important purpose until part 4). Hook up an AC voltmeter and the unknown inductor to the appropriate binding posts. Adjust SW1 for minimum voltmeter reading. Adjust S1, SW1, SW2, S3 for minimum AC voltmeter reading. The unknown inductance is the value indicated by SW2, SW3 and S3 multiplied by the multiplication factor marked on S1.
You can see the voltage across the inductor by flipping S2 to "voltage across ind" position. Note that S2 has to be in the "null" position in order to find the inductance of your unknown inductor.
I've listed two alternates for some of the parts (marked with an asterisk). The first alternate assumes you will be building the unit do that it will ultimately be upgraded to add DC and perhaps variable AC capability later. The second alternative is slightly lower in cost for those who want to build the basic bridge but do not need the added features. I have not included chassis, line cord and hardware items. I'll let these up to your imagination. Source part numbers were from recent catalogs (July 2001) and may have changed by the time you read this. For the capacitors listed, they are actually 250 volt parts. The basic unit can probably be built for less than $100USD, and considerably less with appropriate surplus shopping. In the table below, all part numbers refer to the Mouser catalog (www.mouser.com)
|C7,13,14||3||1.0uF 200V||CDE||5989-250V1.0||01.77||C14 needs to be selected for 1%|
|J2-7||6||"5 way" terms||DGS||164-11102||09.36||All "black". You can use red/blk.|
|R1||1||10 1% 3W
10 1% 0.25W
|R2||1||100 1% 5W
100 1% 0.25W
|R3||1||1k 1% 10W
1k 1% 3W
|R4-14||11||1k 1% 3W||Vishay||71-RS2B-1.0K||06.82|
|R15-25||11||10k 1% 3W||Vishay||71-RS2B-10K||10.56|
|R26,27||2||100k 1% 0.5W||Vishay||71-RN60D-F-100k||00.42|
|R28||1||1M 1% 0.25W||Vishay||71-RN60D-F-1M||00.21|
|S1,3||2||3pos rotary||Mountain||10WW033||03.08||The part spec'd is multiple section. The other sections of S1 will be used in later phases.|
Totals: 73.20 USDor 69.84*USD
Setup / Calibration
S1 is marked Lx0.1 (the switch is in that position on the schematic), Lx1, and Lx10. SW2 is marked 0, .1 .2 .3 ... 1.1Hy. SW3 is marked 0, 1 2 3 ... 11Hy. SW3 is marked 0 10 20Hy. SW1 is uncalibrated and may be marked "also set for minimum AC reading".
The unit essentially does not require any calibration. Standard 1% resistors are used throughout, and this determines the basic accuracy of the instrument. Probably the largest source of error is the one capacitor that also plays a part in determining the inductance value (C14).
Since is difficult to obtain precision 1uF capacitors, there are two ways to ensure the accuracy of the unit. If you have access to a capacitance meter of good accuracy, you can select capacitors until you find one that measures within 1% of 1.0uF. You can also do this by paralleling values until you obtain this. For instance, you can use 2 - 0.47uF capacitors in parallel. Chances are the combination will be slightly less than 1.0uF. You can then start adding smaller values until you get to exactly 1.0uF. As an example, if you measured 0.93uF, you could try a 0.068uF to get close, then perhaps an 0.001uF or 0.002uF to get exactly the right value.
However, if you do not have access to a capacitance meter, you can also try another method. At 60Hz, the reactance of a 1.0uF capacitor is 2653 ohms. (At 50Hz it is 3183 ohms). The closest standard 1% resistor values to those are 2.67k or 3.16k. Simply connect that resistor and your capacitor to be calibrated in series and to the 24 volt AC source (T1). Now, with your AC voltmeter, measure the voltage across the resistor and then the capacitor. Adjust the capacitor as described in the previous paragraph until the voltage measured across the resistor and the voltage measured across the capacitor are the same. Note that these voltages will NOT add up to 24 volts (actually, they will add up to about 34 volts).
One additional note affecting the accuracy/calibration of the unit. Notice that the Lx0.1 range resistor is 10 ohms. This means that the wiring resistance of that path needs to be less than about 0.1 ohms. Keep the leads short and direct, using relatively heavy wire.
|0 to 3.21Hy in 10 mH steps||0 to 32.1Hy in 100mH steps||0 to 321Hy in 1Hy steps|
Inductance is measured at about 5V, 60 (or 50Hz) AC level. [This will be altered in part 4 of this series]. The actual AC across the inductor can be read by flipping S2 to the "AC across ind" position.
0 DC applied in this version, part 3 will add DC capability of about 0 to 400mA on the 3.21Hy range, 0 to 200mA on the 0 to 32.1Hy scale, and 0 to 100mA on the 0 to 321Hy scale.
In part 3, we'll add a DC current source which will allow measurement of inductance as a function of DC current flowing through the inductor. This allows you the ability to judge how much DC current you can pass through an audio transformer without severely degrading its inductance.
In part 4, we'll add variable AC level and frequency. This allows you the ability of determining how the inductance changes with signal level or with signal frequency.