- #1
r_pogo
- 1
- 0
I'm an old digital guy working on a little analog oscillator project so I'm a bit in over my head. The objective is to detect alloy deviations in samples of aluminum scrap: a very simple go/no-go detector
I've cobbled together an inductive meter design using a dual op-amp sine wave generator feeding a single coil Maxwell-Wien bridge, followed by a difference amplifier, then a full wave rectifier then a low pass filter, leaving a DC voltage to feed a couple comparators which power some display LEDs. A "known good" sample is placed on the coil, the bridge is nulled via a pot, then the test sample is placed on the coil for comparison. Green LED is good, red LED is hold for test. Easy, so far
But the design of the of the coil is a bit confusing.
The op amp powering the bridge can deliver 20 mA at +/- 8 Vdc using +/- 10 Vdc rails. The coil form will be in contact with the sample so there will probably be a minimal 2 mm distance offset or gap between the coil and the sample. The sample will be about 3-4 mm thick so the there needs to be good flux penetration through the sample while avoiding flux punch through to the other side. This suggests a low frequency. The sample will be a minimum of about 30 mm square which should suggest a nice big outside diameter coil for good sensitivity.
At this point it would seem that the factors controlling the optimum coil design are more mechanical than electrical. Any suggestions on the optimum dimensions and inductance? I'll be using 18 gauge magnet wire and wind the coils myself so they can be "tuned" if required.
I can easily match the bridge capacitance and resistances to match the coil to get the null so those specs should follow the optimum coil inductance.
Thanks in advance...
I've cobbled together an inductive meter design using a dual op-amp sine wave generator feeding a single coil Maxwell-Wien bridge, followed by a difference amplifier, then a full wave rectifier then a low pass filter, leaving a DC voltage to feed a couple comparators which power some display LEDs. A "known good" sample is placed on the coil, the bridge is nulled via a pot, then the test sample is placed on the coil for comparison. Green LED is good, red LED is hold for test. Easy, so far
But the design of the of the coil is a bit confusing.
The op amp powering the bridge can deliver 20 mA at +/- 8 Vdc using +/- 10 Vdc rails. The coil form will be in contact with the sample so there will probably be a minimal 2 mm distance offset or gap between the coil and the sample. The sample will be about 3-4 mm thick so the there needs to be good flux penetration through the sample while avoiding flux punch through to the other side. This suggests a low frequency. The sample will be a minimum of about 30 mm square which should suggest a nice big outside diameter coil for good sensitivity.
At this point it would seem that the factors controlling the optimum coil design are more mechanical than electrical. Any suggestions on the optimum dimensions and inductance? I'll be using 18 gauge magnet wire and wind the coils myself so they can be "tuned" if required.
I can easily match the bridge capacitance and resistances to match the coil to get the null so those specs should follow the optimum coil inductance.
Thanks in advance...