Metal Detector circuit not really working

[lordvader88]

I'm making a little 5 BJT MT


The coil I made is 19.4uH at 100kHz w/ 25.7Q , and I have that Colpitts oscillator working after upping the gain by changing C2/C1.

I'm using an 8MHz crystal, and I manged to get it working after upping the base resistor to 680k and the base cap to 1n.

Without the coil or crystal on the breadboard, the mixer works great as a radio for a local 800kHz station. You can barely hear it until I touch my finger on either base resistor of the mixer, and then it's sound great just from the piezo.

So now both oscillators are working, and if I place a 2.5lb plate weight over the coil, the Colpitts will drop most of it's amplitude. And on the mixer output, there's a complex but repeating wave at the 1.2MHz of the colpitts. And it changes around when the weight is on the coil.

But there's no audio frequency at all, I don't see anything that seems like ultrasonic or up. But IDK, glad I made it this far.

In the guys video he had about 2:1 between osc. Mine is basically random, 8MHz to 1.26MHz = 6.35

In this type of MT, what is it that makes the audio frequency when they mix ? Should I try  for 2:1, what about 3:2 What does the ratio really do ?

For all I know there's a nice ultrasonic sound there

[vk3yedotcom]

As the detector relies on the relationship between frequencies to work, the ratio between them is critical.

A harmonic of your lower frequency must beat against the higher frequency. In other words it needs to be an exact submultiple.

In other words 2:1, 3:1, 4:1 etc. Not 6.35:1.

Otherwise you won't get a beat note and won't hear anything.

[iMo]

For example, if your coil oscillator would be 2MHz (or 4MHz or 1MHz or 100kHz), and the crystal one 8MHz, you would get an audio beat because the metal changes the frequency of the coil oscillator by a few hundreds Hz or a few kHz.

Code: [Select]

2kHz audio beat = 4 * (2MHz-0.5kHz) - 8MHz

2kHz audio beat = 2 * (4MHz+1kHz) - 8MHz
2kHz audio beat = 2 * (4MHz-1kHz) - 8MHz
2kHz audio beat = 8MHz - 2 * (4MHz-1kHz)
2kHz audio beat = 8MHz - 2 * (4MHz+1kHz)

2kHz audio beat = 8MHz - 8 * (1MHz+0.250kHz)
2kHz audio beat = 8 * (1MHz-0.250kHz) - 8MHz

2kHz audio beat = 80 * (100kHz-0.025kHz) - 8MHz
2kHz audio beat = 8MHz - 80 * (100kHz+0.025kHz)etc., etc.

As written above, the "N" in N*(..) must be an integer, as you mix the 8MHz crystal freq with Nth harmonics of the coil oscillator.

[lordvader88]

Ok I'll try and change the caps in the search osc. to get 1 or 2 MHz, or whatever works. For now I'll keep the 8Mhz crystal osc. Thats' pretty high but it's the lowest I have (I plan to order a kit of them). Yes I could switch to a non. crystal osc of course.

I've seen the eq. for a mixer where's it
V1 x V2     =   (A x B) / 2   [cos (2 pi {f1 - f2} t ) - cos ( 2 pi {f1 + f2} t )]

Time to learn it for real


I tried some random inductors, and the closest it came to working was when I would make the mixer lock into the 800kHz radio station loudly, I could get a bit of a Theremin effect going in the background.

[lordvader88]

I changed the caps in the Colpitts osc and its at 2MHz, and now when I place the 1kg of iron over the coil, the frequency increases to about 3Mhz, the amplitude still decreases. For a moment I heard a beat sound

The osc. freq never changed before


Both osc. are about 5Vpp

[iMo]

Mind the audible beat is in a few kHz region.

Your crystal oscillator is stable, say, +/- 50Hz based on temperature and voltage.

Your "2 MHz coil oscillator" is highly unstable, however. It can move +/- a few 100kHz based on temperature and voltage changes and/or capacitive or inductive coupling easily..

Add the freq changes caused by large metallic objects.

In order to get something of a practical use, you must be able to "zero" the coil oscillator's frequency such you get a "zero beat" just before the actual measurement.