BFO Revisited

[Duncan]

Hi all,
I have been thinking about revisiting the classic BFO circuitry which originally used the outputs of two oscillators say, at 100KHz, mixed and detected to produce an audio beat note. One oscillator was the "reference" oscillator, whilst the other, the "search" oscillator had the search coil as part of its frequency-determining circuit.

When the detector was set up and in use, when the search coil found a metal target, the pitch of the audio beat note increased or decreased, depending on which side of the beat note the detector was tuned to.

The electromagnetic field from the coil induced eddy currents in the metal target and changed the inductance of the coil, which had the effect of changing the frequency of the search oscillator and consequently, the beat note.

My idea is this. If the change in inductance at 100KHz was sufficient to effect a beat note change of 10Hz, what would be the result of beating the 100th harmonic of 100KHz with a 10MHz crystal controlled reference oscillator? My reckoning says that for the same change of 10Hz at 100KHz, the result at 10MHz would be 1000Hz change in the beat note.

What do others think? I would be pleased to hear your views.

Duncan

[Skippy]

Take a look at the ETI561 project (details are on this website), which achieves a similar effect. It digitally mixes a xtal reference that is 8 times the LC sense oscillator to produce an audio output that varies at 8 times the LC frequency. Clever. Actual frequencies used are 110kHz LC, mixed with 880kHz Xtal ref. If you went too sensitive you would see unwanted effects, eg. audio changing in response to: capacitive coupling with the surroundings, temperature drifts of components/environment, mechanical vibrations etc etc.
Regards.

[dfbowers]

Take a look at the ETI561 project (details are on this website), which achieves a similar effect. It digitally mixes a xtal reference that is 8 times the LC sense oscillator to produce an audio output that varies at 8 times the LC frequency. Clever. Actual frequencies used are 110kHz LC, mixed with 880kHz Xtal ref. If you went too sensitive you would see unwanted effects, eg. audio changing in response to: capacitive coupling with the surroundings, temperature drifts of components/environment, mechanical vibrations etc etc.
Regards.

See page 7 of Carl's article.

[Skippy]

Interesting article ,that, I'd not seen it before. I studied the BFO projects on this site because I was building the Scarborough/EPE mini pinpointer, and was looking at ways to improve the (poor) diode-OR mixer. I decided after testing the breadboard version that a multiplying mixer wasn't needed, but substituted a tiny-logic XOR gate (SOT23-5 package) which improves the audio volume, as the XOR's output has a 50/50 duty cycle, compared to the 75/25 of the OR type mixer.
The references to capacitive coupling/ temperature stability etc are all standard problems of any sensitive instrument - the EPE pinpointer exhibits all these problems, being a simple design, and handheld. I found graphited paper shielding inside the case was needed to provide Faraday shielding. It does work, though, just detecting a 25mm copper coin at 50mm.
While BFO's make an interesting project, I know a couple of 'old-timers' who used them in the 1970's; they tell me they were awful things!

[CrotalusAtrox]

A small BFO circuit

I'm nearly finished with this one. It's based on the project 561 design and an similar one that I made. I'm just using it to find native copper in an open pit mine dump.
I'll let you know how this rev 2 works in a week or two.

I don't recommend trying to reproduce it unless you have experience with small SMD parts. Those varactor diodes are pretty hard to find, too.

PJ

[CrotalusAtrox]

Yes, BFO detectors are obsolete, but I had fun building this one. It does manage to find large chunks of copper, which is all I need at the moment.

I had one PCB screwup - a trace was accidently deleted.

The oscillator stability is pretty good. The drift is tolerable using a 16x reference frequency. Power consumption is low.

I think 16x is probably about the practical upper limit for this scheme.

With a shielded 150mm coil, in air, I can detect:
penny at 4.5"
beer can at 12"
jeep at 20"

PJ

[Geo]

So nice pcb for a BFO detector????
It is crazy

[CrotalusAtrox]

I'll make the next one a PI.

PJ

[kt315]

agree with Geo. you are able to solder SMD board in time much people are fearing to look at this... crazy man. I have HH boards only for $5. at the board you can do what you want to get... Garrett PI, Surf PI, and someone asked Tesoro Shark... but BFO... that is crazy matter for a world with Bounty Hunter at sale price for $15...