Monday, March 8, 2010

Matchless Metal Locator

This is one of my best super-simple metal detector designs, which a publisher generously donated to the public. Before this design came along, induction balance (IB) metal detectors typically used 50-100 components. The simplicity was achieved through several tricks -- but in summary, I stripped out the usual analogue electronics. Copyright restrictions prevent me from publishing the design on my blog, but it may be downloaded free here: Matchless Metal Locator. OBSERVATION: What I can offer on my blog is nearly ten years of hindsight. I would now improve the design, most importantly, by increasing the battery voltage to 12V or 15V, and using a 9V regulator. This would greatly assist stability. The biasing around IC1 pins 8 and 12 could be simplified by combining VR1 and VR2 in one 500k component. And if I had the chance, I would like to apply a wobbulator to the IC1b input circuitry. This has been a popular design. It went into production in Australia, in two versions.

17 comments:

Jenny Hillebrand said...

Wobbulator? For real?

Thomas Scarborough said...

Very sensitive circuits run into trouble at the boundary between detection and non-detection. A well known example of this is something called frequency lock. By introducing a slight frequency-wobble into a signal (hence "wobbulator") one can improve detection. Similarly, one may improve detection by introducing random noise. You can try this in Church. Hold steady notes as you sing in the choir, and no one will notice you. Sing an octave higher or lower, and ditto. But wobble your voice, and everyone will pay attention! :-)

markpenrith said...

Hahaha, I'm with Jenny: A Wobbulator :).

Thomas Scarborough said...

Philistines all!

jare said...

Mr. Thomas did EPE Bounty Treasure Hunter (160mm-240mm detects old Victorian peny) is better than the matchless Metal Locator (110mm-150mm)

Thomas Scarborough said...

I ask you. How much trouble can one get into over one post? ;-) The EPE Bounty yielded better performance. However, constructors experienced difficulty with the setup. The Matchless was conceptually elegant -- a nice minimalist design -- a design which broke with tradition -- a design which generated excitement -- and at that time it was the simplest self-contained metal detector (I improved on that with my BB designs). This is why it is one of my favourites. Incidentally I usually state 10% less performance than I actually achieved with a design.

jare said...

Ok mister Thomas do not be angry.More only one question: What would you improve the EPE Bounty (except the stabilizer 7812)

Thomas Scarborough said...

The EPE Bounty had three important adjustments, VR2, VR3, and VR4. These needed to be adjusted in combination with each other, which made the procedure potentially difficult. VR4 would best have been a multiturn preset potentiometer, for better precision. Yes, the circuit would perform better with a regulator, and maybe a crystal earpiece instead of the piezo sounder.

Thomas Scarborough said...

And I would try a wobbulator, experimentally. :-) My guess is that it would significantly improve performance.

jare said...

Thanks for the reply.

kristina said...

Wobbulator theoretically: "You can try this in Church. Hold steady notes as you sing in the choir, and no one will notice you. Sing an octave higher or lower, and ditto. But wobble your voice, and everyone will pay attention!".How wobbulator looks practically?Set the scheme on your blog.

Thomas Scarborough said...

Kristina, thank you for the question. I have provided an example of a wobbulator at http://thomasscarborough.blogspot.com/2010/03/bfo-with-wobbulator.html In the early days of electronics, these used to be mechanical devices -- variable capacitors turned by motors.

f3l said...

Thomas, can I ask why you advise against the NE556 in the article?

I'd like to try this design but I only have plain old NE555s available.

Very best regards, Iain

Thomas Scarborough said...

Than you Iain. Note that the NE556 and NE555 devices are bipolar. You need the CMOS 7556 device (or two times 7555 -- see below -- as the 7556 is a dual 7555). The difference is that CMOS devices have a very high input impedance, while bipolar devices typically do not. Bipolar devices tend to greedily swallow input signals, while CMOS are far more delicate with them -- important with this metal detector. The CMOS 7556 is shown in the design, however you could use two 7555's. In fact this might be better, as you then have separate supply decoupling (that is, capacitors wired across the power terminals of the IC's). Supply decoupling is important with this circuit, as it prevents frequency lock. To put it simply, it prevents a "wobble" in the power supply which might disturb the circuit in operation, and could improve sensitivity say 10%.

f3l said...

Thomas, thanks so much for your detailed reply. Aha, so it's the input impedance that's the problem. I may try an opamp preamp stage then instead of the BJT; I think this should alleviate that issue. I'll let you know how I get on! Best regards, Iain

f3l said...

Thomas, just to add that one very quick Farnell order later, I now have all the correct parts for both the Matchless and the Bounty; so I can start by trying them per spec :) One happy weekend of building ahead; many thanks for the great designs. Best, Iain

Anonymous said...

how to increase a 9v voltage battery to 12v? since 12 v are for cars