HH tunning questions

[Bill]

I finally finished my stock HH and got it working (sort of). I am using a 500uh(498uh measured) 7.25" diameter coil, 26AWG, shielded. My coil is hooked up to the board via dual conductor shielded microphone wire. The shielding is grounded to the battery ground. When I power it up the threshhold pot really doesnt effect tunning very much, pulse width and freq have a lot more impact. Anyway, once I tuned it for just a few minutes I can detect my platinum/gold wedding band at 11". I am very happy with the early performace of this unit. But for some reason, when I hold the ring close to the coil (5" or so) the things frequency goes super high and stays there until I remove the ring. I thought it was supposed to stop if the coil was held still over the target and only do stay high when the autotrack was on. I still have not inserted a switch to use that part of the circuit. What could be causing this? Should I increase the value of the threshold pot as well? I am doing this in my apartment where I have a wireless router and alot of other noisey electronics.
Thanks for the help guys!
Will

[Bill]

Well I figured out why the autotrack was being problematic, the short I soldered in lost connection. That explains that!
Should I have used the battery ground for shielding or should I use the VB- ?

[Carl-NC]

Shielding is connected to "ground", not "-VB". Glad to hear you got it up & running!

- Carl

[Bill]

What kind of sounds am I looking for when I have it running with no metal near the search coil. I get a few clicks here and there, its not constant at all. And there seems to a higher tone beep every now and then at a regular rate. The unit still detects objects though, there really isnt any slope of freq increasing though, it goes from a slight, and I mean slight, inconsistant crackle to high pitch ring. What should I be hearing? I expected a low constant tone with a noticable freq increase slope. Could all the noise in my apt building be causing havoc?
Will

[Carl-NC]

You should be able to adjust the threshold to get a constant low pitch, somewhat like the low growl of a BFO. If you are getting sporadic beeps, it is probably due to interference. Indoor testing will do that. If the response goes straight to full bore, there is too much gain. Try taking down the sensitivity pot.

- Carl

[Bill]

Thanks Carl! Yes, I was expecting a slow constant tone, now it goes from a few sparadic clicks and burps to silence, and repeats, a high pitch beep in the backround happens at regular intervals. I have a suspision that its the interference, I live in an apartment in the city(Houston), I and most of my neighbors have wifi, so thats more than likely the problem. I am going over to my folks house, they live in the middle of nowhere on a few acres so there should be minimal noise. I hope that will do it. As for the very little freq slope which pot is the sensitivity pot? Is it R35? Right now I think I have it at under 2Kohm. I just purchased a osscilliscope for my metal detector projects so I can get a better view. I quickly found out this is a must have when dealing with these things. Thanks for the help Carl!

Once I get this unit running I am going to take it to our deer lease, we have an old wagon trail that runs from east to west. I alreay have found a few old knife blades and silverware just by walking around, cant wait to see what I find with the HH!

Will

[Bill]

HH mod article comming nicley.

After speaking with a few members, gathering info, and trying things on my proto boards, my HH Nugget shooter is comming along nicley. I am using a krynar wire coil, 10" 290uh (mono) with 3M 1190 tape (Cost $$$$$) and plastic tubing for spacing. I decided to use the mono for both units. I am buying one of Gary's pancake type coils and will give it a try as well. The stock HH is functioning well and I am very impressed with it. If I can keep the same depth that it has, and just decrease sample time to about 8usec I will be very happy. I will post all my findings/notes/sources as soon as I finish it. Thanks for all the help guys!
Will

[bbsailor]

HH mod article comming nicley.

William N

To get the delay down to approximately 8uS you will need to reduce the value of R42a, on the Hammerhead Schematic, by a few hundred ohms. If you find that when you reduce the delay, that the machine stops responding to targets, you have reached the limit of the coils response. The quickest fix is to reduce the total capacitance of the coil, coax wire or MOSFET by a 100 or more pfs. A double D coil may be a little faster as the RX coil does not see the MOSFET capacitance. Try to use some of the MOSFETS that Reg suggested (do a forum search) that are less than 50 pf in the output. Another trick to try to squeeze out a faster response is to increase the value of R12 to about 2K. This will reduce the gain but also reduces the effective loading of R12 on the coil. This is a simple, quick mod that is worth a try.

Once you get the machine working take some measurements and post your results and I'm sure you will get come constructive comments.

Which version of the Hammerhead did you build?

Here are the recommended measurements to take.

Coil self-resonance, including cable (try to use the method recommended by Eric Foster by inducing a pulse into the coil under test. This method avoids the probe loading which can throw the self resonance off (lower) by 25Khz or more.)
Coil resistance
Coil to shield capacitance
Coax wire used and length
MOSFET part number used
Battery current draw at minimum and maximum pulse width and minimum and maximum frequency
Maximum detection distance for a U.S. nickel and man's gold wedding band
Maximum peak flyback voltage at minimum and maximum pulse witdh settings
Minimum pulse delay in uS
Damping resistance value (R11)

Others who have built the Hammerhead can then compare similar measurements and provide objective comments.

I did some mods to my Cscope CS6PI and got the response down to 7.5uS at 13.5K PPS and a detection range of 15" to 16" on a nickel at my workbech which is full of noise. I mount the coil on a wood shelf (no metal nearby) and have a small plastic basket that I place targets in. I then swing the basket from a string, over the coil at various test heights and listen in the headphones for a signal. It is pretty crude but it works.

You can actually solder a ground wire to 3M 1190 copper fabric tape. Just put a small piece or metal under it to act as a heat sink when you solder. Cut a small piece off and practice first.

Some of my next experiments are to use an e-bay purchased lock-in amplifier to take the place of IC8A. I will feed the time-differentiated outputs of IC7a and IC7b into the differential inputs of the lockin amp, then feed a synchronizing signal off the main clock to supply the lock-in signal and see what happens. I can adjust the integration time from 1 ms to several seconds to see how much signal I can extract from the noise. At long integration times I realize it becomes impractical for actual detecting use but I just want to see where some limits are located.

Unfortunately (or fortunately) sailing season is starting and I may put this one on the back-burner while I get my sailboat ready and move in. I may pick this up in the Fall.

These are just some thoughts to ponder for the curious PI tinkerers.

bbsailor

[Bill]

Thanks for the info BB! I will put all of those measurements here as soon as I round them up all into one place. I took my finished HH out to a remote beach last weekend. I used a fanny pack to hold the actual detector unit along with the 3 Ah gel cell. I really like the unit configured this way. It took me about an hour to get the settings configured to maximum dept and sens and get used to it. I am VERY pleased with its performace. I dug every target and found tons of trash but it was fun! I also found a Bra (metal wire in it), belt bukle, car keys, and some coinage. The depest target I doug was at least 24"(belt buckle). The add on amplifier worked flawlessly. I did run into a problem with the detector cable though. Every time I bumped the cable or moved it to quickly it gave a false signal. A little different sounding than what an actual target gives but still a loud signal. This is the shielded cable running from the coil to the housing. Any thoughts on how to fix this problem.

Will

[strujas]

Does anybody have directions for use of ranger tell examiner?my friend had build one but we dont know how to use it....

[Unregistered]

try to use the method recommended by Eric Foster by inducing a pulse into the coil under test. This method avoids the probe loading which can throw the self resonance off (lower) by 25Khz or more.)
Does anybody know this method and do i have to do?

[bbsailor]

To use the method that Eric Foster recommends do the following.

1. Connect a square wave pulse generator or a PI metel detector to a coil that will be used to stimulate another coil under test.
2. Place the coil under test near or against the stimulation coil. This will transfer energy to the coil under test by induction.
3. Connect the coil under test directly to the scope.
4. Adjust the scope amplitiude and frequency to see an induced, declining ringing set of frequency oscillations in the coil under test. If using a PI detector to stimulate the coil and the PI is operating at under a few hundred cycles, it might be hard to see the ringing. In that case, use a square wave pulse from a pulse generator at about 1KHz to 10 KHZ to better see the ringing in the coil under test. Any chage in frequency to the stimulation coil will not chage the induced ringing pulses in the coil under test.
5. Measure the period between peaks on the ringing cycles and you will then be able to calaculate the unloaded resonant frequency of the coil under test, without the loading of the probe lowering the measurement.

Here is some additional information that I find useful.

Another method recommended by Gary is to place a 1 to 5pf capacitor in series with the coil uner test and use a probe connected to this series capacitor. This method can be useful with any scope probe to minimize probe loading. This series capacitor can be made by twisting two pieces of hookup wire together for about an inch. This will make a gimmick capacitor that will limit the probe loading on the coil under test to about 1 to 5 pfs. Just use a signal generator with a 1M ohm resistor in series with the coil to stimulate and sweep the coil between 500 Khz and 1.2MHz and look for the resonant rise on the scope signal at the peak resonant point. Place the coil in mid-air, away from metal objects, to obtain an accurate self-resonance.

On my coils I calculated the loading effect of my probe on the coil that I measure. My 10X probe lowered the self-resonance by about 25Khz to 30Khz in the self-resonant area of about 700Khz to 1Mhz range (depending on your particular probe). I just add 30 Khz (for my probe) to any measurement in this range to get a quick, (more) accurate self-resonant frequency. Make your measurement directly on the coil as a coax cable can add anywhere from 50 to 100pf to the coil circuit and will lower the coil's self resonant frequency by quite alot.

When doing measurement, do all your measurements the same way so you can see the effect of any changes that you make in coil wire size, wire type, insulation thickness, coil style, dielectric constant, number of turns, coil diameter, shield spacing from coil, or other variables. Keep good notes and you will begin to develop a good intuitive sense for how changes in coil design can change the overall coil performance. Remember, that higher self-resonant PI coil frequencies require higher values in damping resistors. This is good for making faster coils that can operate at lower delays.

I hope this helps?

bbsailor