Hammerhead Works but not detecting Deep. Help please.

[Lucas Kazuaki]

Use a small pot in Parallel and series with other resistors. Example:
Put a 1 or 2k fixed 1/2W resistor across coil (primary dumping R) then a 5k pot in series with 100-330 Ohm resistor- this is then put across the 1/2W resistor and is adjusted for critical damping.Remove Pot/series and measure the resistance of the combination to find the optimal value that goes in parallel with primary R.

Circuit in this thread: http://www.geotech1.com/forums/showt...aming-resistor

Are you measuring the damping at the Coil out or the pre-amp out?

You may want to try an un-shielded coil to see if the shield is the problem as 6666 suggests.

Another area of the circuit that is critical in the integrator stage. Ensure the 100nF caps are good quality. Next is to ensure the first and second Samples are far enough apart, at least 35usec and up to 100usec (remember that the second sample subtracts from the first).

Is the Coil Current leveling before reaching a peak?
Measure the Pulse ON waveform across the 1 Ohm MOSFET series R without the 1000uF cap.
With short pulses this will ramp up (looking like a triangle) to a peak (Ohms law and R value give the coil current).
With longer pulses the Voltage should level out indicating the coil current is near maximum.
Some experiments I and others have done show that having the coil current level out even with a lower peak coil current improves sensitivity (distance).
This to an effect of the Coil Tau (L/R) and adding a little series R to the coil. This allows the Coil current to level out in a shorter pulse.

Last- The Soil can have a great effect of detection depth so hard to compare to other's results.

I put my probe on the output of the pre amp but i also do check the input side sometimes.

the capacitors are polyester type. i heard that ceramic arent very stable for that operation.

the samples are in that range yes. about 100us. what is the effect of the secondary delay? why keep it between 35 and 100? here i tested it being really close to the primary sampled pulse and also tested it being far with the same results...

I will try experimenting like you said and check the current on the coil.

UPDATE/EDIT : i did test on the 1 ohm resistor like your said. it takes about 80us to charge and level the coil current and yes they look like sawtooth waves. ramping up (charging the magnetic field) then a sudden drop where the mosfet cuts the current

[Lucas Kazuaki]

Is the Threshold low enough?
Should be adjusted so there is slight 'chatter' or tone.

I keep the sound low but not completely silent. Just like comercial PIs, it should be a low like a whisper but still hearable.

[Digger429]

playing with the delay time from 10us to around 60us

You want the first sample delay to be as small as possible, so as much of the signal from the object can be sampled before it dies away.
By "as small as possible" I mean under 10uS. 8uS is better, 6uS is great, 4uS would be fantastic but probably not attainable. The lowest achievable limit is limited by the coil, whether is it a "fast" coil or not. In practice you test just by trying smaller and smaller sample delays until the metal detector doesnt detect anything any more, then you know how fast your coil is.

the secondary delay? why keep it between 35 and 100?

The secondary sample acts as a reference standard comparison for local electromagnetic interference.

Any time after the signal from the sample has completely died away is fine, in theory the longer the better, but in practice that means any time after 50us is good. I wouldn't expect any difference at all between 100uS and 200uS.

For your testing on figuring out why you don't get much depth, use a coil of about 300uH, a sample delay of 8uS, a secondary sample delay of 100uS, and post pictures of the trace you get on the oscilloscope of the preamp output.

---

As for damping resistor, my own testing indoors shows maximum range if it is slightly underdamped, as seen on the image in message 14 of this thread:
http://www.geotech1.com/forums/showt...mping-resistor

[Lucas Kazuaki]

For your testing on figuring out why you don't get much depth, use a coil of about 300uH, a sample delay of 8uS, a secondary sample delay of 100uS, and post pictures of the trace you get on the oscilloscope of the preamp output.

Here it is Blue is the decay, yellow is primary delay. i couldnt get it to 8us but i got it to around 11us or 12us its probably because the limiting resistor in series with the trimpot. will have to change it later.
Finally the white is the secondary delay (100us delay).

[waltr]

With the Decay out at about 20usec I would move the Delay for the first sample out to about 20usec.

With the scope running as in your picture try moving metal near to the coil. You'll see that the rising edge of the decay (blue trace) does not move much but as it get above -2V you'll see movement on the trace. Put the First sample there.
Also note how much the trace in moving where the second sample is being taken. If trace moving then move second sample out a bit more.

[Digger429]

With the Decay out at about 20usec I would move the Delay for the first sample out to about 20usec.

If it was me, I'd be leaving the timings where they are and just changing the damping resistor.

IMHO the curve looks overdamped. As you move towards a more underdamped state the decay becomes quicker as seen in the image of the other link I posted earlier. For a more underdamped state of the one we have here, I suspect it would look like this: (with the hand drawn red line)

[kt315]

to cover with aluminium foil. I know its not the best

hard and interest thing - writing to people do not use the foil in PI coil but they use it again again and again.
do you see at least once Eric Foster used foil in PI coil? HEY THERE!

[Lucas Kazuaki]

hard and interest thing - writing to people do not use the foil in PI coil but they use it again again and again.
do you see at least once Eric Foster used foil in PI coil? HEY THERE!

I didn't find anything to use. only foil. i tried finding some kind of mesh or braid materials but all the stores that i went didn't even know what i was talking about

[kt315]

use kontakt chemie graphit 33

https://www.google.lt/search?q=konta...w=1000&bih=888

at last time i add graphite powder - that gives better result and economy of the spray.

[Lucas Kazuaki]

Thanks for all the replies. They were really useful. Although the detector is still not going as deep as I expected it is working which is the most important. Maybe i will create a new circuit with all those changes in mind...

Today I will be traveling and will bring the detector with me to the real test. No more air tests, now it will see real soil.

Thanks again to everyone

[Carl-NC]

Hammerhead was never intended to be a performance design, it is a learning platform. As such, I specifically chose easy-to-source parts and simple construction (like foil shielding) to keep the project attractive to novices. I make no apologies for that. If you want performance, there are better public-domain designs available now and better construction techniques that will get you there.

All that said, if your posted waveform is off the preamp then it looks slow. Your minimum practical delay time appears to be about 25us. What opamp did you use for the preamp? What diodes for the input clamps? Is this built on a PCB?

[Lucas Kazuaki]

Hammerhead was never intended to be a performance design, it is a learning platform. As such, I specifically chose easy-to-source parts and simple construction (like foil shielding) to keep the project attractive to novices. I make no apologies for that. If you want performance, there are better public-domain designs available now and better construction techniques that will get you there.

All that said, if your posted waveform is off the preamp then it looks slow. Your minimum practical delay time appears to be about 25us. What opamp did you use for the preamp? What diodes for the input clamps? Is this built on a PCB?

It is on a pcb. i am using the diodes and op amps that the project recommended.

[Carl-NC]

Well, something is slowing it down. As others suggested, you might want to kick up the damping resistor. I generally found that 680 ohms worked well for this design, assuming L=500uH.

[Lucas Kazuaki]

Well, something is slowing it down. As others suggested, you might want to kick up the damping resistor. I generally found that 680 ohms worked well for this design, assuming L=500uH.

Now i am on a farm really far from any scopes or any other eletronics equipment.
the last thing i did as suggested by the others was to increase the damping resistor but i noticed that the decay was the same and weird spikes started to show up in the firsts microseconds (0 until 5). so i looked with the scope on the input and it was just the coil that started to ring meaning the resistor i was using before was near the limit that i could be without any ringing.
the value that is installed now is 600ohms. pretty close to 680...

i noticed that using the detector where i am now is practically useless.
the soil is affecting the coil A LOT. i had to keep sensitivity, threshold, auto track all the way to minimum values. and delay had to be set way farther than the 10us where the performance was best. this way the detector worked but the maximum depth was less than an inch.... the soil effects are really strong here. I built the hammerhead thinking that the autotrack feature could minimize the soil effect but it looks as crazy as when i use my old manual tuning pulse induction.
Maybe i am just sitting on a huge gold deposit, i cant lower my coil near the soil.