Greetings,
OK! I have attached the schematic for the VCO version. It is the one in the pictures on the other postings. As I indicated in my first posting it is a take off from Carl Moreland's Hammerhead design. As I go through the circuit I will explain the differences and the reasons behind them, right or wrong.
I will start with the coil pulse generator and the coil. The coil pulses are generated using a Cmos NE555 timer. As you will notice, I again used a diode to isolate the charge and the discharge currents in the timing components. This eliminates any interaction between the pulse with control and the frequency control. The transistors that drive the coil driving FET are connected to provide active coil pulse turn off. Both Reg Sniff and Carl Moreland favor this set up and so do I as the fast turn off results in a faster and higher fly back spike pulse. The IRF740 FET that drives the coil has a reasonably high breakdown spec of 400V. The coil winding information is on the schematic and the resulting coil has around a 2 MHz self resonate frequency and a high "Q". With either the 10" or the 8" coil I have to keep the coil pulse to no more than 50 us or the coil fly back spike with exceed 500 V quickly and break down the FET. At 50 us pulse width the fly back pulse spike is about 400 volts. I am counting of the safety margin in the specs for the FET by using a 50 us pulse width. I never come close to reaching the calculated saturation current in the coil at only 50 us. Fast high "Q" coils and fast pulse turn off are necessary to detecting small gold targets. As you can see, this part of the circuit is pretty much like a lot of other PI machines.
From the coil through the receiver preamp is where some major differences start showing up. The preamp Op amp is a NE5534 and is noted for its low noise and ability to recover from overloads quickly. Notice that I set the DC gain of this stage at only 212 with the 4.7k input resistor. With the lower gain I can get my Target sample in further on the fly back spike settling curve where small gold signals show up. The lower gain also reduces the ground mineralization signal, that is at its highest in that area and prevents overloading the preamp. If the preamp gets overloaded from the ground signal you have to increase the target sample pulse delay to get your sample out of the overloaded area to detect anything. When you do this you will loose the ability to detect the tiny gold targets. Normally you will still be able to detect normal size gold rings, coins and such. So far, knock on wood, but I have not run into any place that has overloaded my preamp and I have been able to detect small gold everywhere I have been. The third value in reducing the preamp gain with the 4.7k input resistor is that it reduces the pulse power loading on the 4.7k input resistor and reduces the current through the clamp diodes. The last thing associated with the receiver preamp is that I added a high pass filter between the preamp and the gating IC. There are two reasons for the filter. First is it AC couples the signal and makes it very easy to set up the detector without the setting of one stage effecting the other. Second, I set the 3db roll off of the high pass filter to be at 159 Hz. This is high enough above 60 Hz to provide some attenuation of 60 Hz power line noise but low enough to still provide good signal integrity in the area where I want to sample for my small gold targets. I have also attached a picture take of my oscilloscope showing the output signal of my Preamp and the target sample Gating Pulse. It shows that I am sampling in at 10 us from the end of the coil pulse. If you are not familiar with the wave forms associated with a PI machine, don't worry about it. It only shows that I am sampling in the right place to detect small gold targets.
We will stop there for now and see what everybody thinks. Please feel free to give feedback. We will carry on further into the circuit in the next post.
God Bless and HH!
Smitty II
OK! I have attached the schematic for the VCO version. It is the one in the pictures on the other postings. As I indicated in my first posting it is a take off from Carl Moreland's Hammerhead design. As I go through the circuit I will explain the differences and the reasons behind them, right or wrong.
I will start with the coil pulse generator and the coil. The coil pulses are generated using a Cmos NE555 timer. As you will notice, I again used a diode to isolate the charge and the discharge currents in the timing components. This eliminates any interaction between the pulse with control and the frequency control. The transistors that drive the coil driving FET are connected to provide active coil pulse turn off. Both Reg Sniff and Carl Moreland favor this set up and so do I as the fast turn off results in a faster and higher fly back spike pulse. The IRF740 FET that drives the coil has a reasonably high breakdown spec of 400V. The coil winding information is on the schematic and the resulting coil has around a 2 MHz self resonate frequency and a high "Q". With either the 10" or the 8" coil I have to keep the coil pulse to no more than 50 us or the coil fly back spike with exceed 500 V quickly and break down the FET. At 50 us pulse width the fly back pulse spike is about 400 volts. I am counting of the safety margin in the specs for the FET by using a 50 us pulse width. I never come close to reaching the calculated saturation current in the coil at only 50 us. Fast high "Q" coils and fast pulse turn off are necessary to detecting small gold targets. As you can see, this part of the circuit is pretty much like a lot of other PI machines.
From the coil through the receiver preamp is where some major differences start showing up. The preamp Op amp is a NE5534 and is noted for its low noise and ability to recover from overloads quickly. Notice that I set the DC gain of this stage at only 212 with the 4.7k input resistor. With the lower gain I can get my Target sample in further on the fly back spike settling curve where small gold signals show up. The lower gain also reduces the ground mineralization signal, that is at its highest in that area and prevents overloading the preamp. If the preamp gets overloaded from the ground signal you have to increase the target sample pulse delay to get your sample out of the overloaded area to detect anything. When you do this you will loose the ability to detect the tiny gold targets. Normally you will still be able to detect normal size gold rings, coins and such. So far, knock on wood, but I have not run into any place that has overloaded my preamp and I have been able to detect small gold everywhere I have been. The third value in reducing the preamp gain with the 4.7k input resistor is that it reduces the pulse power loading on the 4.7k input resistor and reduces the current through the clamp diodes. The last thing associated with the receiver preamp is that I added a high pass filter between the preamp and the gating IC. There are two reasons for the filter. First is it AC couples the signal and makes it very easy to set up the detector without the setting of one stage effecting the other. Second, I set the 3db roll off of the high pass filter to be at 159 Hz. This is high enough above 60 Hz to provide some attenuation of 60 Hz power line noise but low enough to still provide good signal integrity in the area where I want to sample for my small gold targets. I have also attached a picture take of my oscilloscope showing the output signal of my Preamp and the target sample Gating Pulse. It shows that I am sampling in at 10 us from the end of the coil pulse. If you are not familiar with the wave forms associated with a PI machine, don't worry about it. It only shows that I am sampling in the right place to detect small gold targets.
We will stop there for now and see what everybody thinks. Please feel free to give feedback. We will carry on further into the circuit in the next post.
God Bless and HH!
Smitty II