Help with Garrett ADS Deepseeker Upgrades

[mikebg]

Modification of Garrett TX

How should TX section of ADS work?

The circuitry must "dismodulate" TX current. Parasitic parametric modulation of TX tank occurs when TX coil moves. The ground acts as a mobile core for TX coil, whose distance is modified. Amended and conductivity properties and permeability of this core. TX current meets a variable coil impedance. Maximal modulating frequency is by the order of 12Hz. It is obtained by rapidly waving small sensors.
Minimal modulating frequency is about 0,1 Hz. It occurs in slow motion of EMI sensor at pinpointing.
Unfortunately, the same bandwidth has modulation of the ground signal, which is caused due to a change in the distance ground - RX coil.
From these observations follows that RX coil receives two signals modulated by ground:

Signal A. This is primary field modulated by ground. It induces voltage in RX coil due to insufficient induction balance between RX and TX coil. In a stationary sensor lacks modulation. Hence, it is possible to suppress parasitic modulation by three ways:
1. Dismodulation of TX current, which somehow make attached TX circuits. They should act so quick as to suppress even modulating frequency 12Hz.
2. Precision induction balance. From this follows that for suppression of ground signal A, the worst type of sensor is Monocoil.
3. Fixed TX loop or passive metal detector. Eric Foster has a design Superscan, to which may be connected a large stationary TX loop, inside which searches a small mobile RX coil. Passive metal detectors are described in patents of Vaino Ronka.

Signal B. This is generated by ground secondary field. However the secondary field is reflected from ground primary field. If the primary field is modulated, then the secondary field is also modulated with the same spectrum. The suppression of ground signal B can be done with differential sensors such as TWIN LOOP. These are two RX coils connected in series but in opposite direction so that subtract signals from ground and from distant interference sources.

[mikebg]

The input transformer

Lucifer,
The resistance of RX coil generates TN - thermal noise. An input transformer can improve significant signal / TN ratio if is properly designed to transform RX coil resistance. The formula is
Req=Rcoil x N^2
where Req is equivalent resistance of Rcoil transformed in secondary winding. If Rcoil=20ohm and N=15, then Req=20.15.15=4500ohm. This value is used to calculate optimal collector current of input transistor for minimum noise.
However this is useless. In pracrice the TN does not affect the depth of detecting for several reasons. The main reson is that the circuitry uses synchronous demodulator. It acts as lockin amplifier, suppressing so efficient TN, that can detect signals much smaller than TN.
However in the input of RF amplifier there are two signals, which are much stronger than TN and they limit its gain.
1. The signal called AIR is due to mutual inductance between the TX coil and RX coil. Even with very precise induction balance, the AIR signal is hundeds times larger than TN.
2. The signal called GND is due to conductivity and permeability of the ground. Sometimes the GND signal is even larger than the AIR signal.
So the gain of the RF amplifier is not limited by the level of TN, but the level of AIR and GND signals. In an ideal metal detector, the AIR and GND signals are zero, the gain is so large that the output of RF amp is near rail-to-rail noise as shown in left of attached figure. That means maximal high gain of RF amplifier without saturation in output.
A metal detector with sine induction is generally narrow band device. Required bandwidth is less than 12Hz! If done so narrow band RF amp, it will be quite weak in the input noise. For example if RX coil and RF amp generate TN=1nV/sqrtHz, then the noise voltage is 12nV. If the AIR and GND signals are zero, the output signal of RF amp without target and with deep buried target should be for example 6V noise if Vcc=5V and Vee=-5V. Then the possible gain of RF amp is
6/12e-9=5e8=500000000.
However target signals are buried in AIR and GND signals. If they are 120mV, then the possible gain of RF amp without saturation is
6/12e-3=50 times only.
In this case no sense to use input transformer for noise matching, no need to use special low noise opamp.
Now you know how must extremely sensitive metal detectors work. They need an input transformer for low noise matching of RX coil and a slow acting automatic balance system to compensate AIR and GND signals in input of RF amplifier.

[mikebg]

Idea for modification of TX

How TX of ADS works

The circuitry "dismodulates" TX coil voltage instead TX coil current. It does a good job because in tank circuit, the capacitance voltage almost completely transforms in coil current.
The opamp U1 acts as P-I (proportional-integral) controller. However for minimal settling time should be connected as P-I-D (proportional-integral-derivative) controller.
Integration timeconstant is (R1 || R6) x C2 = 278K x 10e-9 = 2,8 ms.
Voltage divider R1-R6 sets reference voltage Vref for P-I control. Due Vf of D1, the stabilized amplitude is Vst = Vref +0,6 V. It can be increased until Q1 start to enter in saturation region and tank start to return energy to supply rails. This can be observed with an oscilloscope connected across R5.
Voltage divider R3-R4 increases settling time because reduces gain of feedback loop.
Resistance of resistor R5 reduces efficiency.

TX circuit of ADSIII
Integration timeconstant is increased to 4,7 ms.
Reference voltage is fixed at 7,5 V. The stabilized amplitude is adjusted with voltage divider R2-R7.

TX circuit of MH7X
Integration timeconstant is reduced to 0,1 ms (if value of C2 is shown correct).
Resistance of R5 is reduced 5 times because it is connected partially to tank circuit. This is accomplished through the connection of capacitor C4, which increases the tank capacitance 3 times; only 1 / 3 of tank current flows trough C5.

Idea for modification
The circuit for experiment is shown right below. It is powered by unstable battery voltage because it does not affect the operation of the P-I controller. Amplitude can be stabilized to 14V if
U1 is powered also by battery voltage.

[conan]

Garrett ADS master hunter three

Hello Mike if this post is not to old I'm looking for the schematic for this model Deepseeker ADS M.H 3 it must be very similar to the normal VLF TR Deepseeker Ads but because they started to spray the I.C's with black paint to cover the numbers it's a lot harder to figure or replace them.I see you have a lot of knowledge on modifications and schematics is there one or do you know where to get one with the I.C's numbers on it.Like the one floating around for the normal one.Or are you able to figure the numbers on this other detector Goldseekers 15000 Minelab.Thanks

[amtech2005]

Hello Mike if this post is not to old I'm looking for the schematic for this model Deepseeker ADS M.H 3 it must be very similar to the normal VLF TR Deepseeker Ads but because they started to spray the I.C's with black paint to cover the numbers it's a lot harder to figure or replace them.I see you have a lot of knowledge on modifications and schematics is there one or do you know where to get one with the I.C's numbers on it.Like the one floating around for the normal one.Or are you able to figure the numbers on this other detector Goldseekers 15000 Minelab.Thanks

Hello Everyone . Goldseeker's 15000 is Minelab by the way !!
Clear Pictures of circuit boards are so important !!! Then you can still build a new circuit board and on your prototype put IC sockets with the correct number of pins on the board and then work on every stage one at a time . It can be done as long as your new IC's are the right type and supply voltages are at the proper pin locations You may NEVER find a schematic with all the IC numbers , but that should not stop you from building a prototype. 98% of everything else is on the schematic already [caps resistors etc...], why quit . By studying all schematics we have collected over the years it will not be hard to figure out the unknown IC's . Minelab made a few models very close to the 15000 . As an example look at the Relic Hawk and the Muskateer how close they are in design .

You can figure out the unknown IC's , do not let anyone tell you otherwise !!
Have a Great Day !!

Best Regards..............Eugene

[conan]

GARRETT ADS DEEPSEEKER MASTER HUNTER 3

Sorry Eugene
But the way you reply with the excitement I think I confused my post a bit.Since yes I know the 15000 is a Minelab and I should have separated from the Garrett ADS question.I was asking Mike about the schematic you have posted thanks.That is the Garrett Deepseeker ADS Master Hunter 3 .And also the numbers on the Goldseeker 15000 Minelab as I said two separate questions but made a mess of it.I did not think the schematic for the Minelab was going to come out that big I have one myself Hipmount model and I'm getting another normal 15000 from someone to look at.
And regarding the Garrett ADS 3. I have fixed several of the First ADS Deepseekers with the full numbers schematic showing all the I.C's numbers ,since they did not painted over the numbers on the physical units yet.Like in the attachment above for the Garrett ADS 1980.But since I have now the more improved model the ADS M.H 3 1984 they started to blank the I.C's numbers and I wanted to get a similar schematic with numbers even do they must use a lot of similar ones as the previous models.As you say the one you posted if you save it and transfer to the picture and fax viewer I have to zoom and zoom on it but never the less is a start.
Cheers
Miguel

[amtech2005]

Thank-You Conan for posting your questions in a good place !! We will find the missing chip numbers and always more new schematics , but it also would be nice to have clear circuit board photo's to study .

Best Regards Everyone........Eugene

[mikebg]

Professor SGS found that missing a startup resistor in the circuit diagram of low-drop voltage regulator:

http://www.geotech1.com/forums/showt...103#post168103

The circuit with missing resistor is given above in post #8:

http://www.geotech1.com/forums/showt...389#post104389

Here is the correct diagram with startup resistor R6. Thank you Mr. Professor! I hope you to find missings and errors in my other posts, and in the circuit diagram of QED:

http://www.geotech1.com/forums/showt...087#post168087

[scs]

By mikebg
I think it is correct.

[scs]

Another change...

[scs]

Garret voltage regulator-
Here is the correct diagram with startup resistor-the best variant.
.

[scs]

Correct schematic...