LOOKING FOR GOLDSCAN 4 Schematic.

[lucafa]

to reg
you really need to change chip to offset?
I think RV1 hides other function.
I tell you how I think.
U12B is an analog memory, the voltage on TP6 is loaded in C29.
On pin 7 U12B I have a amplified voltage of about 6 times and this goes to R40 and R52 + SW1-A.
U12C is a comparator voltage between R40 and RV4 voltage, the comparator trips when it goes to -4.31 V and forbids FET Q10 and voltage at C29 can indeed be amplified (object detected).
In the event that no object is internal loop and in the presence of noise, U12C corrects, by sending into saturation Q10 (approximately +4.12) for a short time, and prevents noise to hear the speaker.
How do I connect this with my claim?
I checked the voltage on pin 7 of U12B and it is always positive, this makes sure that pin 8 of U12C is always at -4.31 V and then interdiction for Q10 and it does not work automatically.
What to do?
RV4 rule only negative voltages.
We must adjust RV1 so that we shift the voltage at pin 7 of U12B towards negative values so as to be able to operate properly U12C.
Everything you said is true Goldscan original.
Google translator.

[Reg]

lucafa,

The purpose of RV1, RV3, RV5, RV6, and RV7 are to compensate for offsets caused by the opamp used. The actual adjustment procedure for these pots is to hold down the reset switch while adjusting. Why hold down the reset switch? Simple, the reset switch activates the FET's, which short out any possible signal coming from the previous circuitry including the coil, thus leaving only the offset as a possible signal. So, any output will be the because of the offset of the opamp.

With the Reset switch pushed, the above mentioned pots are adjusted for 0V at the output of the associated opamp. Since the circuitry is dependent upon the previous stage, the adjustment procedure should start with RV1

How do I know my adjustment procedure is accurate? Simple, I was given it by Eric Foster many years ago so I could adjust a GS 4 I had that was out of adjustment.

Now, if you use a precision low offset opamp, then there will be little to no offset to compensate for, thus no need for the pots mentioned. Does my procedure work? Yes, I removed the above pots once I changed opamps and the detector works just fine. In fact it is better because the adjustment of the pots is extremely difficult to do correctly.

A person doesn't have to change the opamp, but it does make adjustments and design easier.

You are correct, C29 does hold the signal coming from the coil in memory on the original design. Unfortunately, the detector drifts terribly or at least, mine did with no resistor across the FET Q10, thus requiring that I constantly push the reset button. Adding the resistor across the FET causes the voltage on C29 to ramp down to 0v at a determined rate. This addition of the resistor creates a form of autotune that tries to keep the voltage at 0V by ramping down any voltage applied to C29.

On the original mod, the resistor value added across the FET is a 470K. This resistor causes the Capacitor C29 to discharge to 0V quite rapidly. Changing this resistor to a 4.7meg, slows that rate down considerably, so it is much closer to the original concept.

Reg

[lucafa]

reg

Nothing to say about the offset calibration and even the use of operational low ofset.
The point is: RV1 must be set to zero pin 7 U12B or to set value voltage suitable to work correctly even U12C?
To confirm what I say and if you have a goldscan unchanged at hand, you can verify that the output of U12C is always at -4.31 V, with or without metal.
A similar circuit is used in to SuperScan (also in my hand) and it works as I say even with its limitations.
This fact makes the MD unstable because any noise, even in small quantities, is played while being itself voltage variation.
This evening I will try to adjust RV1 to run U12C and check on my theory.
I expect is a loss of depth but an improvement against noise.

The resistance of 470K Q10 is interesting, U12B from analog memory becomes non-inverting amplifier, makes the MD definitely more stable (tested), but changes the revelation NOT MOTION to MOTION, this sincerely I do not like.

With my post I want to ask if you have any readings for TP6 and TP7 with different metals.

A thought.
Eric Foster has designed, built and sold these units for gold (not only) on very mineralized ground.
If you have sold MD, the project worked otherwise I do not think that people spending money unnecessarily.
This leads me to think that there is a particular alignment to make maximum use of this metal detector.
At least I hope .

lucafa

[Reg]

lucafa,

If you use the TL064 you will have to adjust RV1 just like you said to obtain 0v at pin 7. However, if you use a precision amplifier with an extremely low offset, then the voltage at will be 0V already at pin 7, so no further adjustment would be necessary. Keep in mind the offset voltage of the opamp can be as high as 20mv depending upon the opamp used. Now, one of the opamps I use has an offset voltage of a typical 30 uv up to 200 uv. That is 100 times smaller.

The offset voltage is the internal error voltage that adds to any incoming signal. If not compensated then the offset voltage can be amplified and become a much higher value, thus making it impossible to properly set the threshold.

The voltage at each of the opamps having an offset adjustment should be 0V at the output and that is why the offset adjustments are there. Because the TL064 or similar opamp is used, the opamp will have an internal offset voltage that will cause the voltage at the output to be more than 0V. Thus, an external adjustment is necessary.

Now, if you use a precision opamp, there will not be any signficant offset voltage so the output voltage will already be 0V.

Again, to properly adjust RV1 or any of the other mentioned pots, the reset should be pushed to eliminate any incoming signals altering the measurements.

Now, for those reading this post, of the outputs are not 0V, then it can become difficult to impossible to properly adjust the threshold. Also, the threshold level will vary dramatically if the gain (also called Sensitivity) is adjusted. To eliminate these potential problems, then all opamp output voltages should be adjusted to 0V.

BTW, this 0V is really 0.000V +- 1 mv or so, so a good digital or a very sensitive scope should be used for calibration.

As for your statement U12C is always at a negative saturation. That is true because that opamp is being used as a comparator that is either full on or full off. The pot RV4 sets a voltage of about 4 to 15mv offset to keep the U12C at a negative level. This is why the 0V out of opamp U12bis critical and should be at 0.000V. This comparator is what triggers the FET's to eliminate any voltage of the wrong polarity getting through. In this case, if the output of U12b becomes a greater positive voltage than the threshold set by RV4 (plus any offset error), then the comparator will switch, thus turning on the FET's. Turning on the FET's will short out any signal at the opamp side of the cap connected to the FET. In theory, the left side of this cap should be at 0V or very close to it because of the adjustment at the preamp when there is no target signal applied.

BTW, on a different note, the sensitivity can be increased by increasing the value of the 10K sensitivity pot. You might try a 25k or even up to a 50K but keep in mind that any value greater than the original 10K pot further amplifies any error voltage.

Reg

[Reg]

Lucafa,

I am not advocating people buy the precision opamp. What I am saying is that using such an amplifier can make adjustment of this detector much easier by eliminating several of the adjustments that are confusing at best.

Most people don't know the reason behind the adjustments of most of the pots on this detector or the proper way to adjust them. If not properly adjusted, you can lose depth, have difficulty with the threshold and other issues. By eliminating some of these adjustments, the detector becomes much easier to adjust, thus making it easier to make work correctly.

Also, there is little to no need for the Selector Switch used if the design is simplified. The SEL 1 and 2 are there to make it easier to detect certain English coins and not there for gold hunting. So, these two positions can be eliminated for simplicity since trying to find the proper selector switch can be difficult. BTW, SEL1 is basically half of the high conductor/low conductor design, since it only allows low conductor signals to be amplified and shuts off high conductor signals.

Now, a basic design that doesn't have the multi position selector switch or most of the different pots to adjust would certainly be simpler to build and align.

The adjustments of Rv1 and the others I mention do not have anything to do with eliminating noise. The noise generated is from external noise and any internal noise created by the preamp and associated circuitry. If you want to eliminate some noise, you can build the high conductor/low conductor switch into the design because this will reduce the noise by eliminating somewhere around half of it by not amplifying half of the signal.

Did you know that the simple NE5534 used as the preamp can cause problems? Yes, there are differences in the amplifiers depending upon the manufacturers and some work better than others. Some are simply not as fast as others among other things.

To find out, you have to test the opamps using a technique mentioned by Eric Foster quite some time back on this forum. Right now I can't tell you where that information is but it should be there.

Reg

[mschmahl]

Thank´s reg for your hint´s. Now i have to order some parts to make your mod´s. Do you have some information (how much uH and ohm for tx and RX) for the dd-coil?

[Orbit]

http://www.youtube.com/watch?v=cVJK8HczDdQ this is half gs 4 modified tone level ,the other half taken from another pi detector ,,and that the real it works perfectly !

[Orbit]

http://www.youtube.com/watch?v=avlY7...CfpDWw&index=1 .....http://www.youtube.com/watch?v=avlY7...CfpDWw&index=2 GS 4 with significant changes

[Jose]

Hi Everyone
Very interesting the topics covered in the previous post.
Built the GS4 and I would like to make some comments
The biggest problems I had to solve were related to noise.
As to the added resistance in Q9 and Q10, I use 1M, with lower values as k 470 is more stable with less gain, does not work well without the resistances mentioned.
As mentioned Lucafa, RV1 adjustment is complicated, was adjusted to 5 millivolts negative RV1, measured in leg 7 of U12B, so it worked perfectly U12C.
Also in the same way adjustment (5 millivolts negative) RV3, measured in the leg 14 of U12D, in order to have equal voltages at the ends of potentiometer GND.
Thus threshold audio remains stable, when you move the potentiometer GND.
As Geo mentioned, nails, very rusty wires are detected as high driver.
Coins larger, can be rejected as iron, depending on the setting GND.
Jose

[ABDW9]

Dear brothers to you my sincere thanks
And I wish to inform you I am new to this forum I wish I were old with you until I see the maximum benefit you
And yet the first thing I have done is to design the painting because've preached to buy all the ingredients and now waiting to leave I collect
It then compared Mathsalt him including Thsaltm mechanism you
Greetings to all of you

[Reg]

mschmahl,

You can use a wide range of values for the TX and the RX coils of a DD type coil. I would recommend you start out with about 300 uh for the TX and maybe 450 uh for the receive to begin with.

Now, since you own a TDI and a GS 4, you might replace the connector on the GS 4 to a 5 pin connector identical to the one on the TDI. This way you can use all coils on either detector. Pins 4 and 5 are for the transmit, with pin 5 being the ground or shield connection. The receive pair are pins 1 and 2 with pin 1 for the ground (shield) connection and pin 2 for the signal.

Leave pin 3 not connected on any coil you build. I may have a different use for it in the near future.

Now, I don't know if you order from Ebay but if you do, you might wait until one of the opamps I mentioned go on sale. Like right now, the LT1014 are reasonable from a vendor but other IC's do show up on sale regularly.

Reg

[Geo]

Lucafa,


Did you know that the simple NE5534 used as the preamp can cause problems? Yes, there are differences in the amplifiers depending upon the manufacturers and some work better than others. Some are simply not as fast as others among other things.

To find out, you have to test the opamps using a technique mentioned by Eric Foster quite some time back on this forum. Right now I can't tell you where that information is but it should be there.

Reg

Here eric speak about op.amp for front end...

http://www.geotech1.com/forums/showt...P-I&highlight=

[lucafa]

Hi all.

I'm happy that these posts are appreciated.

Before continuing the discussion I mean to follow the instructions of reg set to zero volts in the calibration points.
This is to avoid confusion.
I summarize here the calibration sequence:
I remind you that switch SW1 has the following function 1=OFF (first position is at the top, look at the arrow), 2=ALL, 3=SEL1, 4=SEL2.

RV1= switch SW1 to ALL, pot GND minimum side R52 , press and hold RESET button, adjust RV1 for zero volts on pin 7 U12B.

RV3= switch SW1 to ALL, pot GND minimum side R52 , press and hold RESET button, adjust RV3 for zero volts on pin 14 U12D.

RV5= switch SW1 to ALL, pot GND minimum side R52 , press and hold RESET button, adjust RV5 for zero volts on pin 7 U13B.

RV7= switch SW1 to ALL, pot GND minimum side R52 , press RESET button and hold, pot SENS maximum, adjust RV7 to have zero volts on pin 14 U13D. Check if it remains zero volts on U13D varying the pot SENS.

RV6= switch SW1 to OFF, pot GND minimum side R52 , press RESET button and hold, pot SENS maximum, adjust RV6 to have zero volts on pin 14 U13D. Check if it remains zero volts on U13D varying the pot SENS

RV2= use oscilloscope channel 1 probe to the gate Q4, trigger channel 1, channel 2 probe to the gate Q8, pot DELAY (or FOIL REJECT) at minimum, adjust RV2 for a delay of approximately 17uS (values of my MD), vary pot DELAY to the maximum and read delay about 95uS.

RV8= use oscilloscope channel 1, show at least two pulses, put gnd trace oscilloscope and place on a line screen (zero), back in DC, adjust RV8 to keep track of a two pulse and the other on the line of previous gnd.

RV4= from various posts and forum how to do is this: to adjust a value that ranges from -4 to -15 mV, find the point at which snaps pin 8 of U12C from -4.3 V to +4.3 V with micrometric movements and MD for an answer without disturbance without an object, the audio signal with the subject at a good distance.
(And this thing that you are 'talking in these posts, at least I do )

[Reg]

Geo,

I use different IC's in place of the NE5534 than those mentioned in your link. I have used the AD797 successfully but you have to place diodes across the feedback resistor to limit the IC from going into saturation. It will work fine that way. However, there are a few ICs' that are almost as good or even better.

The LME49990 has lower noise specs and works without any problems. The OPA211 works well and is almost as low noise specs plus the current draw is much lower. The OPA 228 works fine also but is not quite as fast as the LME and the 211. All three are precision opamps having very low offset. This eliminates the need for any offset compensation.

The LME49990 and the OPA211 do not come in the dip package so an adapter is required. The OPA228 does come in the dip package so it can be used directly. In fact, I recently modified a Goldquest by replacing the 5534 with the OPA228. The detector was having drift problems and this eliminated that issue.

Now, as mentioned, one advantage of these three IC's is the fact no offset adjustment is required. So, the preamp circuitry becomes simpler because no compensation is required either This does away with several parts and basically leaves the opamp, feedback and input resistor as the key parts left except for the PS leads and they should be changed also. THE 47 ohm resistors in series with the power supply leads can be removed. Just use a tantalum cap on each lead and parallel the tantalum with a .1uf to take better care of the higher frequencies.

BTW, if you read Eric Foster's posts you will see where he dropped the input resistor to the preamp from 2.2k to 1K to reduce noise on his later designs.

One thing I didn't mention is the GS 4 has a limit on the current it can produce for the 5V supply. So, one has to be careful when selecting parts to substitute. I found this out the hard way, when I suggested the OPA4277 as a replacement for the quad opamps used in the GS 4. Well, a guy ordered the OPA4227, which is also a quad opamp. The difference is the 4227 draws about 4 times as much as the opa4277. So, the 4227 draws about 16ma each chip and with two of them, the 32 ma was too much for the designed power supply when combined with all the other circuitry drawing current from the 5V supply. Thus the 5V supply was loading down. That took a while to figure out.

One final issue you might check and that is the GS 4 I have would have a hum in headphones because of the pulse rate of the detector. To eliminate this problem I raised the pulse rate to over twice the original. This eliminated the buzz and still worked well. If you decide to do this, remember that simply raising the pulse rate by changing the timing cap on the 555 will also shorten the pulse length.

Reg

[Geo]

Ohhh Reg!!!!
So much effort for nothing.
I have not a GS 4. I attached the address because you did not known where was the thread of Eric.
But if you have any Tips for TDI are welcome

My wishes