The TP1 Waveform

[1843]

Hi,

I changed all parts around IC8, But it had no effect...

I noticed that changing the frequency can affect on the beeps;
Even I can set it at a position that TP9 voltage changing will be high...(These occur whether the MOSFET is installed or not)
Also while a target become very close to the coil (under 10cm), the sound damps.

What do you say?

[Reg]

Ok 1843,

I have gone back and read each of your posts but don't understand what you are trying to say in a couple of them.

Now, you see a voltage varying at TP 9 and if it reaches 3V you get a beep. Ok, the beep is happening probably because it reaches a voltage great enough to overcome your threshold setting since you have built option 2.

The variation is the problem and it occurs before TP 9.

If you are still getting the variation with the coil disconnected, then the problem still could be in that area.

If you remove IC 6 and the noise stops, then you need to look around IC 6 or before.

Make sure all resistor values are correct around IC 6.

Also, make sure your damping resistor R 11 is less than 1K.

If the noise is still there, then take a small jumper and jumper across D3 and D4. You can do this by jumpering TP 3 to ground with a short wire. Make sure the wire length is short. If the noise stops or reduces greatly, the noise is probably coming from the FET or associated area. Once you do this, check the voltage at TP 4 to make sure it is less than 3V. If it isn't, adjust R 18 until it is. Now, watch the TP 9.

If the noise is still there with the jumper from TP 3 to ground, then the noise is coming from IC 6 or associated circuitry. Try changing IC 6 and associated resistors. Make sure they are low noise types.

Also, did you use low noise resistors around IC 6. A cheap resistor could cause some of the problems, but normally a bouncy signal is caused by an IC.

If jumpering TP 3 to ground eliminates the noise problem, then check everything around the FET. R 11 and R12 should be low noise resistors.

Remove the jumper.

If you are using a heatsink on the FET, remove it and see if the noise reduces. It makes a good antenna and can pick up noise and reradiate it into the circuitry. In other words, the FET and heatsink can be a major problem and still be good.

Now to help minimize this posibility, you can buy an insulating kit for the heatsink and insulate it from the FET. Then ground the heatsink so it now becomes a noise shield. If you do this and plan on using a long pulse, then make sure you use a heat sink compound to help transfer the heat.

You can also buy FETs that have insulation on them so they can be connected to a grounded heatsink. Normally, they have reduced current ratings because of the insulation. If you use this type of FET, then use a heatsink compound between the FET and the Heatsink and ground the heatsink.

DO NOT GROUND THE HEATSINK WITHOUT INSULATING THE FET if you use a regular 740.

Try these things and let us know what you find.

Reg

[1843]

I shorted TP3 to ground and cheked damping resistor and R12.

Now the threshold works right like an IB metal detector.

The TP4 voltage is about -3.2V and the TP9 varying reduced very much and TP9 is close to zero, But I can still set the frequency until TP9 is varying (the varying skirt reduced greatly) and sound has a little modulation.

The noise does not depend on MOSFET, because these tests got done without the MOSFET.

[Reg]

The 3.2V at TP 4 is to much and the IC could be in saturation which will reduce the signal at TP 9. That is why I said you may need to adjust R 18 to make sure that voltage at TP 4 is less than 3V. This assures IC6 is not in saturation.

So, to perform the proper test with the jumper across the diodes, adjust R 18 until TP 4 is less than 3V and then see if the noise is reduced. If the noise is still there, then the noise is probably coming from IC 6 or associated circuitry. A voltage of 3.2V at TP 4 is very close to the saturation voltage of IC 6. If saturated, then it will reduce the noise if coming from that area because the IC can't amplify any more and hide the problem.

If the voltage at tP 4 is less than 3 V and the noise is still there, I would try changing IC 6. Check to see if all resistors are the proper values around IC 6. Then I would change C 14 and C 15 to tantalum type caps. Best to buy them surplus since they are quite expensive otherwise.

Tantalum caps will reduce high frequency noise about "twenty five times" better than electrolytics, according to National Semiconductor. Personally, I have found them to be the best to use.

I would reduce R 21 and R 20 from 27 ohms down to 5 ohm or maybe as high as 10 ohms.

Make sure the 1 meg feedback resistor (R13) isn't greater like 10 meg or so. In fact, I recommend this resistor be reduced in value to something between 220K and 470K. One could go as low is 100K and make up the difference in gain in the next stage.

You could change R 13 to a 100K and then change R 26 and R 27 to 1 meg resistors. You will probably have to reduce the value of the two caps, C 17 and C18 then.

Make sure R 16 is a low noise type resistor. In fact, all resistors around IC 6 should be metal film type to reduce noise.

If you are wondering why one should reduce values of resistors or use special types, you might want to go back and read some of the previous posts about preamp noise. Eric Foster pointed out a few things and a lot of people provided a lot of good information in a previous thread.

Now, you should realize it doesn't take much of a signal to cause a couple of volts deflection at TP 9. The gain of the preamp is 1000. The gain of the next stage is 100 if the sampling time is long enough. The next stage has a gain of a 100 also. So, after the second stage, you have amplified things 100,000 times and after the third stage signals are amplified a total of 10 million times (if my math is correct). Now, in reality, that gain is probably much less because of the short sampling time, but even with taken into account, one can see at a microvolt noise signal becomes a signal that is in the volts range.

So, any external noise such as testing the detector near something that generates noise will cause voltage deflections at TP 9. That is why it is a good idea to keep anything that causes noise away from the test area. This includes things such as computers, TV's, fluorescent lights, radios, or any other noise generating piece of equipment.

It is possible you have some company nearby creating sufficient noise to cause problems you are experiencing and there really isn't anything wrong with your detector.

This is also why some people place their circuitry in a metal box so shield the circuitry from the noise. Other people like myself will add a shielding material to the inside of my plastic case, especially around the FET and preamp area to minimize such problems.

Anyway, it is something to think about.

PI detectors are not like IB types. Because the amplifier stages are broad band, a PI will amplify noise much more than other types of detectors and as such, will have an unstable threshold in noisy environments where a IB will be stable and quiet.

Reg

[1843]

Thank you dear Reg,
But the TP4 voltage (after the TP3 shorting) is about -3.2V (minus not plus)

[1843]

When TP3 is shorted to GND, R18 cannot affect on TP4...
I replaced the R13 with a 100k, but the TP4 voltage still is -3.2V.(Only effect of it was reducing the detection depth.)
If I remove the TP3 jumper,again HH beeps...
I am confused...

[Carl-NC]

OK, let's back up a little bit... remove Q3, but leave the coil hooked up. Measure the DC voltages at TP2, TP3, and TP4 and IC6 pins 4 & 7.

- Carl

[Reg]

You are right, when TP 3 is shorted, you can't adjust the voltage. Sorry about that, my mistake. I worked 12 hours last night and was still up, so I was not thinking clearly.

Remove the jumper at TP 3.

Is the coil attached? If it is, remove it and see if the pulsing stops. If it doesn't, then proceed to the next step.

You mentioned you have the FET removed. If that is the case, then jumper TP 2 to ground with a short jumper and then adjust for 0V at TP 4. See if the noise is gone then. Do not jumper TP 2 to ground if the FET is mounted on the pc board and working.

Also, check the voltage at pins 4 and 7 of IC6 to make sure the +5 and -5 supplies are correct.

You are correct, changing R 13 to a 100K will reduce the sensitivity but should reduce the noise considerably if it is coming from IC 6.

Now, changing R 13 to a lower resistance value will make the IC 6 work better at shorter delays. If you increase R 26 and R 27 to 1 meg, you get your gain back or close to it.

Reg

[1843]

TP2 & TP3 are zero
TP4 is zero if jumper be open
IC6 pin4 = -5V & pin7 = 5V
(R13 = 100k)

[Reg]

Sorry Carl,

I didn't see your post. If I had seen it I would have waited to post mine.

Reg

[1843]

The day after tomorrow I am trying with a shielded box for my HH...
I think that Reg is right,
The Op-Amp is very sensitive... and must be shielded from any noise...

Best regards

[Carl-NC]

" TP4 is zero if jumper be open"

What jumper?

OK... continuing with Q3 removed... adjust R18 to one extreme, measure the DC voltage on TP4... the adjust R18 to the other extreme, measure the DC voltage on TP4. What are these voltages?

- Carl

[1843]

R18 changes the TP4 voltage +0.15V & -0.15V if R13 be 100K.

[1843]

Hi,
I found that when HH beeps the pin4 voltage of IC6 is unsteady...
But I can see its instability only when the VOLTS/DIV is set to 5mV or less.

What do you say?

[Carl-NC]

R18 changes the TP4 voltage +0.15V & -0.15V if R13 be 100K.

This sounds good. Put it back to zero.

I found that when HH beeps the pin4 voltage of IC6 is unsteady... But I can see its instability only when the VOLTS/DIV is set to 5mV or less.

Probably whatever is causing the beeping is pulling hard on the -5v rail and glitching it.

When the beeps occur, is the output of IC6 glitching?

How is your HH built? Breadboard? PCB?

- Carl