TP4 very noisy

[nick_f]

Hi All,

While I consider that now most functional blocks in the HH work as expected, I have a lot of noise at TP4 and I don't know how to get rid of it. I am using NE5534 on socket (IC is good, I have more chips and I replaced it). I cleaned the flux on the board.
The level of noise on IC6 is about 0.6Vpp. Considering the gain of IC6 is about 1000, it means the noise level at the input is 0.6mVpp (assuming it is indeed noise and not oscillation of IC6).
Did anybody else encounter these problems, and if yes, how did they solve them?
I place in here some shots of the CRO, with explanatory notes attached.
I will remove the 7660 and replace it with a 9V battery, to check if it introduces noise.

Regards,
Nicolae

[hobbes_lives]

Perhaps you can try disconnecting the coil and cable from the PCB to see if it's just external noise that you're picking up?

[nick_f]

Perhaps you can try disconnecting the coil and cable from the PCB to see if it's just external noise that you're picking up?

Hi Hobbes,
That's exactly what I was testing now. (I also removed the IC6 socket and it didn't improve). Without the coil, the noise is much lower, but it tends to have the same frequency.
What I would like to know is if other people who built this circuit had such problems or not. Because if it is a problem related to the PCB or the circuit design, then I will approach a different strategy (replace the preamplifier IC6with two opams with lower gains, as in QED circuit, and shield it in a separate enclosure, away from the proximity of high voltage peaks at TP3). I notice if I put the finger on TP3 with the coil installed, the noise at TP4 can even double in amplitude, so it must be related to the voltage spikes near TP3.

Regards,
Nicolae

[nick_f]

Hi All,
It looks like I found a convenient way to reduce the noise, increase the sensitivity to varios targets, including the 1 square inch aluminium foil, maintaining the same setup (the same bad coil I was using).
The solution was pretty simple: I added a 100k resistor on top of R13 and now the circuit is much more stable (I don't get all kind of random screaming noises when no target is nearby, as before). Maybe Carl should have added a pot to adjust the gain of IC6 as well.
Interestingly, I can adjust the circuit for even greater sensitivity (beer cap detected at 23cm and 1" square alu foil at around 2cm).
You can notice that also the pulse at TP4 is about 8us width, compared to about 12us before.
I don't know why the pulse now has that distortion on it, I will investigate (it is not due to pulse at TP6). I found out with big surprise that lowering the value of the coil resistor R11 increases dramatically the width of the pulse (I would have expected rather the opposite, or no change), so there are multiple advantages in using a resistor as high as possible.

Regards,
Nicolae

[hobbes_lives]

Nicolae.

You've adjusted the voltage gain of the preamp from 1000 down to 100, so you should definitely see a reduction in the noise. If it was 0.6Vpp before, take a look and see if it is now 0.06Vpp now. The noise may still be there, just greatly reduced. The target signal will be reduced by a similar amount. It sounds like it may be a balancing act to get a good signal gain with noise reduced enough to not affect the sensitivity overall.

If it seems like it's coming from the opamp, you may want to try a different manufacturer. They can vary quite a bit, even from lot to lot from the same manufacturer. What is the exact manufacturer and part number that you are using?

[nick_f]

Hi Hobbes,

In the end I set a resistance of 330k (gain = 330). I agree with you, most likely the noise is still there, but doesn't cause problems anymore. I made the first version of PCB (ver C) and I don't know if other people managed to build this version successfully and to get good performance. It may have to do with the layout of the PCB as well.
The IC I am using is made by Texas Instruments, and it is a NE5534P. At the top there is some other code, 81AX27M, not sure if it matters.
The sensitivity of the MD hasn't decreased by reducing the gain, from what I can see it didn't make any difference in performance, I just don't get the random audio chattering as before.
I also tried in the past a LM318 and it was different output of TP4 (can't remember exactly what was it like). I don't know if I can just replace it without any other component changes. Maybe I should give it another try.

Regards,
Nicolae

[hobbes_lives]

I believe that's a standard NE5534 in a plastic package. You may want to try a NE5534A from the same manufacturer. It has lower input noise and current ratings. Take a look here...

http://focus.ti.com/docs/prod/folders/print/ne5534.html

[nick_f]

Yes, it's a plastic package. Carl didn't mention anything about using NE5534A. I know they have better noise specs, but the noise I am getting seems to come from outside the chip. I don't think that the problem, but thanks for the suggestion. Meantime I will try again the LM318 and see what's the deal with it.

Regards,
Nicolae

[Carl-NC]

The ripple you see is interferer-type noise, not thermal. If you remove the coil and it's still there, then it is caused by the circuitry itself. The charge pump is a possible source.

- Carl

[nick_f]

Hi Carl,

Thanks for explanation. I already removed the 7660 and replaced it with a 9V battery. Results didn't change, so it's not 7660. I also believe the noise is due to an interferer, but what to look for?
When I removed the coil, the noise was much smaller. I suspect is the flyback pulse that generates this trouble.
I am using ver C of the PCB (single sided), does that one work properly? I also have the capacitor C1 and resistor R10 in circuit.

Regards,
Nicolae

[nick_f]

TP4 very noisy - not anymore

Hi All,
I have been able to identify the cause of noise at TP4 and get rid of it (for testing purposes). It was an external interferer, which I can't locate. Maybe it could even be in one of the units surrounding me. I got aware when I started rotating the mono coil and I could see the noise changing amplitude. Just to go to the extra mile, I build a figure-8 coil (following advice from Carl) and now the noise is much smaller. As long as I know that the noise is not generated by the circuit, I am happy, as Carl said, outside should be much more quiet. The preamplifier gain was set to 1000, as in the original design.
Thanks to everybody who helped me with this matter. Also, greetings to Woody, for posting the explanatory video about Fig 8 coils on youtube!
With this coil (368nH), despite Al shielding and large shield-coil capacitance (272pF), HH can detect the 1 square inch Al foil at 8cm and the beer cap at 21cm.

Regards,
Nicolae

[Unregistered]

still

Hi All,
I have been able to identify the cause of noise at TP4 and get rid of it (for testing purposes). It was an external interferer, which I can't locate.
Regards,
Nicolae

U have taken lots of bandwidth on this site to identify this (but in vain), waht happened to max

[nick_f]

I posted all that information and photos with the hope some people might find it useful. Unfortunately you are not one of them.
If Carl considers it's all junk, he can delete it all if he wants so, I don't mind.
Eventually I found the reason TP4 was noisy, so to me it wasn't in vain. People on the forum helped me a lot to understand a bit more about metal detectors. I have no idea about Max.

Nicolae

[Sean_Goddard]

Good effort!

[N1PS]

I have been chasing similar gremlins with the Surf PI. I noted the noise I am encountering is low freq in nature. It appears to be below 100Hz. Very sporadic and not a sine wave (like hum). I did get some improvement by carefully looking at the DC rails and added more bypass caps to remove the noise on those rails. Added bypassing around the DC amp (LM35 and had some good results. I also noted my audio likes to loop back into the board.

The noise itself is a mystery. It may be flicker noise or even brown noise. Seems many of these PIs including your HH suffer from noise. I am of course playing with this MD in my house and not out in a quiet field. But a mystery on the noise.

My .02.

Peter