TP4 waveform wrong?

[exciter]

Hi,
after i figured out, that i soldered a 10ohm instead of 1ohm for R10, i have the next problem.
after i calibrate IC6 for 0V with R18, my threshold pot. isnt working correctly...
i can´t tune it for silence. my HH just beeps around ( with and without coil).
My Coil: 25cm diameter 26turns 0,5mm coppper shielded wire.
I measured 2,4Ohm for the Coil. Maybe its to low? My flyback voltage is around 300V.
I compared Carl Scope-Pictures at TP4 with my and they look different.
I mean it works, but poor and there is the problem with the threshold.
I have the Option2 non VCO HH.
I use a 10,8V LiON battery. With 13,8V Power supply its the same. Also outside.


here a pictures of TP4 and flyback:

[Reg]

Hi exciter,

Your tp 4 is correct and Carl's is not. If you change your sweep speed it will look more like Carl's except for where your signal goes positive. It is supposed to go positive like yours does.

The audio on all the time is more likely a problem with one of the resistor values being incorrect around the threshold control. You might want to check them to see if you don't have an incorrect value for R 30 or R 32. Also, check to see if the soldering is good around C 19 and C 19 is good.

Make sure the autotrack switch is closed and the voltage at pin 5 on IC 8b is 0V. with a digital meter. If it is not, then C 19 may be bad or you may have a solder problem shorting it.

Let us know what you find.

Reg

[exciter]

Hi Reg,
i checked R30 and R32. 150K and 100K, also the solder points. IC8 pin 5 is zero volt.
I recorded the sound http://www.ewetel.net/~bernd.hille/hammer/HH.mp3
I turned threshold from maximum to next. But i cant get silence...
This was without a coil connected. But with coil its the same...

[Reg]

Hi Exciter,

Did you check the pot value also? Next, you want to check the voltage at TP 9 when you adjust R 31 from maximum to minimum and note the voltage. It should go from a - V to a +V. The voltages should be small, maybe .4v and -1.2V or so at the extremes.

Let us know what you have for voltages.

If the voltages are correct, double check your transistors, Q5, Q6, and Q11 to make sure they are correct and mounted in correctly.

Finally, check the values of R 38 and R 51 to make sure they are correct.

Let us know what you find.

Reg

[exciter]

okay...
my pot is 4K7 for threshold.
TP9 goes from 0.1V - 1.70V, but i can only measure it, when PulseWidth is around middle position. When i decrease width, sound stutters like in the recorded mp3, and the voltage on TP9 is jumping.
Tested without coil.
R51 is 10k and r38 100k.
Transistor values are right.

regards

[Reg]

Hi Exciter,

You mentioned your threshold pot is a 4.7K. It is supposed to be a 100K. You might try replacing it and see what happens.

Reg

[exciter]

my fault, got a 100k. i looked at R35.
I tryed it once again in the garden. HH seems to work alot better.
i could regulate the threshold and was able to detect 10€ cent in ~17cm under potting soil. is this good?
i will try my shielded coil, maybe its better.
Hopefully it will work tomorrow, when i completely assemble it.
first i tryed with 9V battery, but there was the problem with the threshold...even outside.
with the 10,8V 4Ah LiON battery, everything seems to be good.
its hard for me to find the best setting. Some told they are able to detect a can about 80cm, im not able at all.
But maybe because of the settings...or are there some mods, which i can do? my coil resistance with 2,4ohm is okay???

Thanks for your help, will tell you how my HH performs.

[Reg]

Hi Exciter,

I am glad it is working better. Now, you don't want to use any battery with a lower voltage than what you have. You may experience problems with the other power supplies not obtaining the proper values. Also, you may be getting some interaction if the battery you are using can't handle the current draw and drops in voltage a lot. This could be part of your problem.

As for your depth on your coin, it doesn't sound to bad for a beginning. Now, you need to try to find the best possible adjustments for the delay, the second delay, the pulse length and sensitivity adjustment. All can interact at times and cause changes in the depth capability.

For starters, I would try to get the main delay down as short as possible. I would start out with a 100 usec pulse length, a 10 usec sample time and a second sample at a couple of hundred usecs later than the main sample.

For silver and copper objects, you will need a longer pulse length for the best depth, especially on larger objects.

Reg

[gef12]

Sensitivity

WHile you are here Reg
I was chatting to you a while ago about my HH and you helped me with a few issues like the ground balance etc.
Wondering if you could advise on a good combination for the feedback resistors and caps for the opamps. Currently have mine at 320k for the IC6 (wont go lower with the DD coil but with a mono will go down to 100k for some reason) and yes have played with the balance resistors too.
Also using 470k for IC8A and B. Am getting about 4 inch max with Minlab 11 inch DD coil on .9 gram gold nugget and maybe a 5 to 6 inch on 10" coiltek eleptical coil. Its close to what I expect but am greedy I WANT MORE
Do you think this is what I should expect. Delay is better with the DD than the Mono , currently about 8uS with DD and 10 with Mono. Anything else I can tweek or play with. Am using li-ion 15volt battery going thru 12v Regulator and have separate Audio amp with 9 volt supply. This is also non-vco option. Thanks in advance.
CHeers
gef

[gef12]

PS

Also have reduced my sensitivity to ironstone hotrocks to almost nil, which I feel is a good thing.
I have a 22ohm in series with the mosfet to the coil to reduce current.
Am using low coss mosfet at 600v and 2A. I have made slight mods to the coils my cutting at least a foot off the cable to reduce the cable capacitance.
My targets are gold nuggets. Small ones for the region I go to.

[Reg]

Hi Gef,

One thing you didn't mention is if you built a form of ground balance for your PI. Ground balance is great but it can reduce the sensitivity.

As for your sensitivity, you are close to the limit on the small stuff. However, there are a few tricks that help a little. Again, the strides will not be great, but you can gain a little more.

So, to try to answer your question, you should be able to reduce the gain even more on IC 6 and use a DD coil. I am not sure just why you can't. It should be possible since I am doing it on my HH.

The one thing that will allow you to reduce the delay (which can result in detecting both smaller gold and a little deeper) is to reduce the gain on the preamp some more. Also, the gain is not linear over the range of nuggets you should be able to find. If you search this forum but search for the following; "Low Noise Front for P.I." by Ferric Toes you will see how the gain tapers off on targets with a very short delay. This results in two big problems. First, the gain on small gold is reduced dramatically and the second is the earth field effect can cause problems. Both are not good. This shows up on his pic of the output of the preamp where the "fat" signal becomes thinner at shorter delays.

Here is a link to the discussion. http://thunting.com/geotech/forums/s...ighlight=front

I hope it works. The reason I linked to it is because it discusses the importance of reducing the gain. What isn't discussed is the fact some brands of IC's are better than others. This is another thing that Eric found out. So, reducing the gain can be more important with some IC's.

Now, I used Eric's test technique and found I could reduce the gain to about 100 and my opamp worked fine with any coil. Reducing the gain does allow for faster sampling, but more importantly, increases the gain for detecting smaller nuggets, once the gain is made up in a later stage.

You might also consider building Carl's double amp preamp circuit as an add-on and not use a single amp for the preamp. This would allow for additional gain without the problems involved with a single amp. It isn't necessary, but should help overall.

One other thing you might want to try is something that I came up with that was posted on another forum and that is to add a diode and resistor combination across the feedback resistor of the preamp. The basic design was posted on the PI forum by another guy to whom I had sent the idea, so I will show it here. Now, if the gain is changed from the original, then the 22K resistor will have to be changed. You could experiment by using a 25K pot, and may find you will have to go down to a low value of 2K or maybe less at reduced gain.

What this little circuit does is reduce the gain and speed up the preamp which also allows for a little faster sampling, but more importantly, it reduces the ground signal. In some cases, the reduction is enough that no ground balance is needed, which effectively increases the depth capability.

Finally, the IC noise is still a problem when searching for small gold because it interferes with hearing the very weak signals. My little diode circuit allows one to use some other low noise IC's that won't work without it. Right now, I have another IC coming my way I want to try. I have been able to use the AD797 which won't work without the diode circuit, but have had to reduce the resistor value even more. Unfortunately, the AD797 is quite expensive and does require a little more reduction than I would like to use. So, I am always looking for the ideal replacement for the NE5534A.

If you are hunting in the field, shielding of the electronics isn't quite as critical, but is still a good idea. I use a plastic housing but add shielding inside the case to reduce signal pickup. It is amazing just how much it helps and where it helps. If you have the resistor in series with the FET, you might not need any form of heatsink on the FET. That becomes a great noise antenna and makes noise a bigger problem in some areas by re-radiating the noise inside the housing. I am going to use a heatsink but do something different and that is to isolate the heatsink from the FET and then ground the heatsink. This way, the FET is "shielded" from radiating noise. Using one of the newer plastic encased 220F types helps because the FET case is already isolated by the plastic.

So, as you can tell, there are always things to try that can help. I hope this helps you a little more.

Cheers,

Reg

[Reg]

Hi Gef,

For some reason the pic of the diodes and resistor circuit across the preamp didn't attach. So, I am trying again. The addition of the two 4148 diodes and the 22K work well on the high gain of the original HH. If the gain is reduced significantly, then the 22K will have to be reduced accordingly.

Reg

[exciter]

Hi all,
will try these settings.
I tried also an LM318 instead of NE5534 but i couldnt get Tp4 to zero, so i think R17 and R18 must be changed...right?
I shielded my coil with aluminium-tape. Is it good or not so good? Should i go for another shield material?

[gef12]

Hi Reg
Yes thanks for all your valued comments.
Interestingly I replaced my 5534 with another and suprisingly my sensitivity shot up a little, not much but noticable. Must be different batch that I had purchased... would be a good thing to try. Also I did try an AD797 previously but it didnt work ...it just would not see anything.... but will try that suggestion of yours in the near future.
I have my unit crammed into a small alluminium box which should give it ample shielding, and had to get that box from the U.S. Just doesnt seem to be quite what you want these days on the shelves so we have to improvise.
Next step will be to construct the rod .. have to have it all ready by June when I go out bush.
thanks for all reg
cheers
gef

[Reg]

Hi Gef,

Check and see if your two NE5534's that have different gains are not made by different companies. I suspect they are. If they are, which company seems to have the fastest one?

It appears the NE5534's are being phased out. NJM and Philips are no longer making them. This only leaves TI and ON semi that I can find.

Making the AD797 work well enough that one can sample quickly takes a lot more work. It is tough and that is why I have decided to try to find a different IC that will be almost as quiet. I will let all know what seems to work.

Reg