Damper Resistor Reference!

[1843]

Hi,
Here is the AD844 configuration in my PI.



Regards

[Carl-NC]

The AD844* has some nice specs, but my experience with opamps is, you just can't tell until you try them out. The preamp gets badly abused, and it's the large-signal behavior that matters. That is, full-power BW, slew rate, and overvoltage recovery. The last is rarely spec'd.

On the funky little hump, I've sometimes seen that, and never quite figured out what it is. I suspect opamp output stage nonlinear recovery.

What are your diodes (eg D5)? Bidirectional schottkys?

- Carl

*I know the designer of the AD844, smart fellow... that part is pushing 20 years old, and it's still a competitor.

[1843]

D2 and D3 are 1.5KE400CA. They limit (800V) the flyback peak to protect the pulse switch (IGBT 1200v).
D5 is SA5.0CA.

Thanks for the reply

[1843]

Carl,
Do you think that the coil is underdamped?
What modification do you suggest for the preamp's feedback?
I am not an electronic expert; so I don't know what to do. I'm waiting for your reply.

Best regards

[Carl-NC]

The coil is not underdamped, if it were you would see an overshoot (start of ringing). Try Reg's suggestion, reduce the feedback R to 470k or even 330k. Most high-speed opamps are datasheet-tested with a much lower feedback R, so performance can be quite different for a high R.

- Carl

[1843]

The coil is not underdamped, if it were you would see an overshoot (start of ringing). Try Reg's suggestion, reduce the feedback R to 470k or even 330k. Most high-speed opamps are datasheet-tested with a much lower feedback R, so performance can be quite different for a high R.

- Carl

Ok, I will do that. But how about adding a capacitor in parallel with the 1meg resistor?

[Reg]

Hi 1843,

I am not sure just why you are so concerned about the damping resistor and the hump. If you are sampling very late, then what happens before the time you sample is not important and will not affect your sensitivity.

Now, if you reduce the damper resistance just to try to get a nice perfect curve, then you may be able to do that, but you will also reduce the sensitivity of the detector.

You will want the damping resistor to be as large as possible for maximum depth capability. So, if you are trying to figure out the best damping resistor value so you get the greatest depth capability, then you will try for use the largest possible that can be used and have the signal stable when you sample. The reason is simple, the signal from the coil is dissipated two places. This coil energy is shared by the preamp and the damping resistor. So, the more energy that goes into the damping resistor, the less there is for the preamp. If you want more signal o go to the preamp, then you have to have less go to the damping resistor. To do this, you will want a larger resistor.

If you are trying to obtain maximum depth capability and you are sampling very late like you mentioned earlier, then I would use the largest damping resistance I could and still have things stable when I sampled. I wouldn't care what happened before then. In other words, if it oscillated before I sampled but was stable when I did sample, then I wouldn't worry and use it.

The last thing I would be worried about would be humps or other distortions that occur well before I sampled.

Try different values and see for yourself just how it can affect the depth.

Reg

[1843]

Hi Reg,

I thinked that the hump can cause poor sensivity at higher samples like 120us. So I will try the highest damping resistor untill the coil is stable.
My worrying was due to the coil. Because the my coil has a higher inductance and capacitance than other PIs. So I doubted if it is damped critically or not.

Many thanks to you and Carl!

Now I have another question about my PI;
If I add another diode in series with D2 and D3 to get maximum flyback voltage, can cause increasing in the detection capacity?

Best regards,
1843

[Reg]

Hi Again,

You asked about a capacitor across the feedback resistor, so I will try to explain a couple of things about such a feature.

Most people want to be able to sample as soon as possible to obtain maximum depth capability. The earlier you sample, the greater the signal you get to amplify. The down side of this is if you sample too early, then you also get a ground signal to contend with and that can be a problem.

Now, adding a capacitor across the feedback resistor can do a couple of things. First, it will reduce the high frequency response of the preamp. This will have two different effects. One is, reducing the high frequency signals, will reduce the signals from very small gold or other targets that have a very short time constant. Since you are already sampling late, then this is of no concern.

One other thing a capacitor can do is reduce high frequency noise by reducing the bandpass. If there is less noise then, hearing deep targets will be easier. you want your detector signal to be as quiet as possible so you can hear the weak signals.

Now, if you are sampling late, you don't need any shielding, but if the large coil is picking up too much noise signals, then you might try shielding to see if it gets quieter. Whether shielding will help and allow more depth will be determined more by how much external noise you have to contend with.

One other thing you might want to try and that is to move your subtract signal out to a later delay. When you sample late, then the decay curve is quite linear and nearly flat. At this point, then your subtract sample can reduce your sensitivity because the signal on that sample can be closer to the first sample. This will result in a depth loss.

So, we need to know your objective. If you are trying to get the most depth capability and are sampling late, then you need to modify the detector to do that. This means, you can add a capacitor across the feedback resistor.

If you are looking for large objects, then use a longer pulse width and don't worry about what is happening to the decay curve before you sample. A longer pulse width will allow for a stronger signal from large objects that have a longer time constant.

I personally, would sample as soon as I could where the ground signal was not a problem. Usually, this is about 30 to 50 usec or less.

Reg

[Reg]

I thinked that the hump can cause poor sensivity at higher samples like 120us. So I will try the highest damping resistor untill the coil is stable.
My worrying was due to the coil. Because the my coil has a higher inductance and capacitance than other PIs. So I doubted if it is damped critically or not.

Now I have another question about my PI;
If I add another diode in series with D2 and D3 to get maximum flyback voltage, can cause increasing in the detection capacity?

Anything that happens before the time you sample will not have any impact or cause a problem. If there is oscillation but it stops before your sample, then the oscillation is not a problem.

This doesn't matter regardless of the inductance or the capacitance of your coil. It is what is happening at the time you sample that is important.

You are missing the point of the sampling process. You sample the signal at the output of the preamp at a certain time. The only thing that will happen is you will sample the signal only at that time and anything strange that has already happened and is gone will have absolutely no effect. So, sampling very late could care less about the initial flyback signal, strange humps, etc.

Now, you can try adding diodes to see what happens, but be careful, some amplifiers have built in diodes to protect the amplifier. If this is the case, then adding more diodes in series with D2 and D3 will only cause the internal diodes to have to protect the input. This could result in the preamp being destroyed.

I personally wouldn't do it because it will not gain what you think it will gain.

Reg

[1843]

OK, Now I want to ask another quetion;

The my PI works at 500Hz and its integrator TC is 220ms. Do you think that lowering the TC to 47ms and adding a capacitor across the feedback can improve its detection?

[1843]

Now, you can try adding diodes to see what happens, but be careful, some amplifiers have built in diodes to protect the amplifier. If this is the case, then adding more diodes in series with D2 and D3 will only cause the internal diodes to have to protect the input. This could result in the preamp being destroyed.

Reg

But I think they don't affect the preamp. Because there are 3 other diodes at the input path. D2 and D3 are parallel with the coil only to limit the flyback peak.

Best regards

[1843]

OK, Now I want to ask another quetion;

The my PI works at 500Hz and its integrator TC is 220ms. Do you think that lowering the TC to 47ms and adding a capacitor across the *feedback can improve its detection?

I wanted to write 'preamp feedback resistor'!

[ivconic]

The coil is not underdamped, if it were you would see an overshoot (start of ringing). Try Reg's suggestion, reduce the feedback R to 470k or even 330k. Most high-speed opamps are datasheet-tested with a much lower feedback R, so performance can be quite different for a high R.

- Carl

Now i agree with this.
But i do have further problems; at Delta Pulse is easy to correcrt damper resistor. Coil is not ringing and everything is looking fine. Detection and sensitivity are great. But...playing with knobs(pots) i can reach point where ringing occuring again!? Usuallly if some of pots is turned fully at one end.
So i tried to find exact values for damping resistor,feedback etc. to achieve proper damping and avoid any further ringing but without succes. At the end i put most proper resistors with which detector has greatest depth and most of the time fet remains pretty cool. So (only in educational purposes) i ask you is there something i missed? Is there any other solution for (ocassionally) ringing coils?
Probably all those DP presets are more than sufficient for everyday prospecting. I was thinking to fix few of those with fixed resistances...like Germans done that in DBP2010...

[Reg]

Hi Ivonic,

I would look at all the power supplies to see if one or more of them are going unstable. I found this happens quite frequently if there is no regulator to regulate the main power supply voltage.

When you try to pulse with a lot of current and then increase the gain and audio levels, there can be an interaction which causes oscillation or instability.

I haven't looked at the schematic of your detector, but I would decouple the audio with a low value resistor and large capacitor, and, if there is no regulator on the main power supply, I would add one. By main power supply, I mean the battery supply.

One of the nice things about using li Ion batteries is the battery voltage of 4 batteries in series is 14.4V which regulates down to 12 volts very nicely with a simple regulator.

Always use quality capacitors, especially around the preamp. In fact, I use tantalums for the best results.

Now, one other thing, some amplifiers will change a little when overloaded or the delay is changed. In other words, all may look ok, but not really be ok. I have found this to be true when using the NE5534. Also, there are differences in NE5534's also. For some reason some brands work better than others. TI's appear to be the worse when trying to sample very early. This is a really hard problem to predict because what happens can vary a little.

Reg