chance pi build

[green]

.[ Has anyone built a similar two stage op-amp using two of the OP37 chips and a nominal +12V & -12V supply?]

The answer is no but I would try a simple modification to what is there. If it works then add the second stage. Using the input resistor as R damping saves a couple of usec. Remove R13, add diode across R25, change R25 to 20k, adjust R16 for R damping.

[baum7154]

Thanks Green,

This is essentially what I did yesterday in my testing to see how fast the output of the OP37 would come out of saturation at a gain of 21(6us). I just paralleled the 470K resistor with a 22k and left the original diode mod and 470k in place on the board. I didn't remove the damping resistor as I have an adjustable damping network permanently in place that allows me to operate at the minimum Guard Interval (10, 8us)with the fast coil over varying types of soil. The testing did prove that it is worth while to do the 2 stage amp and I will use the original amp as a much lower gain (20 to 25) first stage .

Regarding using R16 as the only damping resistor, it is not tied directly to ground, but has the protection diodes and opamp between it and ground. I'm afraid flyback would eat the diodes and amp without my adjustable damping net in place directly across the coil.

I envision a daughter board connected through the 8 pin socket for the OP37 now in place on the main board. Both OP37 amps will be moved to the daughter board and hopefully this will make the mod a 'plug and play' situation perhaps with some additional grounding/shielding/bypassing. If the amp can be made to perform with stability and allow 6us or less in order to come out of saturation, then the fastest allowed start of sampling at 8us will be improved by not having any vestige of the coil flyback decay present in the sample.

Input at the 1st stage amp is about 130mv for the coil flyback and this drives the amp to saturation, not sure what the desired signal voltage typically is for a 10 grain gold target, but the detector definitely does see it as it is now.

The question I am trying to resolve now is the optimization of the non-inverting second stage and it's connection to the 1st stage. (1k resistor coupling?) Should I include a diode across the 2nd stage feedback resistor as is done in the first stage? I think this is to limit stage gain during output excursions above 0.6v.

I'm thinking I need to just etch a small board to allow it to be 'breadboarded' and optimized in place in the detector.

I'm glad to hear any suggestions.

Regards,

Dan

[green]

[Regarding using R16 as the only damping resistor, it is not tied directly to ground, but has the protection diodes and opamp between it and ground. I'm afraid flyback would eat the diodes and amp without my adjustable damping net in place directly across the coil.]
The op amp ties R16 to ground with the diode added across R25. I'm thinking you already tried disconnecting R16 when you measured circuit resonance. The diodes still see the fly back through R16 , but now they have to dissipate all the energy. Removing R13 should decay faster if wanted.

[The question I am trying to resolve now is the optimization of the non-inverting second stage and it's connection to the 1st stage. (1k resistor coupling?) Should I include a diode across the 2nd stage feedback resistor as is done in the first stage? I think this is to limit stage gain during output excursions above 0.6v.]
I would first try connecting a 1k from first amp out to + in of second amp. With a 200 ohm to ground and a 5k feedback resistor to - input of the second amp (10k for more gain if it would work). No feedback diodes. Add 100n bypass caps to power pins. Probably need a balance adjust to - in of the first amp. Just my thoughts

Looking at the OP37 specs. Max diff. input .7 volts. Thinking + input of first amp needs 50 ohm to ground.

[baum7154]

Thanks Green,

I forgot to mention that I don't have the D9 diode to ground at the input to the A/D because it is a modification called out for the use of the 1.2.1 software along with the addition of the diode across the 470K feedback resistor that I already have.

I did disconnect R13 for a short term self resonance test with no damage to the diodes or amp. Looked at the specs on the 1N4148 and the repetitive reverse voltage is 100v and 150ma forward current/2A surge so they probably took the brunt of the flyback. In any event my coil damping is good at 6us and I like the ability to tweak damping over wet/mineralized soil so I'll keep it as is. With this coil the typical R13 damping resistor value in air is adjusted to 1040 ohms. This mod was recently added and since the pictures were posted a while back.

1K second stage input sounds good to me too.

No feedback diode even on the first op-amp? Are you thinking it won't be necessary due to the low stage gain?

Bypass caps are easy and good practice.

The OP37s I bought are balanced trimmed at the factory to 10uv & 10na, a little pricey. Will this negate the need for the balance adjust you mention?

Is the 50 ohm first stage + input resistor recommended to insure that the .7v differential input is not exceeded?

It will be interesting to see what effect going beyond the design gain value of 470 to as much as 1000 will have.

I'll get to the task of board design and build.

Thanks!

Dan

[green]

[No feedback diode even on the first op-amp? Are you thinking it won't be necessary due to the low stage gain?]
Put diode on first amp.

[The OP37s I bought are balanced trimmed at the factory to 10uv & 10na, a little pricey. Will this negate the need for the balance adjust you mention?]
Might not need adjustment. I see OP37 has adjustment input pins that might be better than the - input.

[Is the 50 ohm first stage + input resistor recommended to insure that the .7v differential input is not exceeded?]
The 4148 drops over a volt at 200ma. 20 ohms is enough to keep the current below 25ma. With such a short duration maybe isn't needed.

[green]

[I forgot to mention that I don't have the D9 diode to ground at the input to the A/D because it is a modification called out for the use of the 1.2.1 software along with the addition of the diode across the 470K feedback resistor that I already have.]
Do you have two diodes back to back. A diode to replace D9 would have the cathode connected to the output not to - in.

[baum7154]

No the only back to back diodes are across pin 2 inverting input and pin 3 non-inverting input.
On the feedback resistor there is one with the cathode end also on pin 2. It is installed just as in your earlier attached drawing.

[green]

[On the feedback resistor there is one with the cathode end also on pin 2. It is installed just as in your earlier attached drawing.]

Not what I would expect. I think the added diode in the drawing limits output volts to +.6 volts. D9 limits A-D input to -.6 volts.

[baum7154]

However it was never intended that they both be installed at the same time…they are mutually exclusive. D9 is in place for the old V0.8.4 software and it is removed and another installed across the 470k feedback resistor when running the latest V1.2.1 software…and we don't have the source code…yet.

[Waikiki_Sweep]

Did a bit more scope time in the CHANCE PI with an 8" fast spider coil today. Measured the time it takes for the OP37 to come out of saturation on the flyback signal at various gain settings. The coil itself measured at the protection diodes decays in 6us. With the designed gain of 470 the OP37 amp takes right at 16us to come out of saturation and smoothly back to zero with proper damping. Dropping the gain to 43 the amp settles to zero in 8us. Further dropping the gain to 21 allows the amp to settle to zero in about 6us, virtually the same as the coil itself. I took pictures of these test measurements for future reference.

I intend to build a 2 stage op amp, 1st stage inverting with a gain of 20 to 25, and the 2nd stage non-inverting with a gain of 20 to 25 to allow better/faster sampling for small gold targets. Has anyone built a similar two stage op-amp using two of the OP37 chips and a nominal +12V & -12V supply? No sense re-inventing a good wheel!

I appreciate any responses.

Dan

I have a version to tell. May be saturation and slowness of OP37 is not that problem.
What if you can see just an overload of input of preamp.

When you reducing gain you just increasing the range of signal you will amplify.
So you see decay curve in 10 microseconds earlier.

Than if you add a second amplifier you will have that old gain value and the same input overload and same 16 microsecond flat top on your decay curve.

Sure two stages amplifier will have higher slew rate but may be it will give only couple microseconds speedup not a 10.

Anyway results of such experiments are always interesting.

[baum7154]

Yes this is a concern and that is why I have to build the 2 stage amp to give it a try and see if it can be optimized. I see many comments on this forum that the dual op amp will overcome the saturation problem and even allow a bit more gain if noise is kept within acceptable limits. I have the board geometry developed now for an easy implementation of the 2 stage amp. Now I have to build it and test it. Will keep you informed on the results here.

Regards,

Dan

[baum7154]

Than if you add a second amplifier you will have that old gain value and the same input overload and same 16 microsecond flat top on your decay curve.

Sure two stages amplifier will have higher slew rate but may be it will give only couple microseconds speedup not a 10.

Anyway results of such experiments are always interesting.[/QUOTE]
---------------------------------------------------------------------------------------------
I am thinking it might help to put a second back to back diode limiter pair at the input of the second op-amp. That way the flyback signal i limited to about .6v but the weak target responses may not be limited at all and so sail right past the diodes to be further amplified by the second stage and still avoid saturation with the low gain of approximately 25.

Dan

[Waikiki_Sweep]

I am thinking it might help to put a second back to back diode limiter pair at the input of the second op-amp. That way the flyback signal i limited to about .6v but the weak target responses may not be limited at all and so sail right past the diodes to be further amplified by the second stage and still avoid saturation with the low gain of approximately 25.

Dan

You may try to put that diodes between output of opamp and its negative input like D9 has been placed after Chance update. It will limit input to that very low range which limiting output to + - 0.6v.
Because ADC is used to read voltages after amplifier so output of + - 0.6v looks OK. Anyway you can try.

[baum7154]

I may try that but I think it is better to clamp the amplified flyback input to the second stage at its input thus stopping saturation before it starts. I may add the feedback diode too…have to see how it works in the real world.

[Waikiki_Sweep]

I may try that but I think it is better to clamp the amplified flyback input to the second stage at its input thus stopping saturation before it starts. I may add the feedback diode too…have to see how it works in the real world.

If gain=25 than 0.6v x 25 = 15 v

Looks like clamping it at 0.6 v is not enough because opamp will be saturated anyway.

0.6 v / 25 = 0.024 v With feedback diodes input voltage will never left + - 0.024 v range as well as output will always be in + - 0.6 v range. As you can see there is no any saturation at all at any moment of time.