Differential Amplifier

[pm4698]

I want to take the signal from OUT label and hold its peak. I do not know how much time i have to hold it.

I want to keep the last circuit you sent me (Delay_Int3.asc ) but the signal from OUT label to be at about 10Volts peak (50mV input - Gain 200 - 10V output) and i want to have the output of the image you attached at the message that you refer that the integrator is too slow. In this image it seems that when the pulse appears the ouput gives the peak but with an amount of delay. I do not want this delay (or minimum delay) and i want the peak to be at 10Volts and not 6V as shown in that image.

If i want to use this value of Vout, should i save it somewhere or what?

P.S. Thank you a lot for your help

 

[ericgibbs]

I want to take the signal from OUT label and hold its peak. I do not know how much time i have to hold it.

If i want to use this value of Vout, should i save it somewhere or what?

hi pm,
I keep asking what is the output connected to.? is an ADC etc or what.?

 

[pm4698]

Sorry about not noticing

I would like to connect it to an ADC and then store the values somehow

 

[ericgibbs]

I want to keep the last circuit you sent me (Delay_Int3.asc ) but the signal from OUT label to be at about 10Volts peak (50mV input - Gain 200 - 10V output) and i want to have the output of the image you attached at the message that you refer that the integrator is too slow. In this image it seems that when the pulse appears the ouput gives the peak but with an amount of delay. I do not want this delay (or minimum delay) and i want the peak to be at 10Volts and not 6V as shown in that image.

Just seen your edit.
The output from Mikes amp is 8V, because the input signal is 40mV, that is a gain of 200!

I dont see a 6Vpk signal in my posted image, unless you mean the image just showing the integrator.??
They are just test inputs, they show that the integrator is not suitable for fast operation.!

The time you want to hold the peak is important.

 

[pm4698]

The whole circuit (MDL etc) will function as a sensor. When the magnetic field changes i receive an output Voltage from the coil (50mV). I want to amplify this voltage (10V) and hold its peak. This voltage is proportional to the change of the magnetic field.

As a result, at the end of my circuit i want to see a peak and then i can calculate a result about the magnetic field (not relevant with the circuit). How can i calculate the time i want to hold this peak?I just want to see it or save it somewhere (through an ADC or something like that) and then conclude a result about the magnetic field

 

[ericgibbs]

The whole circuit (MDL etc) will function as a sensor. When the magnetic field changes i receive an output Voltage from the coil (50mV). I want to amplify this voltage (10V) and hold its peak. This voltage is proportional to the change of the magnetic field.

As a result, at the end of my circuit i want to see a peak and then i can calculate a result about the magnetic field (not relevant with the circuit). How can i calculate the time i want to hold this peak?I just want to see it or save it somewhere (through an ADC or something like that) and then conclude a result about the magnetic field

hi,
Look at this pdf, this a sample hold ic.

 

[ericgibbs]

hi pm,
This circuit would be an option for an amp and sample and hold.

It removes the 2.4V offset from the output signal.
I am assuming that the output signal of the amp will always
be greater than about 5mV in order to trigger the comparator which generates the 'hold' pulse.

In practice the Reset circuit is not required, unless you want to have a 'zero' sample ref for the circuit.

 

[pm4698]

The circuit you sent works just fine!

Thank you very much!!

I have a few questions..

1)At sample gate, before 'E' end, should i place a pull-up or pull-down (47K)?
2)Because my provider cannot find easily the LT's, could i replace them with AD8072?
3)From the coil, i get two pulses. One positive and one negative. So, i suppose that if i connect at + end of the first LT1215 the coil the other end will go to ground. This means that i will not be able to see the negative pulse. In this case, i suppose that i will use the first amplifier - circuit (from previous replies) and at the end i will use the sample-hold circuit.
4)If i do as said at 3), then, should i use a current supply between the coil and the first LT1215 (because of the large resistors)?

Thank you again for your time

 

[ericgibbs]

The circuit you sent works just fine!

Thank you very much!!

I have a few questions..

1)At sample gate, before 'E' end, should i place a pull-up or pull-down (47K)?
2)Because my provider cannot find easily the LT's, could i replace them with AD8072?
3)From the coil, i get two pulses. One positive and one negative. So, i suppose that if i connect at + end of the first LT1215 the coil the other end will go to ground. This means that i will not be able to see the negative pulse. In this case, i suppose that i will use the first amplifier - circuit (from previous replies) and at the end i will use the sample-hold circuit.
4)If i do as said at 3), then, should i use a current supply between the coil and the first LT1215 (because of the large resistors)?

Thank you again for your time

hi,
I have been comparing the two OPA datasheets, it looks as though the AD8072 should work OK.

The coil sense should be such that the negative 'end' is connected to 0V and the coil end that produces the positive output goes to the OPA input.

You CANNOT make a hybrid of the two circuits.!!!

I dont understand what you mean you 'cannot see the negative pulse'.??
You originally stated that is was a positive going pulse that you wanted to amplify and integrate.?

This statement dosnt make any sense

4)If i do as said at 3), then, should i use a current supply between the coil and the first LT1215 (because of the large resistors)?

 

[pm4698]

I read a paper where only the positive pulse (peak) was used for measurements. So, i begun with this skeptic. But, at the laboratory i realized that i need to take in mind the negative pulse too - to consider that at the two ends of the coil, a voltage appears without grounding the one end of the coil.

 

[ericgibbs]

I read a paper where only the positive pulse (peak) was used for measurements. So, i begun with this skeptic. But, at the laboratory i realized that i need to take in mind the negative pulse too - to consider that at the two ends of the coil, a voltage appears without grounding the one end of the coil.

hi,
I dont wish to sound unkind, but I think you are well out of your depth with this project.

A coil will always have a ' negative and positive end'.

A voltage will appear across the ends of the coil.

If you had a amp that produced a pos and neg swing, if you integrated the two polarities you would get zero!

What produces the electromagnetic field that couples with your coil.???

 

[pm4698]

Don't you worry. You don't insult me!As i said i do not know a lot about electronics but i try to follow ( i am aware of the theoretics but in action i do not get along! )

The MDL is actually a ferromagnetic ribbon. When in some place up or down of MDL someone approaches a magnet then at this spot, on MDL an elastic pulse will be produces and its going to travel to both directions of MDL. I

If at one spot, i place a search coil then i get the Vpp of the transmitted pulse. The Vpp is between the positive and the negative pulse of the elastic wave.

 

[ericgibbs]

Don't you worry. You don't insult me!As i said i do not know a lot about electronics but i try to follow ( i am aware of the theoretics but in action i do not get along! )

The MDL is actually a ferromagnetic ribbon. When in some place up or down of MDL someone approaches a magnet then at this spot, on MDL an elastic pulse will be produces and its going to travel to both directions of MDL. I

If at one spot, i place a search coil then i get the Vpp of the transmitted pulse. The Vpp is between the positive and the negative pulse of the elastic wave.

hi,
If I understand your description correctly.

Your sensor coil will be placed at say one end of the MDL, aligned so the coil length is along the axis of the MDL.?

If a magnet is passed close to the MDL a electromagnetic wave will travel towards the coil and be detected by the coil.?

Question about the magnet passing close to the MDL, will the same POLE of the magnet be moved in the same direction relative to the MDL.?

If its a electomagnet causing the remote pulse on the MDL, will the current and coil direction relative to the axis of the MDL always be the same.?

Please let me know the answers, also a simple diagram would be helpful.

EDIT:
MDL were once often used in computers, the MDL was simple coil, wound like a clock spring. the driver coil was at one of the coiled MDL and the detector at the other.

They were used as simple data string memory devices, are you using the MDL in the same way.????

 

[pm4698]

1-The sensor coil will be placed so that the coil length is along the axis of the MDL.
2-If a magnet is passed close to the MDL then an elastic wave will travel towards the coil and to the opposite direction as well. So, the whole voltage that the coil will detect will be at about 50mV.
3-The place(height from MDL) of the magnet will determine the magnetic field. If it approaches more the MDL, different voltage will be seen. It doesn't matter which POLE, as long as magnetic field exists.
4-The whole magnet-story, is the general way that an MDL-sensor works. In my circuit i will not have any magnet but i will detect changes on magnetic field with some other way.So, it doesn't matter the type of magnet but the general idea.

At the image attached, you can see the basic idea of MDL. You have an excitation coil (or a moving magnet) which produces elastic wave which travel to both directions. There is a sensor coil, which senses a Voltage(of elastic wave). If you change the parameters of excitation coil (or magnet for example) then different elastic wave travels, so we get different voltage at the sensor coil.

So, as the elastic wave has positive and negative pulse, if we ground the coil, i suppose that we could not sense the negative pulse and we would take only the positive pulse peak.

I hope that this helped

 

[ericgibbs]

hi,
As understand that OK.

Mike's or my circuit will not work with a negative [reversed input].

In order to get a +/-10v swing you need +/-12v supply rails to the OPA's.

Also you should consider the ADC, I dont know any that require +/-10V input!

Typically +/-2.5V or +/-5V

How many bits does the ADC have.?

 

[pm4698]

For the first stage, i am not going to use an ADC.

I just want to conclude to some results for my experiment.

After that, maybe i will use an ADC, so i haven't thought how bits should it have.

Couldn't i place some resistors to drop the voltage if i wanted to use an ADC later?

 

[ericgibbs]

For the first stage, i am not going to use an ADC.

I just want to conclude to some results for my experiment.

After that, maybe i will use an ADC, so i haven't thought how bits should it have.

Couldn't i place some resistors to drop the voltage if i wanted to use an ADC later?

hi,
If you dont have to go to a +/-10v signal I would advise against it, I would consider +/-2.5V.

This will simplify the amp design and help to reduce the amps response time, also it will keep it within the specification of more OPA's ADC's and logic gates.

 

[pm4698]

The +/- 10 Volts is not critical. Of course i can use +/- 2.5 Volts. The +/- 10 Volts was a first choice, but i can manage the signal with +/- 2.5 Volts too.

 

[ericgibbs]

The +/- 10 Volts is not critical. Of course i can use +/- 2.5 Volts. The +/- 10 Volts was a first choice, but i can manage the signal with +/- 2.5 Volts too.

hi,
OK, I'll look at the circuit design.

I think MikeMI must be busy on another project.

 

[pm4698]

Thank you for your patience and time!