Small signal to 5V output amp

[camerart]

Hi,

I have 2X pick up coils with outputs varying 0.1 to 10V Approx AC.

I need to switch a motor forward or reverse depending which coil reading is highest or lowest.

I can input 2X A/D converters on a PIC chip and program the motor to turn. Or I can probably use a 2X channel analogue circuit like a servo.

I have tried Op amp circuits which have a 2n3819 fet Transistor as a pre-amp, but after many attempts, I can't seem to get consistency.

Hope you can help.

Cheers, Camerart.

 

[MikeMl]

Are you asking how to build an analog comparator which compares two varying signals, decides which is larger, and then switches a relay to reverse a motor?

Should the motor always be running full speed, or are you trying to make the motor speed proportional to the DIFFERENCE between the two signals?

What supply voltages do you have to work with?

 

[camerart]

Hi MikeMl,

I am trying compare 2X coil outputs, they should be equal. The coils are sensing a wire with a signal passing through it. If coil 1 is larger than coil 2 then turn a motor one way so that coil 1 moves away and coil 2 moves towards the wire, and visa versa, keeping them central.

The voltage is 12V.

I have been experimenting with an circuit for more than a year, that has a 2n3819 fet amplifying the coil, then into an opamp (5V) which changes AC to DC then into 2X 5V A/D inputs of a PIC chip programmed to compare the A/D inputs. This does work, but doesn't seem to be stable. I think probably it's too complicated. I think that the range of 0.1 V to 10V APPR, is too large making the A/D inputs too fine.

I found when using Relays that the fast switching sometimes makes -+, +-, and they blow. So I made a circuit using Mosfets H bridge and a protective PIC chip, again to stop switching too fast. Again perhaps too complicated, but working ok.

I wonder if a simpler analogue or digital circuit could be made. Perhaps similar to a Servo using the difference between the coils to change the H bridge balance? I am happy to use a Mosfet H bridge, one the comparisons have been made. Proportional is best but full speed is fine as the Motor is geared. If possible a small Null point is good.

Cheers, Camerart

 

[camerart]

If my first reply is too long winded, here's a shorter one:

Is it possible to make a circuit that will convert a pick up coil signal with a range of approx 0.1v to 10v AC to give an output of 0v to 5v DC? (X2)

The Batteries I use are 12V with a 5v Regulator.

I can then send 2X signals to PIC A/D inputs, and use my existing comparison program in the PIC and the existing H Bridge Motor Drive.

 

[crutschow]

To get an accurate DC output from the low 0.1V AC voltage, you can use a precision rectifier op amp circuit plus filter such as this (Figure 12.35).

P.S. For an AC signal you should have a separate plus and minus supply voltage for the op amp. Also if you need 5V from a 5V supply, you will need rail-to-rail type op amps.

 

[camerart]

Thank you, Crutschow,

The WIRE signal is 55khz. Can you tell me what values are the components are shown? OR, I see calculations on the next page, if these are for this circuit, I could probably work them out in time.

(Sorry I'm not sure what I'm asking here) I think I understand rail to rail. Does the chip sit in the middle of + and-v? If so, do both one side of the coil and 0V connect together?

 

[ronv]

Did you get a circuit yet?

So when the coils are in balance both ac signals are like 5 volts or are they just equal?

 

[camerart]

Hi Ronv,

I have been trying to build the circuit from the Crutschow message. but it is an Opamp circuit, very similar to the types I have been trying and failing with.

Basically, the problem is the range of Voltage. 0.1V to 10V approx. If I could get a reliable circuit to amplify a coil, making 0.1V AC = 0V DC and 10V AC = 5V DC, this would be perfect, I would need one for each coil. I then put the outputs through 2X PIC a/d chip to compare them.

 

[camerart]

Hi Crutcshow, I have been trying the circuit (Fig 12,35) And it is similar to ones I have tried and failed, I don't know why, but the circuits I have only have 0V-12V, I now know what Rail to rail means, but can I get this from 0V-12V making 0V half way between, by dividing with 2X same value resistors?

 

[blueroomelectronics]

What are you trying to build? Post the schematic.

 

[ronv]

Peak Detector

Maybe something like this would work for you. The op amp still needs to be rail to rail for the .2 volt signal.
Not sure how accurate you need it to be.

 

[camerart]

Hi Ronv,

Thanks for that, I will have a go at building a test circuit. even though it uses Opamps which I haven't had success with, everybody so far has suggested them, so they seem to be the best answer.

However I will need to buy the components, and I'm away from home at the moment, so it may be a while tillI get back to you, but thanks.

 

[camerart]

Hi Blueroomelectronics, I have been playing with a circuit with Opamp feeding A/D PIC chip inputs, with a pre amp 2n3819, but even though it works, it isn't reliable, and gives wandering readings.

To save me explaining again, see my replies to others, please.

Thanks.

 

[camerart]

P.S. Accuracy isn't so important as long as the 2X coil signal readings are constant enough for PIC A/D comparison.

 

[ronv]

Should be ok. You will need to divide the outputs by 2 to get the 10 volt level down for the PIC. Will you build the deadband into the PIC?

 

[camerart]

OK.

I don't quite understand Rail to rail. I use 12v Battery, so how do I connect the power? Are all of the 0V Arrows connected to Gnd/0V or must they be -6V?

Also What is Deadband?

I can't quite see the component values on your schematic. Is D1-2 1N914 (I have 1N4001 and 1N4148) Is U1 LM339 and is U3LT1013 (I have LM324 LM358 TL081-2)

Thanks.

 

[ronv]

Rail to rail means the input can swing to ground and to Vcc without problems for the op amp. It also means the output can switch from ground to Vcc. The op amp must work with the input at only .2 volts and the output must go down to .2 volts. Your op amps are ok for this. Your op amps may be marginal with the 10 volt signal and the 12 volt power supply. Spec. is Vcc -2 volts.
The ground symbols are 0 volts.
I think you are building a servo with this circuit. You will need some band where the motor is not told to move otherwise it will always be seeking zero error and not being able to find it. I call this deadband. You should be able to do this with the PIC. If for example the A to D is within a few counts (lets say 10) don't drive the motor in either direction.
Bottom line... The circuit will probably be ok with your parts.

 

[camerart]

Ok, thanks, I now understand rail to rail.

What I have already is: A 2X channel coil circuit, going through a PIC chip, and a cmos H bridge Motor Drive. The existing Coil circuit is the problem section. As an option I was quite happy to have a 'New Coil circuit' and add it to a 'new' analogue Servo circuit. The existing PIC chip has a good program for working with a stable Coil circuit and has a 'Deadband', and if your circuit is good, it will be fine with my existing circuit. P.S. it also copes with, if there is no Signal it stops the Motor drive.

With my existing Coil circuit, one of the problems I came across, that I was baffled with, it couldn't cope with the Input going more than 0.3V negative. I am an amateur, but read in the LM339 information PDF that it also suffers from the same problem. Does your circuit get round this problem?

 

[ronv]

The resistor and diode on the input of the comparator protects against the negative voltage.
See page 8 for diode clamp or page 12 for a better transistor clamp.

https://www.electro-tech-online.com/custompdfs/2011/03/LM339.pdf

 

[camerart]

Ok, Ronv,

I'll make a circuit, with what I've got. Also I have also ordered the components from your Schematic, my normal supplier has LM139 not LM339, I think it will be ok though. Let me know if definitely not.

Thanks.