Looking to build a single supply OP Amp one shot

Thread Starter

bonanzaman

Joined Oct 15, 2010
6
Hi all,

I'm trying to build a single power supply op amp based one shot.

I'm interfacing to a accelerometer which has a 10mv/g ac wave riding on a 12 dc bias.
I've used a single op amp from a quad package op amp to amplify the signal and now I would like to use that to light an LED.

Since the accelerometer requires 27 volts, I'm using three 9v batteries for power for both the accel and the op amp.

The pulse I'm getting is too narrow to see easily on the LED so I'd like to trigger a pulse from the one shot and use that to keep the LED lit longer.

The input will be a small ac pulse. i would like to extend the pulse to about a second.

Any suggestions?

Thanks

Dave
 

wayneh

Joined Sep 9, 2010
17,496
Couldn't you just use a peak detector to rectify the pulse and charge a cap, which would discharge over a second and hold the op-amp "on" during that time?
 

Thread Starter

bonanzaman

Joined Oct 15, 2010
6
I thought about using a 555 but would like to keep it down to a single IC and I'm currently not using any voltage regulation for the op amp. The package i'm using is a quad op amp.
Could the 555 trigger from the 10mv signal generated by the accel and then I wouldn't need the op amp?

Wayne, I'm not sure about using a peak detector. Not sure what it involves. Could you add a little detail?


Thanks

Dave
 

wayneh

Joined Sep 9, 2010
17,496
Wayne, I'm not sure about using a peak detector. Not sure what it involves. Could you add a little detail?
A brief spike is passed across a diode to the top of an RC tank, a resistor and capacitor in parallel between the op-amp input and ground. The spike takes the voltage high , the diode prevents the voltage dropping, and the cap holds it there, depending on the choice of R and C, with the product of R-times-C being the time constant. For example, 100K times 10µF = 1 second. Output from the op-amp then depends on the voltage on the other pin.
 

Thread Starter

bonanzaman

Joined Oct 15, 2010
6
A brief spike is passed across a diode to the top of an RC tank, a resistor and capacitor in parallel between the op-amp input and ground. The spike takes the voltage high , the diode prevents the voltage dropping, and the cap holds it there, depending on the choice of R and C, with the product of R-times-C being the time constant. For example, 100K times 10µF = 1 second. Output from the op-amp then depends on the voltage on the other pin.
Hi Wayne,
What I guessing is that I still use the initial op amp for amplification then put the peak detector at the input of the next amp.

Sorry If I seem a little slow, but I haven't done a thing with op amps for over 30 years let alone single supply ones.
Need to put my thinking cap on and fasten it down.

Dave
 

Thread Starter

bonanzaman

Joined Oct 15, 2010
6
Hi Wayne and group,

I tried to use the detector and no joy. I got the capacitor to charge some, but when I connected it up to the input to the next amp, it appeared to load it down.
Looks like I may need to use the 555 to generate the pulse.

Any other suggestions?

Thanks

Dave
 

Audioguru

Joined Dec 20, 2007
11,248
Hi Wayne and group,

I tried to use the detector and no joy. I got the capacitor to charge some, but when I connected it up to the input to the next amp, it appeared to load it down.
Looks like I may need to use the 555 to generate the pulse.

Any other suggestions?

Thanks

Dave
Post the schematic you used for a peak detector with its power supply voltages and the part number of the opamp you used.
Any opamp works perfectly from a single supply if its input is biased properly.
 

gootee

Joined Apr 24, 2007
447
Capacitor is too large (and/or resistor too small), probably.

But it depends on the pulse. What is the peak voltage of the pulse? How long does it last?

You might also need to condition the input pulse for the detector, i.e. you might need to remove the DC, and might need to amplify it to be above the diode's threshold (or use an "ideal diode" opamp circuit that rectifies down to zero volts), and might even need to attenuate it with a voltage divider. It all depends on the pulse's size and shape.
 

Thread Starter

bonanzaman

Joined Oct 15, 2010
6
LM148
power is three 9v batteries for 27 volts.

Opamp 1 appears to work ad expected. I get a large signal at the output (10V+). I put the cap at the output to remove the dc and the diode to rectify the signal for the charging of the RC circuit.

I can get the circuit by itself to charge up to 1-200 mv or more but when I hook up the second stage it seems to have a great effect on the charging.
I've played with both the cap and resister.

I've played around with the resistance values on #2 as to not load the rc circuit.

The pulses from the first stage around 300 us and the amplitude is to the top rail or 12 volts peak with lesser peaks after initial stimulus.
 

Attachments

Kermit2

Joined Feb 5, 2010
4,162
try a different arrangement of your 1st opamp output to the second opamp input. Pull out the series cap. or try the diode as a parallel to ground component instead of series. Something is preventing transfer of charge and the second op amp is supplying the problem current by keeping its inputs balanced.

Also the wave shape may need to be cleaned up as suggested earlier. A schmitt trigger buffer might be called for to clean up the 'pulse' you are looking to amplify.
 

Audioguru

Joined Dec 20, 2007
11,248
I removed many useless resistors and removed the coupling capacitor that feeds (and is charged by) the rectifier.

I don't know why the second opamp had a 20k attenuator at its input that cut the signal in half then had two resistors as feedback to amplify the reduced signal back up to the the original signal level.
 

Attachments

Last edited:

gootee

Joined Apr 24, 2007
447
Using Vcc to derive the reference, to "subtract out" the 12 volts so you can amplify only the tiny pulse will probably cause difficulty.

It's a lot more robust if you create a low-pass filtered version of the sensor output and take the difference between that and the original sensor output. That way the absolute DC levels won't matter. Otherwise, tiny differences in the DC levels of the inputs will swamp the output with DC.

My circuit is probably a lot more complicated than it needs to be. But the most important parts are the difference amplifier (U1 and R2 through R4) and the low-pass filter for the sensor (R5, C7, and the U5 buffer). Note that the V3 voltage source is your accelerometer output voltage.

D2 and C3 plus R5, R12, and R9 make up the detector. And I added an adjustable output voltage clamp (U2, D1, and R7, R8) and an output buffer, U3.

You might want to vary some of the component values, to better-suit your signal levels and desired LED on-time. Almost any opamp that can handle the supply voltage should work.

The LTspice circuit simulation file is at http://www.fullnet.com/~tomg/accel.asc.txt . After downloading it, remove the ".txt" from the file name.



http://www.fullnet.com/~tomg/accel.jpg

Edit: My ISP must be having some sort of problems. My link and picture aren't working.
 
Last edited:
Top