How to create a Op amp based Ramp?

Thread Starter

picstudent

Joined Feb 3, 2009
91
I am doing a op-amp based comparator over current cut off. (Hope idea evident) But the comparator keeps on and off because of closed loop.:confused:

How to create a ramp like mechanism in op amp to prevent it from releasing the cut off for a few seconds (adjustable)? Each comparator switching to positive side should hold for a few seconds, then try again.

Thanks
Roy Thomas
 
Last edited:

Thread Starter

picstudent

Joined Feb 3, 2009
91
Hello,
I think you need to build in a hysteresis.
Bertus
I think something more is needed. Hysteresis adds some sort of a tolerance to avoid sudden to and back transitions. (Right?)

But in my case I think that wont be sufficient,beacuse when MOSFET goes off current will be 0 within few nano seconds or microseconds. So substantial change to overcome the hysterisis will be back soon.

I think I need a delay for switching (only in one direction) Comparator fires - but should stay there for sometime irrespective of input conditions.

Inn short circuit condition it will then try connecting every 10seconds or so.
 

Ron H

Joined Apr 14, 2005
7,063
If you provide a schematic of what you have, we might be able to suggest a simple (or complicated :D) solution.
 

bertus

Joined Apr 5, 2008
22,270
Hello,

I think I was wrong with the integrator.
I think you need a monostable multivibrator (like a 555).
When the current trips(the output of the comparator), the 555 will start and can hold the input low, atfher the time if the 555 the circuit will be released.

Greetings,
Bertus
 

Ron H

Joined Apr 14, 2005
7,063
I agree with Bertus.
Here is a way to make a monostable with your existing circuitry. I simulated it. The sim times out in about 12 seconds. Change R1 and/or C1 to change the timeout. Note that I eliminated Q3 and R17 from your circuitry.
I would use a nonpolarized cap for C1. If you use a polarized cap, the + end is to R1.
 

Attachments

Thread Starter

picstudent

Joined Feb 3, 2009
91
I agree with Bertus.
Here is a way to make a monostable with your existing circuitry. I simulated it. The sim times out in about 12 seconds. Change R1 and/or C1 to change the timeout. Note that I eliminated Q3 and R17 from your circuitry.
I would use a nonpolarized cap for C1. If you use a polarized cap, the + end is to R1.
Thanks a lot for the help. Now I can save one op amp or 555. Thanks a gain. Will learn and implement the circuit.

Roy
 

Thread Starter

picstudent

Joined Feb 3, 2009
91
I have simulated he circuit.It works beautifully for my purpose.But while I connected the closed loop setup it fails to simulate."Too many iterations without convergene","Gmin stepping failed". Proteus help says "mathematics of the circuit may be week"
Any clues? The problem starts when I connect the closed loop to control MOSFET.

Where I should search in circuit?
By keeping the the current minimum using a pot I can start simulation,But when I increase current to cut off region it starts oscillating. ON and OFF

Thanks
 

Attachments

Last edited:

Thread Starter

picstudent

Joined Feb 3, 2009
91
What happens in simulation is that ,it keeps on and immediately off every 5 milliseconds.
Wave form of MOSFET gate, load. comparator input and comparator output is attached.

Thanks
 

Attachments

Thread Starter

picstudent

Joined Feb 3, 2009
91
The first charging cycle in simulation gives proper response.After that on and off at 5ms is occurring.
I feel C1 is not getting discharged to start the next charging cycle. Is it the clue?
Thanks
 

Ron H

Joined Apr 14, 2005
7,063
The first charging cycle in simulation gives proper response.After that on and off at 5ms is occurring.
I feel C1 is not getting discharged to start the next charging cycle. Is it the clue?
Thanks
My simulation doesn't do that. How are you simulating your short circuit?
 

Thread Starter

picstudent

Joined Feb 3, 2009
91
My simulation doesn't do that. How are you simulating your short circuit?
What i did was reduce the load resistance(Probe) to a low value so that current will be more than 500mA. Then simulation cannot start. So I tried by short circuiting the probe.Then also cannot start, But then I replaced the probe resistor with a variable one. Started simulation with a safe current and then moved the variac to increase current to above 500mA. Around 500mA it will on and off like a astable multivibrator of about 6 seconds on and same time off. Then I increased current to a high value like 800mA or somewhere, it started showing the 5mS repeated triggering of MOSFET.

I raised the discharge doubt beacuse when we move the POT to bring current to a high value for the first cycle the monooshot will be ok. But after the time out MOSFET will be ON to try again, from this moment onwards it starts repeating at 5mS.

Thanks for the time.
btw if you have or is using proteus for simulation I will send the DSN file if required
Regards
 

Ron H

Joined Apr 14, 2005
7,063
If you remove the short before the oneshot times out, then it works correctly. If the short still exists after the oneshot times out, then it oscillates, as you say. I'll have to think about this some more.
 

Thread Starter

picstudent

Joined Feb 3, 2009
91
I'll have to think about this some more.
Thanks a lot for the help. Take your own time. I learned this much mainly by help from all here. So wait is no problem. In the meantime I will keep thinking on this and move to other parts of the circuit. I don't think I will be able to find a solution by myself.
Thanks a lot
 

Thread Starter

picstudent

Joined Feb 3, 2009
91
If you remove the short before the oneshot times out, then it works correctly. If the short still exists after the oneshot times out, then it oscillates, as you say. I'll have to think about this some more.
Any ideas? Still I am stuck there.

Thanks
 

Ron H

Joined Apr 14, 2005
7,063
OK, I think I've got it. If the short still exists at the end of the oneshot timeout, the MOSFET will turn back on for a few milliseconds before the oneshot is allowed to be triggered again, giving the oneshot time to recover before it fires again.
I can post a schematic if this sounds like what you want.

EDIT: The oneshot will continue to trigger approx. once every 10 seconds (or whatever time you want, within reason) while the short exists.
 
Last edited:
Top