Maker Pro
Maker Pro

Simple comparator with delay?

M

MRW

Jan 1, 1970
0
Hi all!

I have this:
http://bayimg.com/fADbOAABN

The image shows the input and output voltages given certain
conditions. I'd like the output to equal Vsupply+ when the input
voltage passes the voltage reference. But I'd also like the output to
have delay so that it does not go to Vsupply- until a x seconds after
the input voltage drops below the reference voltage.

Will the circuit that I included work? I figured if I charge the
capacitor at the input then depending on the parallel resistor then
its discharge will take sometime.

Any thoughts? Thanks!
 
J

Jamie

Jan 1, 1970
0
MRW said:
Hi all!

I have this:
http://bayimg.com/fADbOAABN

The image shows the input and output voltages given certain
conditions. I'd like the output to equal Vsupply+ when the input
voltage passes the voltage reference. But I'd also like the output to
have delay so that it does not go to Vsupply- until a x seconds after
the input voltage drops below the reference voltage.

Will the circuit that I included work? I figured if I charge the
capacitor at the input then depending on the parallel resistor then
its discharge will take sometime.

Any thoughts? Thanks!
Well.
First of all. You didn't specify if you wanted a delay ramp up to the
Vcc+ voltage or just a delayed digital signal which is a resulting
square wave..
Any ways, you need to have a series R with the input of your existing
circuit.. Also.., You need a little hysteresis to prevent a slop/ramp
effect on the output when the threshold is reached.

Putting a little feed back from the output to the + input can solve
that issue.
Personally, I would use the first Comparator as a buffer stage to
isolate any effects on the source circuit, and then, use the output
to another op-amp/comparator with a series R and bypass cap as your
Time constant for delay on.. also, place a diode across the R so
that reverse voltage will flow to discharge the Cap quickly when the
signal goes off..
If you're trying to construct a Delay on Timer that needs to only
work if there is a minimum of clean signal, this is the way to do it.

P.S.
A 555 timer can solve this issue.
 
J

John Popelish

Jan 1, 1970
0
MRW said:
Hi all!

I have this:
http://bayimg.com/fADbOAABN

The image shows the input and output voltages given certain
conditions. I'd like the output to equal Vsupply+ when the input
voltage passes the voltage reference. But I'd also like the output to
have delay so that it does not go to Vsupply- until a x seconds after
the input voltage drops below the reference voltage.

Will the circuit that I included work? I figured if I charge the
capacitor at the input then depending on the parallel resistor then
its discharge will take sometime.

Any thoughts? Thanks!

When your input voltage goes back down to zero, it will suck
the charge out of the timing capacitor very quickly, making
the 1 meg discharge resistor pointless. If the input had an
ideal diode in series with it, then it could charge the cap,
but not dump it.

It might be easier to use 2 comparators in cascade, the
first producing a pull down pulse when the input was above
the reference, with its open collector output, dumping a per
charged up capacitor to ground very quickly, but just then,
disconnecting itself when the input fell below the
reference. This one way discharge process eliminates the
need for a diode.

That way, the cap could be recharged positive by a timing
resistor, and the second comparator could compare the
capacitor voltage to the reference to produce the delayed
output.
 
J

John Fields

Jan 1, 1970
0
Hi all!

I have this:
http://bayimg.com/fADbOAABN

The image shows the input and output voltages given certain
conditions. I'd like the output to equal Vsupply+ when the input
voltage passes the voltage reference. But I'd also like the output to
have delay so that it does not go to Vsupply- until a x seconds after
the input voltage drops below the reference voltage.

Will the circuit that I included work?
---
No.
---

I figured if I charge the
capacitor at the input then depending on the parallel resistor then
its discharge will take sometime.

---
The source will discharge it in as long as it took to charge it.

Try this: (View in Courier)


+5V>---------+------+--------+
| | |
VIN>---+ [R1] | [4K7]
| | | |
[POT]<--------|-\ |
| | | >------+
| +--+--|+/ LM393 |
| | | | |
| [R2] +---|--[1M]--+-----+------A
| | | | | NAND Y--+-->OUT
GND>---+-----+------+ +-+-+ +-+-+ +-B |
| | | | | |
A B A B | |
| |
Y Y | A--+
| | +-Y NAND
+-----+ B--+
| |
[1N4148] |
|K |
+-----+------------+
| |+
[Rt] [Ct]
| |
GND GND

You haven't mentioned anything about what your input signal looks
like or what OFF delay you want, so I couldn't put in many values.

Also, the pot might be able to be eliminated or replaced with a
couple of resistors.


All of the gates are parts of a quad 2-input NAND with inputs 'A'
and 'B' and outputs 'Y'
 
Top