switch on at threshold and hysterisis circuit

M

M B

Guest
Hello,

I want to switch on an npn transistor when a sensed voltage exceeds a
certain threshold, say 4v, but I don't want it to switch off again until the
sensed voltage drops a bit below that, say 3v - because the signal will not
be clean.

I suppose I need a schmitt trigger, but I have no supply voltage for the
circit - only the sensed voltage itself which is found across a 10k load
resistor. Is what I want possible? Any help appreciated.

Thanks,

Matt B
 
M B wrote:
Hello,

I want to switch on an npn transistor when a sensed voltage exceeds a
certain threshold, say 4v, but I don't want it to switch off again until the
sensed voltage drops a bit below that, say 3v - because the signal will not
be clean.

I suppose I need a schmitt trigger, but I have no supply voltage for the
circit - only the sensed voltage itself which is found across a 10k load
resistor. Is what I want possible? Any help appreciated.

Thanks,

Matt B
Click to expand...
Use a micro power comparator, like the LT1716. Also, use a 2.5V
micropower reference. Then, set up your input so it powers the
comparator and reference, and a voltage divider. The divider should get
to 2.5V when the input rail gets to 3.5V. The noninverting input goes to
the reference through a resistor. There is a resistor of approximately 5
to 10 times the value from the noninverting input to the output (that's
your hysteresis). The inverting input gets the divider. The output goes
to your circuit you want to power.

As the input is rising, the divider will be below the reference (which
will track the rail up to 2.5V). Once the rail gets to 4V, the divider
will be above 2.5 + the hysteresis, and the comparator will start
sinking current.

Note that you shouldn't pull more than about 40uA from the circuit, or
it'll drag down the voltage across the 10k input resistor. Thus, make
sure your reference isn't going to suck up alot of power, and make the
divider very high resistance. When the circuit it on, it'll pull current
from the input out to ground through the comparator.

--
Regards,
Robert Monsen

"Your Highness, I have no need of this hypothesis."
- Pierre Laplace (1749-1827), to Napoleon,
on why his works on celestial mechanics make no mention of God.
 
"M B" <ffp944-a@yahoo.co.uk> wrote:

Hello,

I want to switch on an npn transistor when a sensed voltage exceeds a
certain threshold, say 4v, but I don't want it to switch off again until the
sensed voltage drops a bit below that, say 3v - because the signal will not
be clean.

I suppose I need a schmitt trigger, but I have no supply voltage for the
circit - only the sensed voltage itself which is found across a 10k load
resistor. Is what I want possible? Any help appreciated.

Thanks,

Matt B
Click to expand...
When the NPN is switched on, where will its conducting current come
from if you have no supply voltage? From the sensing circuit? What is
the end objective?

--
Terry Pinnell
Hobbyist, West Sussex, UK
 
"Terry Pinnell" <terrypinDELETE@THESEdial.pipex.com> wrote in message
news:22n231dq5t87srb2fc757ofia7jnqu8lso@4ax.com...
"M B" <ffp944-a@yahoo.co.uk> wrote:

Hello,

I want to switch on an npn transistor when a sensed voltage exceeds a
certain threshold, say 4v, but I don't want it to switch off again until
the
sensed voltage drops a bit below that, say 3v - because the signal will
not
be clean.

I suppose I need a schmitt trigger, but I have no supply voltage for the
circit - only the sensed voltage itself which is found across a 10k load
resistor. Is what I want possible? Any help appreciated.

Thanks,

Matt B


When the NPN is switched on, where will its conducting current come
from if you have no supply voltage? From the sensing circuit? What is
the end objective?

--
Terry Pinnell
Hobbyist, West Sussex, UK
Click to expand...
Hi Terry,

I have a sealed "off the shelf" timer which contains a litium battery. It is
looking for closed contacts but seems to work fine across the collector and
emmiter of a switched on npn transistor.

Matt B
 
"Robert Monsen" <rcsurname@comcast.net> wrote in message
news:bOCdnakEANcXrazfRVn-3Q@comcast.com...
M B wrote:
Hello,

I want to switch on an npn transistor when a sensed voltage exceeds a
certain threshold, say 4v, but I don't want it to switch off again until
the sensed voltage drops a bit below that, say 3v - because the signal
will not be clean.

I suppose I need a schmitt trigger, but I have no supply voltage for the
circit - only the sensed voltage itself which is found across a 10k load
resistor. Is what I want possible? Any help appreciated.

Thanks,

Matt B

Use a micro power comparator, like the LT1716. Also, use a 2.5V micropower
reference. Then, set up your input so it powers the comparator and
reference, and a voltage divider. The divider should get to 2.5V when the
input rail gets to 3.5V. The noninverting input goes to the reference
through a resistor. There is a resistor of approximately 5 to 10 times the
value from the noninverting input to the output (that's your hysteresis).
The inverting input gets the divider. The output goes to your circuit you
want to power.

As the input is rising, the divider will be below the reference (which
will track the rail up to 2.5V). Once the rail gets to 4V, the divider
will be above 2.5 + the hysteresis, and the comparator will start sinking
current.

Note that you shouldn't pull more than about 40uA from the circuit, or
it'll drag down the voltage across the 10k input resistor. Thus, make sure
your reference isn't going to suck up alot of power, and make the divider
very high resistance. When the circuit it on, it'll pull current from the
input out to ground through the comparator.

--
Regards,
Robert Monsen

"Your Highness, I have no need of this hypothesis."
- Pierre Laplace (1749-1827), to Napoleon,
on why his works on celestial mechanics make no mention of God.
Click to expand...

Thanks for that Robert,

It sounds like a very elegant solution.

Regards,

Matt B
 
Terry Pinnell wrote:

"M B" <ffp944-a@yahoo.co.uk> wrote:


Hello,

I want to switch on an npn transistor when a sensed voltage exceeds a
certain threshold, say 4v, but I don't want it to switch off again until the
sensed voltage drops a bit below that, say 3v - because the signal will not
be clean.

I suppose I need a schmitt trigger, but I have no supply voltage for the
circit - only the sensed voltage itself which is found across a 10k load
resistor. Is what I want possible? Any help appreciated.

Thanks,

Matt B



When the NPN is switched on, where will its conducting current come
from if you have no supply voltage? From the sensing circuit? What is
the end objective?

a shocky diode perhaps?
Click to expand...
 
Hysteresis is a great thing. As a matter of fact, I use a low pass filter
along with hystersis to eliminate most noise on the test stands for GM and
Chrysler. Hysteresis in this case means positive feedback. A NPN function
like an inverter. You will have to add another NPN in order to have positive
feedback. Use a feedback resistor that connects the 2nd NPN to the input
signal.
Just adding a RC filter to the input can make big trouble because the input
signal will now remain at the threshold voltage for a longer time.

Harold





"M B" <ffp944-a@yahoo.co.uk> wrote in message
news:39c69cF61j3bmU1@individual.net...
Hello,

I want to switch on an npn transistor when a sensed voltage exceeds a
certain threshold, say 4v, but I don't want it to switch off again until
the sensed voltage drops a bit below that, say 3v - because the signal
will not be clean.

I suppose I need a schmitt trigger, but I have no supply voltage for the
circit - only the sensed voltage itself which is found across a 10k load
resistor. Is what I want possible? Any help appreciated.

Thanks,

Matt B
Click to expand...
 
You must log in or register to reply here.

Welcome to EDABoard.com

Sponsor

Back
Top