Re: discrete comparator
- From: "kell" <kellrobinson@xxxxxxxxxxxx>
- Date: 1 Feb 2006 23:13:34 -0800
Winfield Hill wrote:
> kell wrote...
> >
> > A while ago I posted here with a question about current sensing
> > in a circuit to limit the output of an antique generator. Here's
> > an idea for a comparator with hysteresis to cut off field current
> > when generator output exceeds 10 amps. 36R is an estimate, and
> > the value of the 470K hysteresis resistor is just a guess.
> >
> > I haven't built this yet. Also, I wonder if it would need a cap
> > from the collector to base of the PNP transistor.
> >
> > DC generator output
> > |
> > V
> > -
> > |
> > +--0R1--+----O B+
> > | '-->|-->|---,
> > \e |
> > PNP |--+--36R--+-->|--+
> > /c | | 10K |
> > | | \/\/\/\/
> > | | ^
> > | 470k |
> > | | gnd
> > | | field
> > | | O
> > | | |
> > | | |--
> > | +--|| N channel mos
> > | | |--
> > | c/ |
> > +--| NPN |
> > e\ |
> > | |
> > '--+--+- gnd
> >
> > A voltage divider is connected to the base of the NPN transistor
> > so it serves both current limiting and voltage regulation.
>
> I compressed your circuit slightly. Your drawing is hard to
> understand, what's B+, which is in 0R1's current path, and
> what's "field 0", which is driven by the powerful MOSFET drain?
> Are we to assume a magnetic-field coil goes between these two?
>
> Your circuit is rather tortured. The hysteresis part may be OK,
> but 0R1 with two diodes and pot plus 36R are troublesome. When
> you get the right circuit there won't be any guessing. One thing
> to keep in mind is diode voltage drops don't equal transistor
> drops (although they may match diode-connected transistor drops).
>
> Also, you mentioned a voltage divider "connected to the base of
> the NPN transistor," but we don't see it in your drawing. Plus,
> it's not clear how a voltage divider on the NPN could be helpful.
> A resistor across the NPN base-emitter would establish a current
> in the PNP that might be helpful in setting the PNP Vbe voltage
> at some threshold current. Perhaps that's what you had in mind.
>
>
> --
> Thanks,
> - Win
I'll paste a diagram and some explanatory text for the voltage
regulator the limiter is meant to serve from my "voltage regulator"
post of Jan 21.
+-----------------+------+-----O +Batt
| | |
_/ | |
7.5v /^ 1K - blocking
| | ^ diode
| | |
| | |
| | +-----O Armature
| | |
| | |
| | - freewheel
| | ^ diode
| | |
V +---100K---------+-----O Field
- | | |
| | | |
: | | |
1N4148 : | | |
X10 : | | _|
| | .01 | |
V +----||---+---|| N channel
- | | |_ mosfet
| | | |
| | c/ |
+---R---+-------| NPN |
| e\ |
Rtrim | |
| | |
+-----------------+------+------O Ground
The 100 K resistor and the 0.1 uF cap are for hysteresis, to keep heat
down in the mosfet; I'm putting this circuit into a small, flat
enclosure without fins so it can be installed in a motorcycle's battery
compartment. The string of 1N4148 diodes are there so the regulator
will have a temperature coefficient matching the charging
characteristic of the battery. Another reason to keep heat
down in the circuit. Similarly, I used a schottky for the blocking
diode, which has to conduct all the charging current.
Generator configuration:
One of the armature brushes is grounded. The other brush is wired to
the A terminal on the outside of the generator and is also connected to
one end of the field coils. The other end of the field is connected to
the F terminal on the outside of the generator. It's called a floating
field or "type A" arrangement, common on the old GM vehicles and
Harleys.
* * *
Okay, first of all, I made a boo-boo in the diagram -- the current
sense resistor is 0.01 ohms, not 0.1 ohms.
In answer to your questions, B+ is the battery's positive terminal.
The O next to the word field simply represents the connection point on
the circuit board for the wire going to the field terminal of the
generator.
The NPN transistor and mosfet in the latest post is the very same NPN
transistor and mosfet in the "voltage regulator" diagram. I didn't
want to redraw the whole thing this time, so I just drew the two
circuit elements the limiter mainly affects.
Also, what I lazily referred to as a "divider" in my latest post isn't
strictly speaking a divider.
But that's neither here nor there.
The 10K resistor is a pot with the wiper connected to ground.
With the wiper at the midpoint, the 36 ohm resistor would have about
0.1 volt across it. So with the nominal ten amps through the current
sense resistor, the pnp transistor
will turn on. The pot is for adjusting it exactly.
Hysteresis so the mosfet is always on or off and doesn't generate heat.
I'll use diode-connected transistors instead of diodes.
Well, that's enough for now.
Kell
.
- References:
- discrete comparator
- From: kell
- Re: discrete comparator
- From: Winfield Hill
- discrete comparator
- Prev by Date: Re: Labelling plastic boxes
- Next by Date: Re: simple current source for LED and 3 cells?
- Previous by thread: Re: discrete comparator
- Next by thread: Question for baking chips
- Index(es):
Relevant Pages
|
Loading
