My mosfet regrets...

[Jamie]

Hi,

I have a problem with a circuit I'm developing...I've put a schematic up
here http://www.student-pcs.co.uk/circuit.tif (sorry about the drawing...)
a datasheet for the 4017 is availiable here
http://eu.st.com/stonline/books/pdf/docs/1944.pdf
and for the mosfet here http://us.st.com/stonline/books/pdf/docs/2950.pdf

The two sections indicated by the dotted lines are in seperate enclosures
and are connect by a long cable and run off seperate power supplys. (There
are actually 10 of the sections containing LEDs, one for each of the 4017's
outputs).

Now for the problem, the LEDs if I get a wire, connect it to the gate and
touch the positive supply voltage, the LEDs will turn on and stay on till I
touch the wire to ground. As indictated on the diagram, the measured supply
voltage is 4.8V and the ground is obviously 0V. So I'm figuring that when I
connect up my 4017, what with the HI voltage being 5.1V, every thing will be
just great and LEDs will be going on and off all over the place, but, alas,
it is not so...

So, what on earth is going on? On thing I did think of was to change the LED
section to look like this http://www.student-pcs.co.uk/alternate.tif with a
resistor from the gate to ground, but I'm really just guessing...

Any help you could give me on this'd be, like, totally appreicated

Jamie

 

[Wolfgang Mahringer]

Hi Jamie,

Jamie wrote:

I have a problem with a circuit I'm developing...I've put a schematic up
here http://www.student-pcs.co.uk/circuit.tif (sorry about the drawing...)
...
The two sections indicated by the dotted lines are in seperate enclosures
and are connect by a long cable and run off seperate power supplys. (There
are actually 10 of the sections containing LEDs, one for each of the 4017's
outputs).

Now for the problem, the LEDs if I get a wire, connect it to the gate and
touch the positive supply voltage, the LEDs will turn on and stay on till I
touch the wire to ground.

Since the Gate-to-Source-Resistance of a MOSFET is soooo big, you put
some charge from you body to the gate and turned it on. If you touch GND
with the gate wire, you discharge the gate and the Leds are turning off.

On thing I did think of was to change the LED
section to look like this http://www.student-pcs.co.uk/alternate.tif with a
resistor from the gate to ground, but I'm really just guessing...

Good Idea. I would have taken 10kohm, but ok....

HTH
Wolf

 

[Jamie]

Hi Jamie,

Jamie wrote:
Now for the problem, the LEDs if I get a wire, connect it to the gate
and
touch the positive supply voltage, the LEDs will turn on and stay on
till I
touch the wire to ground.

Since the Gate-to-Source-Resistance of a MOSFET is soooo big, you put
some charge from you body to the gate and turned it on. If you touch GND
with the gate wire, you discharge the gate and the Leds are turning off.

The wire was insulated...what I meant was that I touched the wire to the
positive supply.

On thing I did think of was to change the LED
section to look like this http://www.student-pcs.co.uk/alternate.tif
with a
resistor from the gate to ground, but I'm really just guessing...

Good Idea. I would have taken 10kohm, but ok....

You reckon this will work? That was a good guess! I'm just off to put some
money on a horse..lol

Jamie

 

[Wolfgang Mahringer]

Hi Jamie,

Jamie wrote:

The wire was insulated...what I meant was that I touched the wire to the
positive supply.

In that case it makes no difference. For transfering charge there is no
need of touching a wire directly!

You reckon this will work? That was a good guess!

Yep, I don't see a problem here for a private project.

Have fun,
Wolf

 

[Peter Bennett]

On Fri, 12 Sep 2003 14:57:40 +0100, "Jamie"
<jizzproductions@hotmail.com> wrote:

Hi,

I have a problem with a circuit I'm developing...I've put a schematic up
here http://www.student-pcs.co.uk/circuit.tif (sorry about the drawing...)
a datasheet for the 4017 is availiable here
http://eu.st.com/stonline/books/pdf/docs/1944.pdf
and for the mosfet here http://us.st.com/stonline/books/pdf/docs/2950.pdf

The two sections indicated by the dotted lines are in seperate enclosures
and are connect by a long cable and run off seperate power supplys. (There
are actually 10 of the sections containing LEDs, one for each of the 4017's
outputs).

Do you have a ground connection between the two parts?










--
Peter Bennett VE7CEI
GPS and NMEA info and programs: http://vancouver-webpages.com/peter/index.html
Newsgroup new user info: http://vancouver-webpages.com/nnq

 

[Jamie]

no..should I?

Jamie
"Peter Bennett" <peterbb@interchange.ubc.ca> wrote in message
news:nru3mvk6ob3lnhv65b2nohl9kvkajkpgh2@4ax.com...

On Fri, 12 Sep 2003 14:57:40 +0100, "Jamie"
jizzproductions@hotmail.com> wrote:

Hi,

I have a problem with a circuit I'm developing...I've put a schematic up
here http://www.student-pcs.co.uk/circuit.tif (sorry about the
drawing...)
a datasheet for the 4017 is availiable here
http://eu.st.com/stonline/books/pdf/docs/1944.pdf
and for the mosfet here http://us.st.com/stonline/books/pdf/docs/2950.pdf

The two sections indicated by the dotted lines are in seperate enclosures
and are connect by a long cable and run off seperate power supplys.
(There
are actually 10 of the sections containing LEDs, one for each of the
4017's
outputs).

Do you have a ground connection between the two parts?










--
Peter Bennett VE7CEI
GPS and NMEA info and programs:
http://vancouver-webpages.com/peter/index.html
Newsgroup new user info: http://vancouver-webpages.com/nnq

 

[Wolfgang Mahringer]

Hi Jamie,

Jamie wrote:

no..should I?

Now that was a joke, wasn't it?

greetz
Wolf

 

[Walter Harley]

"Jamie" <jizzproductions@hotmail.com> wrote in message
news:bjt0mo$knj$1@newsg1.svr.pol.co.uk...

Do you have a ground connection between the two parts?
no..should I?

Yes.

These things are called electronic circuits, right? That's "circuit", as in
"circle", as in "the end connects back to the start". Right now, you don't
have a control circuit, you've got an LED galvanometer

Actually, to get technical, you do have a circuit; but the return path is
extremely high impedance (it involves insulators rather than conductors) and
as a result your system is very susceptible to external voltage fields. To
make your system behave reliably you need the impedances of all intended
current paths to be much less than the impedances to any external,
unintended fields. The way you do that, in this case, is with a ground
wire.

Another way of thinking of it is this: you're trying to control the MOSFET
with a voltage. But there's no such thing as a "voltage" in absolute terms;
voltage is a relative measurement. A voltage exists between two points, not
just at one point. YOu have to give the MOSFET and the controller some
shared ground reference, and the way that you "share" those grounds is by
connecting them electrically. Just drawing them with the same symbol
doesn't mean anything; electrons can't read schematics.

See also the responses on sci.electronics.design to this post, because there
are also some other problems that need to be fixed for this to work
properly.

 

[Jamie]

Most helpful, it totally makes sense to me now, I mean, everybody's sayin'
"connect your grounds" and I'm going right, lets stick a wire in here then!
But I know why to stick a wire in there...probably should've know why, but
hey, guess I should've gone to a few more lectures at uni last
year...(promises to go to more in 2nd year). Anyway, thanks every one for
all your most helpful help,

bless you all (and your little cotton socks)
Jamie

"Walter Harley" <walterh@cafewalter.com> wrote in message
news:bjt203$i22$0@216.39.172.65...

"Jamie" <jizzproductions@hotmail.com> wrote in message
news:bjt0mo$knj$1@newsg1.svr.pol.co.uk...
Do you have a ground connection between the two parts?
no..should I?

Yes.

These things are called electronic circuits, right? That's "circuit", as
in
"circle", as in "the end connects back to the start". Right now, you
don't
have a control circuit, you've got an LED galvanometer

Actually, to get technical, you do have a circuit; but the return path is
extremely high impedance (it involves insulators rather than conductors)
and
as a result your system is very susceptible to external voltage fields.
To
make your system behave reliably you need the impedances of all intended
current paths to be much less than the impedances to any external,
unintended fields. The way you do that, in this case, is with a ground
wire.

Another way of thinking of it is this: you're trying to control the MOSFET
with a voltage. But there's no such thing as a "voltage" in absolute
terms;
voltage is a relative measurement. A voltage exists between two points,
not
just at one point. YOu have to give the MOSFET and the controller some
shared ground reference, and the way that you "share" those grounds is by
connecting them electrically. Just drawing them with the same symbol
doesn't mean anything; electrons can't read schematics.

See also the responses on sci.electronics.design to this post, because
there
are also some other problems that need to be fixed for this to work
properly.

 

[tim kettring]

Cmos is so nice for low power consumption , but it is so easy to trash !!!

tim CET-MD238

"Jamie" <jizzproductions@hotmail.com> wrote in message news:<bjt3v2$brq$1@newsg3.svr.pol.co.uk>...

Most helpful, it totally makes sense to me now, I mean, everybody's sayin'
"connect your grounds" and I'm going right, lets stick a wire in here then!
But I know why to stick a wire in there...probably should've know why, but
hey, guess I should've gone to a few more lectures at uni last
year...(promises to go to more in 2nd year). Anyway, thanks every one for
all your most helpful help,

bless you all (and your little cotton socks)
Jamie

"Walter Harley" <walterh@cafewalter.com> wrote in message
news:bjt203$i22$0@216.39.172.65...
"Jamie" <jizzproductions@hotmail.com> wrote in message
news:bjt0mo$knj$1@newsg1.svr.pol.co.uk...
Do you have a ground connection between the two parts?
no..should I?

Yes.

These things are called electronic circuits, right? That's "circuit", as
in
"circle", as in "the end connects back to the start". Right now, you
don't
have a control circuit, you've got an LED galvanometer

Actually, to get technical, you do have a circuit; but the return path is
extremely high impedance (it involves insulators rather than conductors)
and
as a result your system is very susceptible to external voltage fields.
To
make your system behave reliably you need the impedances of all intended
current paths to be much less than the impedances to any external,
unintended fields. The way you do that, in this case, is with a ground
wire.

Another way of thinking of it is this: you're trying to control the MOSFET
with a voltage. But there's no such thing as a "voltage" in absolute
terms;
voltage is a relative measurement. A voltage exists between two points,
not
just at one point. YOu have to give the MOSFET and the controller some
shared ground reference, and the way that you "share" those grounds is by
connecting them electrically. Just drawing them with the same symbol
doesn't mean anything; electrons can't read schematics.

See also the responses on sci.electronics.design to this post, because
there
are also some other problems that need to be fixed for this to work
properly.