Build an in-circuit fuse blown detector?

A

Arch-lab

Guest
Anyone seen a circuit that detects if a fuse is blown using simplest of
analog parts? AND considers the following.

Details:
Circuit: 12V DC input, fuse, diode(A to K), Vfused DC

Catches:
1. If the fuse blows I want a LED to light... catch is that you can't
just put it over the fuse because I don't want to rely on the output
being shorted. I.e. if it isn't shorted to ground then the led still
wont light.
2. Second catch , I don't want to waste current. The whole point of this
is so that we don't have an LED ON all the time while the fuse is good.
Thus to save current I want the opposite to happen. SO, if you use some
resistor divider to bias a LED setup you will have to chose values that
can pass enough current for when the LED turns on... which means you
will be wasting current and may as well just have a LED ON all the time.
 
"Arch-lab" <lab@us.com> wrote in message
news:OVgid.68487$df2.59169@edtnps89...
Anyone seen a circuit that detects if a fuse is blown using simplest of
analog parts? AND considers the following.

Details:
Circuit: 12V DC input, fuse, diode(A to K), Vfused DC

Catches:
1. If the fuse blows I want a LED to light... catch is that you can't
just put it over the fuse because I don't want to rely on the output
being shorted. I.e. if it isn't shorted to ground then the led still
wont light.
2. Second catch , I don't want to waste current. The whole point of this
is so that we don't have an LED ON all the time while the fuse is good.
Thus to save current I want the opposite to happen. SO, if you use some
resistor divider to bias a LED setup you will have to chose values that
can pass enough current for when the LED turns on... which means you
will be wasting current and may as well just have a LED ON all the time.
Click to expand...
Take a look at http://www.fncwired.com/FusedLED/
Tell me what you think?
Brian
 
Brian wrote:
"Arch-lab" <lab@us.com> wrote in message
news:OVgid.68487$df2.59169@edtnps89...


Anyone seen a circuit that detects if a fuse is blown using simplest of
analog parts? AND considers the following.

Details:
Circuit: 12V DC input, fuse, diode(A to K), Vfused DC

Catches:
1. If the fuse blows I want a LED to light... catch is that you can't
just put it over the fuse because I don't want to rely on the output
being shorted. I.e. if it isn't shorted to ground then the led still
wont light.
2. Second catch , I don't want to waste current. The whole point of this
is so that we don't have an LED ON all the time while the fuse is good.
Thus to save current I want the opposite to happen. SO, if you use some
resistor divider to bias a LED setup you will have to chose values that
can pass enough current for when the LED turns on... which means you
will be wasting current and may as well just have a LED ON all the time.




Take a look at http://www.fncwired.com/FusedLED/
Tell me what you think?
Brian
Click to expand...
not very good. LED intensity probably varies with Beta, as you cant (by
definition!) make the pnp saturate - Vce cant fall below Vbe.


put the LED at the top, then the pnp, then the LED current setting
resistor. it looks more like Spehro's nifty cct that way, and its easy
to ensure the npn is off when you want it to be, but turns on properly
when the fuse (or diode) blows:

+12V unfused
|
|
|
| LED (high efficiency, no sense in wasting 20mA)
+------->|--------+
| |
,-. |
||| |
||| Fuse |
'-' | E
| ___ |<
+--|___|--+-----| pnp Beta > 200 for 2.4mA in LED
| 10k | |\
Diode V .-. | C
- | | .-.
| | | 1M0 | | Rled = (12 - 0.7 - 1.7)/Iled
'-' | | say 3k9
+12V fused | '-'
| |
=== ===
GND GND

And it uses the exact same number of bits as they shonky circuit...

Note that in this case their is a "sneak" path through the LED, b-e
junction, 10k and diode into the load, but only base current flows
through it (Spehro's circuit likewise pumps a little current into the
load this way, although via the 1N5245). The 10k could be bigger, too - 47k.

In either case the LED will NOT light if the output diode fails but the
fuse does not - check your I^2t ratings of fuse & diode, and make sure
the diode's is bigger (if they dont give you I^2t then figure it out
from Ifsm). I am sure many people have come across semiconductors which
have given their (oft expensive) lives to protect 10c fuses, I sure as
hell have (actually, its $600 transistors protecting $100 fuses, but the
principle is the same :)

Cheers
Terry
 
"Brian" <bellis350@comcast.net> wrote in message
news:bLidnWM9dv10XxTcRVn-ig@comcast.com...
"Terry Given" <my_name@ieee.org> wrote in message
news:p6jid.3114$op3.128770@news.xtra.co.nz...
Brian wrote:
"Arch-lab" <lab@us.com> wrote in message
news:OVgid.68487$df2.59169@edtnps89...


Anyone seen a circuit that detects if a fuse is blown using
simplest of
analog parts? AND considers the following.

Details:
Circuit: 12V DC input, fuse, diode(A to K), Vfused DC

Catches:
1. If the fuse blows I want a LED to light... catch is that you
can't
just put it over the fuse because I don't want to rely on the
output
being shorted. I.e. if it isn't shorted to ground then the led
still
wont light.
2. Second catch , I don't want to waste current. The whole point
of
this
is so that we don't have an LED ON all the time while the fuse
is
good.
Thus to save current I want the opposite to happen. SO, if you
use
some
resistor divider to bias a LED setup you will have to chose
values
that
can pass enough current for when the LED turns on... which means
you
will be wasting current and may as well just have a LED ON all
the
time.




Take a look at http://www.fncwired.com/FusedLED/
Tell me what you think?
Brian

not very good. LED intensity probably varies with Beta, as you cant
(by
definition!) make the pnp saturate - Vce cant fall below Vbe.


put the LED at the top, then the pnp, then the LED current setting
resistor. it looks more like Spehro's nifty cct that way, and its
easy
to ensure the npn is off when you want it to be, but turns on
properly
when the fuse (or diode) blows:

+12V unfused
|
|
|
| LED (high efficiency, no sense in wasting 20mA)
+------->|--------+
| |
,-. |
||| |
||| Fuse |
'-' | E
| ___ |
+--|___|--+-----| pnp Beta > 200 for 2.4mA in LED
| 10k | |\
Diode V .-. | C
- | | .-.
| | | 1M0 | | Rled = (12 - 0.7 - 1.7)/Iled
'-' | | say 3k9
+12V fused | '-'
| |
=== ===
GND GND

And it uses the exact same number of bits as they shonky circuit...

Note that in this case their is a "sneak" path through the LED, b-e
junction, 10k and diode into the load, but only base current flows
through it (Spehro's circuit likewise pumps a little current into the
load this way, although via the 1N5245). The 10k could be bigger,
too -
47k.

In either case the LED will NOT light if the output diode fails but
the
fuse does not - check your I^2t ratings of fuse & diode, and make
sure
the diode's is bigger (if they dont give you I^2t then figure it out
from Ifsm). I am sure many people have come across semiconductors
which
have given their (oft expensive) lives to protect 10c fuses, I sure
as
hell have (actually, its $600 transistors protecting $100 fuses, but
the
principle is the same :)

Cheers
Terry

After thinking about, I knew there was a better way to do it. The way
it
is
drawn, it will probably work okay but it could be better. I will redraw
it.
Thanks,
Brian
Click to expand...
Okay, I changed it. How does this look. http://www.fncwired.com/FusedLED/

Brian
 
Terry Given wrote:


+12V unfused
|
|
|
| LED (high efficiency, no sense in wasting 20mA)
+------->|--------+
| |
,-. |
||| |
||| Fuse |
'-' | E
| ___ |
+--|___|--+-----| pnp Beta > 200 for 2.4mA in LED
| 10k | |\
Diode V .-. | C
- | | .-.
| | | 1M0 | | Rled = (12 - 0.7 - 1.7)/Iled
'-' | | say 3k9
+12V fused | '-'
| |
=== ===
GND GND
Click to expand...
Note that this circuit could provide enough current to operate CMOS
circuitry via the be junction and the 10k. I prefer a circuit like
Spehero's with a FET to avoid this problem.

And don't forget- the LED can just as well be an optocoupler, which
allows easy connection to a status monitoring system.

Paul Burke
 
Problem is this violates the conditions I mentioned in the original
post.

If your load current is at least 5 - 10 ma, putting the LED in series
with a
resistor across the fuse should still work. Easy enough to try -
remove the
fuse.

Tam
Click to expand...
 
"Arch-lab" <lab@us.com> wrote in message
news:iGzid.77338$df2.27094@edtnps89...
Problem is this violates the conditions I mentioned in the original
post.

Not if your load normally draws enough current. A simple addition is a P
Click to expand...
channel FET. Connect the source to the unfused side, the gate to the load
through a high value resistor, and the LED from drain to ground in series
with a resistor.

Tam
If your load current is at least 5 - 10 ma, putting the LED in series
with a
resistor across the fuse should still work. Easy enough to try -
remove the
fuse.

Tam
Click to expand...
 
Yeah... that won't work because it violtes the constraints I listed...
thanks.


If your load current is at least 5 - 10 ma, putting the LED in series
with a
resistor across the fuse should still work. Easy enough to try -
remove the
fuse.

Tam
Click to expand...
 
Typo there, I meant it violates the constraints I listed... i.e. it will
not work...


Yeah... that won't work because it violtes the constraints I listed...
thanks.


If your load current is at least 5 - 10 ma, putting the LED in
series
with a
resistor across the fuse should still work. Easy enough to try -
remove the
fuse.

Tam
Click to expand...
 
"Arch-lab" <lab@us.com> wrote in message
news:adQid.82247$df2.37656@edtnps89...
Yeah... that won't work because it violtes the constraints I listed...
thanks.

You commented twice on the previous posting. The PMOS can be made to work
Click to expand...
even if the load is open if you connect a second resistor of the same value
as the gate resistor from load to ground.

Tam
If your load current is at least 5 - 10 ma, putting the LED in series
with a
resistor across the fuse should still work. Easy enough to try -
remove the
fuse.

Tam
Click to expand...
 
"Arch-lab" <lab@us.com> schreef in bericht
news:UvQid.82273$df2.4307@edtnps89...
Typo there, I meant it violates the constraints I listed... i.e. it will
not work...


Yeah... that won't work because it violtes the constraints I listed...
thanks.


If your load current is at least 5 - 10 ma, putting the LED in
series
with a
resistor across the fuse should still work. Easy enough to try -
remove the
fuse.

Tam
Click to expand...
Think the circuit below fulfils all your needs. Less then 2uA quiscent
current, no current into the load when the fuse has been blown and a minimal
part count. The BC516 is a PNP darlington. You can try two PNP transistors
as well and you can raise the 1k resistor to lower the LED current.

_
----------+-----o_/ \o-----+------------
| |
V V Si
- LED -
| |
| |
| |
|---+ |
Click to expand...
/| | |
| >| |
|BC516 |--------+
| /| |
| | |
+-----+ |
| |
.-. .-.
| | | |
| |1k | |10M
'-' '-'
| |
---------+----------------+------------
(created by AACircuit v1.28 beta 10/06/04 www.tech-chat.de)

petrus bitbyter




---
Outgoing mail is certified Virus Free.
Checked by AVG anti-virus system (http://www.grisoft.com).
Version: 6.0.786 / Virus Database: 532 - Release Date: 29-10-2004
 
On Thu, 04 Nov 2004 02:58:54 GMT, the renowned "Arch-lab" <lab@us.com>
wrote:

Anyone seen a circuit that detects if a fuse is blown using simplest of
analog parts? AND considers the following.

Details:
Circuit: 12V DC input, fuse, diode(A to K), Vfused DC

Catches:
1. If the fuse blows I want a LED to light... catch is that you can't
just put it over the fuse because I don't want to rely on the output
being shorted. I.e. if it isn't shorted to ground then the led still
wont light.
2. Second catch , I don't want to waste current. The whole point of this
is so that we don't have an LED ON all the time while the fuse is good.
Thus to save current I want the opposite to happen. SO, if you use some
resistor divider to bias a LED setup you will have to chose values that
can pass enough current for when the LED turns on... which means you
will be wasting current and may as well just have a LED ON all the time.
Click to expand...
Here's one way- it only uses 12uA or so Iq.

+12V unfused
|
|
|
| 1N4148
+------->|---------+-------+
| | |
,-. 1N5245 | / |
||| /--- |
||| Fuse / \ |
'-' | | S
| ___ | ||-+
+------+---|___|---+----||-> BS250P
| | 10K ||-+
V .-. | D
- | | .-.
| | | 1M0 | | 1K0
'-' | |
+12V fused | '-'
| |
=== | /
GND V / LED
- /
|
|
===
GND




Best regards,
Spehro Pefhany
--
"it's the network..." "The Journey is the reward"
speff@interlog.com Info for manufacturers: http://www.trexon.com
Embedded software/hardware/analog Info for designers: http://www.speff.com
 
You must log in or register to reply here.

Welcome to EDABoard.com

Sponsor

Back
Top