Fastest AC Electronic Fuse

[D from BC]

This is probably an oldie question on here...

I'd like to break the AC "hot" if there's too much AC current.
This would be due to a short or ground fault.
Or....
Shunting the load to burn a fuse on the AC line would be ok too..

I'm afraid that a fuse maybe too slow to protect the electronics from
burn out..
Let's say the electronics I'm protecting cannot withstand 1.5x
overcurrent.

What's faster than a fuse on a AC line?

Details
Load: The load on the line is not a transformer and is not inductive.
Current draw: 2Amps
Line voltage: 240Vrms, 60Hz
Surge rate: guessing 45Amps in 1uS

Rule: 1/4watt current sense resistor.

The load circuit has a power bridge rectifier.
Maybe I could modify the bridge with transistors to interrupt the
surge current?
D from BC

 

[Jim Thompson]

This is probably an oldie question on here...

I'd like to break the AC "hot" if there's too much AC current.
This would be due to a short or ground fault.
Or....
Shunting the load to burn a fuse on the AC line would be ok too..

I'm afraid that a fuse maybe too slow to protect the electronics from
burn out..
Let's say the electronics I'm protecting cannot withstand 1.5x
overcurrent.

What's faster than a fuse on a AC line?

Details
Load: The load on the line is not a transformer and is not inductive.
Current draw: 2Amps
Line voltage: 240Vrms, 60Hz
Surge rate: guessing 45Amps in 1uS

Rule: 1/4watt current sense resistor.

The load circuit has a power bridge rectifier.
Maybe I could modify the bridge with transistors to interrupt the
surge current?
D from BC

Years ago I used an SCR to blow fuses (on the load side). I think it
was in the microsecond range. Warning, traces need to be fat,
otherwise you blow traces instead of fuses... I had ~100A circulating
momentarily to blow a 1 Amp fuse ;-)

...Jim Thompson

 

[John Larkin]

This is probably an oldie question on here...

I'd like to break the AC "hot" if there's too much AC current.
This would be due to a short or ground fault.
Or....
Shunting the load to burn a fuse on the AC line would be ok too..

I'm afraid that a fuse maybe too slow to protect the electronics from
burn out..
Let's say the electronics I'm protecting cannot withstand 1.5x
overcurrent.

What's faster than a fuse on a AC line?

Details
Load: The load on the line is not a transformer and is not inductive.
Current draw: 2Amps
Line voltage: 240Vrms, 60Hz
Surge rate: guessing 45Amps in 1uS

Rule: 1/4watt current sense resistor.

The load circuit has a power bridge rectifier.
Maybe I could modify the bridge with transistors to interrupt the
surge current?
D from BC

One of our customers recently asked us to do that. He needs,
basically, an SSR that also acts as a programmable circuit breaker
*and* a current limiter. He needs to simulate cable faults and switch
power circuits without exploding anything. We've been scribbling
circuits for days, and it's non-trivial. We just a few minutes ago
finished whiteboarding the design of a breadboard pcb so we can test
the dynamics.

John

 

[John Fields]

This is probably an oldie question on here...

I'd like to break the AC "hot" if there's too much AC current.
This would be due to a short or ground fault.
Or....
Shunting the load to burn a fuse on the AC line would be ok too..

I'm afraid that a fuse maybe too slow to protect the electronics from
burn out..
Let's say the electronics I'm protecting cannot withstand 1.5x
overcurrent.

What's faster than a fuse on a AC line?

Details
Load: The load on the line is not a transformer and is not inductive.
Current draw: 2Amps
Line voltage: 240Vrms, 60Hz
Surge rate: guessing 45Amps in 1uS

Rule: 1/4watt current sense resistor.

The load circuit has a power bridge rectifier.
Maybe I could modify the bridge with transistors to interrupt the
surge current?
D from BC

 

[Spehro Pefhany]

One of our customers recently asked us to do that. He needs,
basically, an SSR that also acts as a programmable circuit breaker
*and* a current limiter. He needs to simulate cable faults and switch
power circuits without exploding anything. We've been scribbling
circuits for days, and it's non-trivial. We just a few minutes ago
finished whiteboarding the design of a breadboard pcb so we can test
the dynamics.

John

There's some work already out there on IGBTs re shorts into hundreds
of volts. IIRC, it's bad to try to turn it off too fast (or too slow,
obviously), plus there are some other subtleties.


Best regards,
Spehro Pefhany

 

[Tim Williams]

I drew a circuit here
http://webpages.charter.net/dawill/Images/Electric_Fuse.gif
which someone built, and evidently works quite nicely. Maybe you could
change the MOSFET to two back-to-back so they'll handle AC. Which I also
had to draw for someone it seems:
http://webpages.charter.net/dawill/Images/Antiseries MOSFET Switch.gif

Alternately, you could put a single FET across the DC terminals of a FWB.

Obviously, the +12V supply needs to be floating. It could possibly be
provided by a resistor+capacitor and diode from the AC terminals of the
thing. Interfacing external control/feedback would probably be best done
with optoisolators...

Tim

 

[D from BC]

Years ago I used an SCR to blow fuses (on the load side). I think it
was in the microsecond range. Warning, traces need to be fat,
otherwise you blow traces instead of fuses... I had ~100A circulating
momentarily to blow a 1 Amp fuse ;-)

...Jim Thompson

Wow!...peaking to 100A before blow out.....
Scary....

I guess it takes time for the fuse element temp to rise to the
melting/vaporize point.
D from BC

 

[D from BC]

One of our customers recently asked us to do that. He needs,
basically, an SSR that also acts as a programmable circuit breaker
*and* a current limiter. He needs to simulate cable faults and switch
power circuits without exploding anything. We've been scribbling
circuits for days, and it's non-trivial. We just a few minutes ago
finished whiteboarding the design of a breadboard pcb so we can test
the dynamics.

John

To break a DC circuit, a pass transistor can be used.. Easy...

But breaking 240V AC...huhhh..

I might look at using a little transformer to sense a current surge
and have that fire a shunt circuit until a fuse blows..

D from BC

 

[D from BC]

I might have to blow up some stuff to find out.

I'm just interested in rough ideas for various levels of cut-out
speed..
Doesn't matter what cost, complexity or parts count.

At the moment, the faster ...the better..

D from BC

 

[D from BC]

I drew a circuit here
http://webpages.charter.net/dawill/Images/Electric_Fuse.gif
which someone built, and evidently works quite nicely. Maybe you could
change the MOSFET to two back-to-back so they'll handle AC. Which I also
had to draw for someone it seems:
http://webpages.charter.net/dawill/Images/Antiseries MOSFET Switch.gif

Alternately, you could put a single FET across the DC terminals of a FWB.

Obviously, the +12V supply needs to be floating. It could possibly be
provided by a resistor+capacitor and diode from the AC terminals of the
thing. Interfacing external control/feedback would probably be best done
with optoisolators...

Tim

mmm...makes sense to see transistors to shut down the mosfet..
Op amp current sensing might be considered generally slower..
The cct looks like it has a ~6A trip level....
It's easy when a sense resistor has enough V to bias a transistor.

Using a floating circuit to control a push pull pair of mosfets
looking interesting..

Thanks..
D from BC

 

[D from BC]

Years ago I used an SCR to blow fuses (on the load side). I think it
was in the microsecond range. Warning, traces need to be fat,
otherwise you blow traces instead of fuses... I had ~100A circulating
momentarily to blow a 1 Amp fuse ;-)

...Jim Thompson

I'm wondering what's more explosive...
Slowly sneaking up the current through the fuse until it blows...
Or.....
suddenly connecting the fuse across a 300V charged 100uF low ESR
capacitor?

Should I worry about fuse shrapnel and BANG! when it comes to fuse
selection?
D from BC

 

[Jim Thompson]

I'm wondering what's more explosive...
Slowly sneaking up the current through the fuse until it blows...
Or.....
suddenly connecting the fuse across a 300V charged 100uF low ESR
capacitor?

Should I worry about fuse shrapnel and BANG! when it comes to fuse
selection?
D from BC

This was a conventional fast-blow fuse. It flashed like a flash-bulb
(but not as much light) and made a tiny "click" ;-)

The filament was completely plated out onto the glass.

After we cracked a few, we went to the ceramic tube type.

...Jim Thompson

 

[John Popelish]

I'm wondering what's more explosive...
Slowly sneaking up the current through the fuse until it blows...
Or.....
suddenly connecting the fuse across a 300V charged 100uF low ESR
capacitor?

Definitely the second. In the first case, the alloy just
quietly melts. In the second, it vaporizes.

Should I worry about fuse shrapnel and BANG! when it comes to fuse
selection?

If you exceed the fault current rating of the fuse, all bets
for containment are off.

 

[Fred Bartoli]

D from BC a écrit :

To break a DC circuit, a pass transistor can be used.. Easy...

But breaking 240V AC...huhhh..

I might look at using a little transformer to sense a current surge
and have that fire a shunt circuit until a fuse blows..

How about an AC current limiter?


.--------------------------------.
| |
| |
| .---+---. |
| | | | |
.---+ | .-. .-. |
| | | | | | | |
| +-|| .----+ | | | | |
- ->|| | | '-' '-' -
^ +-||--' >| | | ^
| | |-+ +-|<-+---. |
o-----+ +--. /| | | | | +-------o
| .-. | | | | | | |
- | | | ___ | |/ --- --- z -
^ | | '-|___|-+-| --- --- A ^
| '-' |> | | | |
AC in | | | | | | | AC out
'---+--------------+---+----|---+----'
.-.
| |
| |
'-'
|
o---------------------------------+----------------o

 

[D from BC]

D from BC a écrit :

How about an AC current limiter?


.--------------------------------.
| |
| |
| .---+---. |
| | | | |
.---+ | .-. .-. |
| | | | | | | |
| +-|| .----+ | | | | |
- ->|| | | '-' '-' -
^ +-||--' >| | | ^
| | |-+ +-|<-+---. |
o-----+ +--. /| | | | | +-------o
| .-. | | | | | | |
- | | | ___ | |/ --- --- z -
^ | | '-|___|-+-| --- --- A ^
| '-' |> | | | |
AC in | | | | | | | AC out
'---+--------------+---+----|---+----'
.-.
| |
| |
'-'
|
o---------------------------------+----------------o

Arrrrghh...
At first I couldn't figure this cct out..

The only way for current to pass is if the N Ch mos turns on.
Otherwise the diodes block..
I see the bridge rectifier in this circuit..
Cool...

I see the negative feedback loop from the Rsense through the NPN-PNP
compound and back to the MOS that limits the AC current..

Is the AZ a zener?

The circuit looks fast..

The only thing I don't like about this cct is that the current sensing
resistor will be dissipating a nominal 1.2Watts in my app..
0.6V*2Amps= 1.2Watts..
0.6V~=Vbe
To use a smaller current sense resistor, I could use an op amp, but
I'll be trading off speed...
However...I might be able to come up with a transistor arrangement
that uses a smaller (cooler) sense resistor..

Maybe I might try an SCR to shut down the mosfet...
Or CMOS logic...
It wouldn't be a limiter anymore but a breaker..

Thanks...

Another circuit I'm looking at is
http://www.google.com/patents?id=CK...om=4&dq=triac+over+current+protection#PPP1,M1
Triac blowing line fuse.
I think this also has a hot current sense resistor (If I=2amps and
Itrip = 3Amps).
D from BC

 

[Fred Bartoli]

D from BC a écrit :

Arrrrghh...
At first I couldn't figure this cct out..

The only way for current to pass is if the N Ch mos turns on.
Otherwise the diodes block..
I see the bridge rectifier in this circuit..
Cool...

I see the negative feedback loop from the Rsense through the NPN-PNP
compound and back to the MOS that limits the AC current..

The positive FB loop between the NPN/PNP bjts make the 'limiter' trip.
With the right time constants, the *small* bypass cap voltage gets down
to the point where the 'SCR' self resets. This obviously has to happen
within one line period (before injecting a new current pulse through the
series RC).

Is the AZ a zener?

Yep. Limits the gate voltage and provides a path for the RC branch
'negative half period' current.

The circuit looks fast..

Should be.

The only thing I don't like about this cct is that the current sensing
resistor will be dissipating a nominal 1.2Watts in my app..
0.6V*2Amps= 1.2Watts..
0.6V~=Vbe
To use a smaller current sense resistor, I could use an op amp, but
I'll be trading off speed...
However...I might be able to come up with a transistor arrangement
that uses a smaller (cooler) sense resistor..

Can be fixed with just one more resistor.

Maybe I might try an SCR to shut down the mosfet...
Or CMOS logic...
It wouldn't be a limiter anymore but a breaker..

It is already. Well a periodic, self resetting one.

 

[Ian]

"Fred Bartoli"

D from BC a écrit :

The positive FB loop between the NPN/PNP bjts make the 'limiter' trip.
With the right time constants, the *small* bypass cap voltage gets down to
the point where the 'SCR' self resets. This obviously has to happen within
one line period (before injecting a new current pulse through the series
RC).


Yep. Limits the gate voltage and provides a path for the RC branch
'negative half period' current.


Should be.


Can be fixed with just one more resistor.



It is already. Well a periodic, self resetting one.

The power dissipated in the resistor is less than that, and as Fred says
can be fixed easily (it isn't steady state, more like a sawtooth).

I used something very similar as an inrush limiter on a SMPS a many
years ago, and got bitten by ** changing the fab that supplied the
Mosfet just after we went into production. I forget the details, but it
was to do with wimpy metallisation to the Mosfet gate - to turn the
Mosfet off fast you have to pull the gate charge back out, which
takes current.

Regards
Ian

 

[Ian]

I might have to blow up some stuff to find out.

I'm just interested in rough ideas for various levels of cut-out
speed..
Doesn't matter what cost, complexity or parts count.

At the moment, the faster ...the better..

D from BC

The fastest fuse in the world is an expensive microwave transistor.
Works a treat to protect line fuses.

Regards
Ian

 

[Fred Bartoli]

Ian a écrit :

The fastest fuse in the world is an expensive microwave transistor.
Works a treat to protect line fuses.

But you'll have to call your power company so that they speed up their
generators quite a bit.

 

[Spehro Pefhany]

I'm wondering what's more explosive...
Slowly sneaking up the current through the fuse until it blows...
Or.....
suddenly connecting the fuse across a 300V charged 100uF low ESR
capacitor?

Should I worry about fuse shrapnel and BANG! when it comes to fuse
selection?
D from BC

I did tests, for example, on 5A 250V rated fuses. The 20mm glass
variety would literally explode (a VERY loud bang) sending shrapnel
everywhere, just on 240VAC/60Hz with an industrial source capable of
enough source current.


Best regards,
Spehro Pefhany