Mosfet switch used to protect a load

Status
Not open for further replies.

Thread Starter

newbieateverything

Joined Feb 25, 2013
95
How does your circuit detect the current?

V9 in series with U1A which makes no sense. If you want V9 to power the op amp it needs to have the negative side grounded and the positive side going to pin 4. There should be no connection to V3.

The LM324 is a slow op amp and will not rapidly turn-on the MOSFET in response to a signal. Better to use a comparator such as the LM339 which has a much faster response.
thanks, i dont have the LM339 in my pspice package thats why i chose to use the LM324.

does the attached file make more sense now? or am i still doing something wrong.

as to current detection for now i have no ideas, wot i do on pspice is to put a current marker on my load(resistor+inductor combination)

please look at graph

The 2nd graph shows what the voltage at the mosfet drain looks like.

i am not sure i am doing this right yet.

i know the max current across my load is 2.3A ( source being 24 and impedance of load being 10.48)

however what i want to show is that the overvoltage causes the current in the load to increase beyond 2.3A thereafter showing that the mosfet by switching off interrupts the circuit and allows the current in the load to dissipate and drop back down to 2.3A
 

Attachments

Last edited:

thatoneguy

Joined Feb 19, 2009
6,359
I guess where I am confused is how do you know when the load current is low enough to turn on the motor/device again? Once it is disconnected, there's no current to measure.

Your original circuit was a crowbar, now it is in series.
 

Thread Starter

newbieateverything

Joined Feb 25, 2013
95
I guess where I am confused is how do you know when the load current is low enough to turn on the motor/device again? Once it is disconnected, there's no current to measure.

Your original circuit was a crowbar, now it is in series.
my original circuit was wrong in many ways have a look at this one and tell me if it makes more sense.

I determine the parameters because i need to control it to react the same to various situations, so in the case of a lighting strike reduce current in load to 2.3 this is the current in the load when the supply voltage of 24 runs through it, it is to react this way for any other overvoltage situations.

2.3 is the normal operating current i want in my load.
The overvoltage will cause this value to increase, the circuit i am trying to design is to reduce the current back to 2.3 after it has increased due to the overvoltage.

hope it makes sense.

The circuit i attached is still wrong, a little bit anywayz, i dont know where to introduce my overvoltage
 

Attachments

thatoneguy

Joined Feb 19, 2009
6,359
In the event of a lightning strike, if the surge isn't stopped by a MOV, do you think the tiny spacing between bonds in a MOSFET will prevent it from passing through?

You may want to look into a line conditioning system, Tripp Lite makes some decent ones. Put in front of UPS. They have great surge suppression (far better than power strip "surge suppressors"), in addition to over/under voltage correction, and filters to make the output as close to a true sine wave as possible.
 

Thread Starter

newbieateverything

Joined Feb 25, 2013
95
In the event of a lightning strike, if the surge isn't stopped by a MOV, do you think the tiny spacing between bonds in a MOSFET will prevent it from passing through?

You may want to look into a line conditioning system, Tripp Lite makes some decent ones. Put in front of UPS. They have great surge suppression (far better than power strip "surge suppressors"), in addition to over/under voltage correction, and filters to make the output as close to a true sine wave as possible.
:mad: its $25000....???? not gonna spend that much on a hobby but good idea in theory.

But its simulation, i need to start from somewhere in reality SCB aren't going to protect a load onboard a plane against a lighting strike at least not without a SCB+MCB combination. (hybrid circuit breaker)

I am just tryign to test the limits of my SCB design, my voltage pulse is set to 240V, i.e 10 times the supply on the load, lightning strikes have far higher voltages.

i just want to show via simulation that a solid state circuit breaker can be used to protect a load from a voltage pulse 10 times the supply voltage.

however the problem is there are still some errors with my design and i am looking for some help.
 

thatoneguy

Joined Feb 19, 2009
6,359
:mad: its $25000....???? not gonna spend that much on a hobby but good idea in theory.
No, $25,000 is their guarantee, if any equipment attached to a line conditioner fails due to a surge/brownout/over-voltage, they will replace up to $25k in damaged equipment you had plugged into it.

The Conditioner itself is around $200, some are a bit less or a bit more, depending on capacity.

For a hobby circuit, a standard "Surge Strip" from a reputable brand (not a no-name Chinese product) protects devices extremely well, simply use your adapter plugged into one with a 900 Joule or higher rating.

Keep in mind that the good surge suppressors protect against a surge on any of the inputs, including ground. Ground Swell is a killer, due to the earth ground raising hundreds to thousands of volts over the line voltage.

More info here
Good information on Lightning strikes in general
 

Thread Starter

newbieateverything

Joined Feb 25, 2013
95
No, $25,000 is their guarantee, if any equipment attached to a line conditioner fails due to a surge/brownout/over-voltage, they will replace up to $25k in damaged equipment you had plugged into it.

The Conditioner itself is around $200, some are a bit less or a bit more, depending on capacity.

For a hobby circuit, a standard "Surge Strip" from a reputable brand (not a no-name Chinese product) protects devices extremely well, simply use your adapter plugged into one with a 900 Joule or higher rating.

Keep in mind that the good surge suppressors protect against a surge on any of the inputs, including ground. Ground Swell is a killer, due to the earth ground raising hundreds to thousands of volts over the line voltage.

More info here
Good information on Lightning strikes in general
I know of surge strips, however i want to build my device to function as a surge strip, i.e protecting the device from voltage spikes, a circuit breaker pretty much but with a mosfet.

Any ideas on how i can improve my design in simulation before i hit the breadboard?
 

thatoneguy

Joined Feb 19, 2009
6,359
This hasn't really been packaged neatly in one post, but a great way to help people help you is to boil down your requirements:

In this case:
1) Power source and variances you want to protect against ("ALL" is too broad when lightning is involved).
2) Type of load, current drawn by load in start-up and in normal operation so that "ab-normal" can be more clearly defined.
3) Approximate ability with electronic design or assembly / skill levels.

To start, I'd suggest looking at the Datasheet for a MOV which shows what they will protect against, and how fast they will respond. They are relatively low cost, and are essentially a fuse. That is but one example, and costs about a quarter.

Look at conditions your circuit needs to attempt surviving in addition to the protection from a MOV, and detail what the limiting should be. The type of load is important here, as well. For example, an incandescent light bulb is more forgiving of varying voltage than a microcontroller.
 

Thread Starter

newbieateverything

Joined Feb 25, 2013
95
This hasn't really been packaged neatly in one post, but a great way to help people help you is to boil down your requirements:

In this case:
1) Power source and variances you want to protect against ("ALL" is too broad when lightning is involved).
2) Type of load, current drawn by load in start-up and in normal operation so that "ab-normal" can be more clearly defined.
3) Approximate ability with electronic design or assembly / skill levels.

To start, I'd suggest looking at the Datasheet for a MOV which shows what they will protect against, and how fast they will respond. They are relatively low cost, and are essentially a fuse. That is but one example, and costs about a quarter.

Look at conditions your circuit needs to attempt surviving in addition to the protection from a MOV, and detail what the limiting should be. The type of load is important here, as well. For example, an incandescent light bulb is more forgiving of varying voltage than a microcontroller.
ill get back to you on these details.
For now what i can say with certainty is that i want my circuit breaker is going to be used for DC applications specifically and not AC.

my power source will be from the mains 240V, i will step this down to 120V and use that to power my load.
For now said load is simply an inductor and resistor in series, as most loads have inductive and resistive properties, as to the load specifically i haven't decided, maybe an (LED, they are cheap!!)

I want my breaker to protect against overvoltages up to 10 times the supply so 1200V, (in theory of course, practically it wont be possible till i combine my design with a mechanical circuit breaker)

2) current drawn by the load when supply is 120V is 11.5A, abnormally would be any currents more than 11.5*root 2= 16 appx

3) my assembly skill level is pretty basic i am no specialist not by a long shot but if i have a design on PSPICE that works i will be able to replicate it on a breadboard

For now this is just the mosfet switch aspect of the circuit breaker, i realize there are other things involved.
 

Thread Starter

newbieateverything

Joined Feb 25, 2013
95
hi, i understand what my circuit requires in theory, however i dont think its smart to start trying to build it if at the simulation level there are still some errors.

1) i am not sure where i should introduce my pulse signal

2) the results i obtain dont show that my mosfet is interrupting the circuit to protect the load from the pulse voltage.


Notes: there is a simple differential amplifier configuration with a gain of 2.
Pulse signal: V2= 240, V1=0, TD= 0, TR= 1us, TF= 1us, PW= 1ms, PER= 2ms
The 90V opposes the 120V supply to ensure that a max of 30V enters Vcc of LM324.

please help, what is wrong with my design

Please note, this is a circuit breaker using a mosfet as the solid state device, when the mosfet turns off it interrupts the circuit protecting the load from the pulse voltage causing the current in the load to increase to more than 2.86A*root 2,
it turns back on when the current in the load has been reduced to its normal operating current which should be at 2.86A
 

Attachments

Last edited:

Thread Starter

newbieateverything

Joined Feb 25, 2013
95
Take a peek at ICs like the LTC4361 and other automatic voltage/over-current protection products.
thanks will use that when i am building it,
however before i get there could you please tell me if my circuit design is correct or are some errors,i.e can i replace the LM324 with the suggested LTC4361. If not possible the design with the LM324 is it correct?

i would rather not blow up the mosfet or other devices, while most are cheap i dont want to have the shards flying about


edit: Looking at the LTC4361 voltage range i am not sure if it would be suitable to use as my overvoltage protection device
 
Last edited:

thatoneguy

Joined Feb 19, 2009
6,359
I should have been more specific (or general?). I suggest looking at the datasheets for those (Automatic Voltage Regulation and Automatic Current Protection) ICs to see the internal block diagrams/schematics to get an idea of what goes in to building one.
 

Thread Starter

newbieateverything

Joined Feb 25, 2013
95
more specific
but thanks, i just ordered one and i will use that to protect my load.
My end goal is to build a complete solid state circuit breaker, so ill just get the LTC4356
 

Thread Starter

newbieateverything

Joined Feb 25, 2013
95
Please let me know if the design in the attached file is correct

the LED would remain on if there is a fault that cannot be resolved
off when there is no fault
and blinking when the mosfet interrupts the circuit continuously till the fault is resolved.

if there is no difference between the overvoltage and the ref voltage (feedback from the mosfet source resistance back to the negative input of the comparator)

no signal is sent to the 555 timer, which leaves the mosfet off.

I think this configuration might be better to use vs using the LTC4356, which makes the task a TAD easier but then wont be a hobby. :rolleyes:
 

Attachments

Last edited:

thatoneguy

Joined Feb 19, 2009
6,359
One slight issue with your circuit, the references are all created by voltage dividers.

You'll need to pick something like a 22 Volt Zener diode to have a solid voltage reference.

Without a Zener/bandgap ref or similar stable voltage reference (independent of source voltage), measuring supply voltage is rather difficult.
 

Thread Starter

newbieateverything

Joined Feb 25, 2013
95
One slight issue with your circuit, the references are all created by voltage dividers.

You'll need to pick something like a 22 Volt Zener diode to have a solid voltage reference.

Without a Zener/bandgap ref or similar stable voltage reference (independent of source voltage), measuring supply voltage is rather difficult.
apart from adding the zener diode to achieve stable voltage reference are there any other errors with my design,
thanks alot for your help btw.

the zener diode would be located after the resistance connected to the source of the mosfet right
 
Status
Not open for further replies.
Top