Trying to find a device

[DerStrom8]

Hi all,

I'm trying to find a device that allows current to flow until it reaches a certain value (i.e. 50mA) and shuts off (much like a fuse). However, this device must have very low resistance when on (~0.01-0.1 ohms), must be self-resettable, and must be surface mount. I was looking at PTC fuses but their resistance is relatively high even under normal operating conditions. Does such a device exist? Unfortunately I'm limited to very few components (preferably just the one) due to space constraints and to the fact that the current flowing through the main line must not be affected much (except to be cut off above a certain current).

Any thoughts? I feel like this should be a simple answer but I have been trying to figure it out for the past day with no results to speak of.

Thanks,
Matt

 

[Ian Rogers]

didn't we do this with a couple of transistors and a 0.1 resistor

 

[DerStrom8]

didn't we do this with a couple of transistors and a 0.1 resistor

There are a couple of issues with that. 1) The resistance should be as low as possible (Ideally it would be 0 ohms, but no real-world device will be able to do that). 2) There should be no voltage dropped (well, due to what was mentioned in #1 I am force to say very, very little voltage dropped, preferably in the range of millivolts), 3) The current must be cut off at 50mA +- 10mA, which is precision you can't easily get from transistors, and 4) I have very, very little room on this board.

 

[ronsimpson]

0 ohms

Hall sensors approach 0 ohms. You will need to add a amp and SR-FF which will drive MOSFETS for the switch.
Tooooo many parts.

 

[Ian Rogers]

If you can't afford 0.6v drop or +0.10 resistance, then I doubt you'll find something..

To be honest, I've never needed anything like this..

 

[ronsimpson]

You will need to add a amp and SR-FF

A SR-FF could be as simple as a SCR or two transistors to make a SCR.

 

[BobW]

Maybe something like this:
https://www.onsemi.com/pub_link/Collateral/NIS5112-D.PDF

 

[Pommie]

Maybe I'm misunderstanding but won't this oscillate?
Current exceeds 50mA - current cuts off.
Current is zero - switch back on.
Rinse and repeat.

Is some time constant involved?

Mike.

 

[DerStrom8]

In the example of the PTC fuse a small amount of current flows to keep the device switched off until the current is removed. This would work, if it weren't for the significant resistances that come with the component. Ideally the device would limit the current to only a few milliamps until it's completely removed.

 

[ronsimpson]

I fount PTC respond v...e...r...y...... .....s.....l....o....w

 

[NorthGuy]

A mechanical relay will definitely have the smallest resistance possible, but it probably will require some room.

 

[simonbramble]

This part should do:
http://cds.linear.com/docs/en/datasheet/4363fb.pdf
It has a few extra bells and whistles (over and reverse voltage protection), but the -2 version has a cool down period after which the power is reasserted

 

[Colin]

You can use a coil with a few turns and a reed switch.

 

[ronv]

What voltage do you have available to run this gadget?
How long do you want it to stay off before it resets itself?

 

[DerStrom8]

This part should do:
http://cds.linear.com/docs/en/datasheet/4363fb.pdf
It has a few extra bells and whistles (over and reverse voltage protection), but the -2 version has a cool down period after which the power is reasserted

Thanks for the datasheet. Unfortunately that device is much too large. This board is about 0.5 inches wide by 1 inch long and it's packed with traces. That's why I was hoping to find an inline device like a polyfuse.

You can use a coil with a few turns and a reed switch.

That was a thought but I'm afraid it won't help me much. It's far too difficult to calibrate and will be very unstable (inconsistent).

What voltage do you have available to run this gadget?
How long do you want it to stay off before it resets itself?

Ideally it would cut (or limit) current until the power is removed.

I think I'm just going to have to make some changes upstream of this board (it connects to another board via a ribbon cable) to ensure no damage can be done if something goes wrong.

Thanks for the suggestions though, everyone.

Matt

 

[Colin]

You can do it with 1 ohm resistor (or less) and a completely isolated circuit.

 

[NorthGuy]

You can use a coil with a few turns and a reed switch.

This is called "relay"

 

[AnalogKid]

Self-resetting can be tricky. It is easy to inadvertently build a power oscillator. Normally I'd say to use an LTC electronic circuit breaker controller (some have a hiccup mode auto-reset), but your midrange voltage drop is only 2.5 mV, not enough to sense without outboard amplifiers, etc. Multiple turns through a closed loop hall effect sensor will get you the low voltage drop, but they are not super-accurate and pretty large. As above, a form C reed relay would work, but without some external control it would chatter like a point-contact buzzer. Hmmm...

ak

 

[nsaspook]

Ideally it would cut (or limit) current until the power is removed.

I think I'm just going to have to make some changes upstream of this board (it connects to another board via a ribbon cable) to ensure no damage can be done if something goes wrong.
Matt

Maybe a crowbar circuit (SCR) on the current-limited power feed to the board that's triggered by a over-current sense to the board.