SparkFun Forums

[gussy]

I am having some trouble getting my head around the PTC parameters for selecting an appropriate PTC fuse.

The terms Itrip and Ihold are what's geting me confused.
Ih - The maximum steady state current at 20°C that can be passed through a PolySwitch device without causing the device to trip
lh - The minimum current that will cause the PolySwitch device to trip at 20°C

Say I have a circuit with a regulator that can only supple 150mA, I want to limit the input current to now use anything greater.

A PTC has a Ih of 150mA and an It of 320mA. Now my question is, what really happens between Ih and It? Is this where the resistance starts to increase, until it becomes open circuit at It?

If so, my question is, is my assumption that choosing a PTC with an Ihold of 150mA the correct one for my given system?

Now for the zener diode. I would like to add overvoltage protection, currently my circuit is something like this;
Vin1----->|------[PTC]-----[VREG]-----
Vin2----->|--^

The two diodes OR the power supply to one of the other, as well as providing protection against reverse polarity connection.

Now I would like to add in a TVS Zener diode. The regulator can handle up to 16v but if Vin1 is greater than 7v, there is a problem.
If I put in over voltage protection to protect against >7v on Vin, that breakdown voltage should I look for, something a bit higher than 7v?
What's the best way to connect this in, parallel with Vin1 and GND?

Some people say that a zener when protecting against over voltage would blow up, these TVS Zeners can take 1000's of watts, so I assume that this isn't an issue?

[theatrus]

In general terms, think of a PTC like your house breaker. At the trip current, the device will rapidly enter shutdown. At the hold current, there is a temperature dependent effect and a slow time delay until the device resistance has gone up. In addition, after the device has been tripped, it would not enter its normal low-on-resistance mode until the possible current has dropped below Ihold.

Note that the PTCs have an on resistance at any current level - you need to weigh the current vs on resistance carefully.

Unless your regulator does not feature a thermal shutdown mode, I wouldn't bother with a PTC on the regulated side.

A zener diode has a very finite energy rating - many can take several hundred watts of energy, but only for several microseconds (common with ESD protection diodes). Simply adding a zener across the rails will lead to a blown up zener and an unprotected circuit. This is where the PTC comes into play - the PTC should be before the zener (after your diode OR). When the zener conducts, the PTC will limit input current to safe levels before the zener catches fire or falls off the board.

[gussy]

Quick reply, awesome!

In general terms, think of a PTC like your house breaker. At the trip current, the device will rapidly enter shutdown. At the hold current, there is a temperature dependent effect and a slow time delay until the device resistance has gone up. In addition, after the device has been tripped, it would not enter its normal low-on-resistance mode until the possible current has dropped below Ihold.

Note that the PTCs have an on resistance at any current level - you need to weigh the current vs on resistance carefully.

Unless your regulator does not feature a thermal shutdown mode, I wouldn't bother with a PTC on the regulated side.

Makes sense. The regulator does have thermal shutdown.

A zener diode has a very finite energy rating - many can take several hundred watts of energy, but only for several microseconds (common with ESD protection diodes). Simply adding a zener across the rails will lead to a blown up zener and an unprotected circuit. This is where the PTC comes into play - the PTC should be before the zener (after your diode OR). When the zener conducts, the PTC will limit input current to safe levels before the zener catches fire or falls off the board.

Awesome, very clever. I will do that then, add in a Zener before the PTC. Coupled with the diode OR, this should provide some pretty good protection against over-voltage/current and reverse polarity. I want this as foolproof as possible, and you should never underestimate the ingenuity of fools
Is this arrangement what is referred to as a crowbar circuit?

If I want this breakdown the occur at 7v, I should pick a Zener with a breakdown voltage of exactly 7v? What about the Reverse Standoff voltage?

Thanks for the advice, much appreciated!

[theatrus]

Eeep. I meant Zener after the PTC. Placing it before the PTC leads to the problems I mentioned above.

[MichaelN]

I do this on my circuits, but you do have to carefully look at the parameters of the TVS and polyfuse. In particular, the TVS needs to be big enough to absorb the energy it will be subjected to in the time it takes to trip the polyfuse.

The minimum value you should choose for the TVS "Standoff" voltage rating is the maximum legitimate power supply voltage in your system. If there is plenty of "overhead" between the maximum legitimate supply voltage and the maximum voltage rating of your regulator, it pays to choose a Standoff rating a bit higher than the minimum, so it doesn't conduct with minor voltage surges.

A very important thing to consider with any overvoltage protection device is the maximum voltage that the device will clamp the supply at under real conditions. A TVS device with a 7V Standoff would clamp the voltage MUCH higher than this under many "real world" surge / spike conditions.

You can actually omit the reverse polarity protection diode in many cases too (if you use a unidirectional TVS) - just make sure your polyfuse and TVS can handle it. When the supply voltage is reversed, the TVS will just look like a forward-biased diode - ie, quite a low impedance, and the polyfuse will trip quickly. Check the maximum current rating of the polyfuse in particular, and factor in the impedances to calculate the current if the supply polarity is reversed. Often the resistance of the polyfuse itself will limit the current to a safe value (but this will depend on the supply voltage).

Another thing to check is the voltage rating of the polyfuse - this is pretty much determined by the power dissipation of the polyfuse when in the "tripped" state. That's another thing to consider with polyfuses - by their very nature, they still conduct some current when "tripped".

[gussy]

Thanks for the reply Michael, that helps a lot.

So here is a bit more info about the system;
Regulator is capable of up to 16v, but should hit thermal shutdown well before 16v. Maximum voltage, worst case scenario would be probably anywhere from 12-24v. Typically it should be regulated 5v. The system consumes, 100mA and the regulator can handle 160mA.

There are 2 voltage input, one that is 5v USB and the other which is the one that could possibly go overvoltage.

I have chosen the following TVS and polyfuse form Digikey, based on your advice;

TVS (VESD08-02V-GS08)
Voltage - Reverse Standoff (Typ) = 8v
Voltage - Breakdown = 9v
Polarization = Unidirectional
Max clamping voltage = 30v
Power rating = Unknown

Polyfuse (1206L016WR)
Trip time = 0.3s
Max voltage = 30v
Ih = 160mA
It = 370mA
Max current = 100A

These look pretty good to me, but I am new to this type of protection, what do you think?
By the looks of the ratings, anything over 30v is bad, but under that everything is ok.

[MichaelN]

The polyfuse should be OK, but that TVS is too small. If it ever reached the Max clamping voltage of 30v, then your regulator would be fried (it is only rated to 16V). This TVS reaches a clamping voltage of 30V at only 4A current. Not only would it faill to protect the regulator, but would probably be fried itself while the TVS is tripping.

Something bigger (eg a SMA sized package) would be better. For example, this device has a standoff rating of 7V, and a max clamping voltage of 12V at a current of 33A:

http://search.digikey.com/scripts/DkSea ... .0ALFCT-ND

Of course, those ratings are only for a very brief time.

To figure out whether the TVS would handle the power dissipation while it is waiting for the polyfuse to trip during various different faults would actually require a detailed analysis using information in the TVS and polyfuse datasheets. Actually, since that TVS datasheet doesn't give detailed curves of Voltage versus Current, you'd need to make some guesses.

For example, an input of 16V would result in a maximum current of probably about 4 or 5A, based on the minimum resistance of the TVS, ie, 1.2 ohms, and a guess of the TVS clamping voltage. Extrapolating Figure 1 on page 3 of the TVS datasheet, the TVS should be able to handle this power dissipation (say, 50W) for a bit more than 0.1 second.

Looking at the trip time curves of page 2 of the polyfuse datasheet, it should trip in under 0.04 seconds with a current of 4 or 5 amps. Thus the polyfuse should trip well before the TVS is in danger.

I wouldn't go any smaller than my suggested TVS device in this application...

[gussy]

Another nicely detailed reply, thankyou!

Here is where I'm currently at:

D1 needs to be changed now to something much bigger. At the moment I believe it will blow before the TVS or Polyfuse ever gets used.

One way of getting around this could be to put the TVS+Polyfuse before D1 and only protect one power input line. This probably makes a bit more sense, seeing as USB will never get above 7-8v.

Thoughts?

[MichaelN]

It's not clear to me where you are connecting the regulator, and what (if anything) is connected to the left-hand diode's anode.

To have the option to connect both 5V from USB & an external voltage supply, I would connect the polyfuse to the common cathode of the diodes, and connect the TVS & regulator to the other side of the polyfuse. The 5V supply would then connect to the anode of one of the diodes, and the external power to the anode of the other. As for the polyfuse, you could connect it between the common cathode of the diodes and the TVS / regulator.

And yes, you'll need to choose another diode type (the BAT54C is only rated to 1A pulsed current for 10ms). There are plenty of small schottky diodes that have much higher peak current ratings. You might not get something suitable in a dual-diode SOT-23 package though...

[gussy]

Picture was labled wrong, didn't pick that up sorry.
Left anode is VUSB, right is +5v (which is the one to protect). V_UNREG connects to the VIN of the regulator.

Post #4 confused me, I originally thought the TVS should be on the side of the regulator, after the polyfuse. Turns out this is how it should be?