From: dbvanhorn on

I'm looking at an existing design that has a polyswitch device to
limit our exposure to shorts on a cable.

The existing part will hold 0.050A and trip at 0.10A which is right
for us, but it's thruhole. All the SMD devices I've seen are 0.150
trip, and they tend to have larger untripped resistance. If I can
find a suitable SMD PTC, then I'm all for that.


I need to emulate the original part, with low on resistance, with some
minimum parts count circuit.


I looked at setting up a CCS with an LDO, but couldn't find anything
that had the right parameters.
Vin is <=14V, Iout nom is 0.050A. I can tolerate up to 5V loss, no
more. I need to survive Vout > Vin by a volt or two.

Ideas?
From: mkaras on
On Apr 14, 10:31 am, dbvanhorn <microb...(a)gmail.com> wrote:
> I'm looking at an existing design that has a polyswitch device to
> limit our exposure to shorts on a cable.
>
> The existing part will hold 0.050A and trip at 0.10A which is right
> for us, but it's thruhole.   All the SMD devices I've seen are 0.150
> trip, and they tend to have larger untripped resistance.  If I can
> find a suitable SMD PTC, then I'm all for that.
>
> I need to emulate the original part, with low on resistance, with some
> minimum parts count circuit.
>
> I looked at setting up a CCS with an LDO, but couldn't find anything
> that had the right parameters.
> Vin is <=14V, Iout nom is 0.050A.  I can tolerate up to 5V loss, no
> more.  I need to survive Vout > Vin by a volt or two.
>
> Ideas?

How about putting a small value resistor in series with a pass FET and
that in series with a Schottky diode. The diode protects for voltage
of output being greater than the input voltage. The voltage drop
across the resistor is monitored for current trip point via a small
single opamp which is setup to turn off the FET when current gets too
high. With proper component selection you should be able to realize a
cutout switch with more accuracy than the PTC, faster acting behavior
and voltage drop of less than half a volt in normal operation.
From: dbvanhorn on

> How about putting a small value resistor in series with a pass FET and
> that in series with a Schottky diode. The diode protects for voltage
> of output being greater than the input voltage. The voltage drop
> across the resistor is monitored for current trip point via a small
> single opamp which is setup to turn off the FET when current gets too
> high. With proper component selection you should be able to realize a
> cutout switch with more accuracy than the PTC, faster acting behavior
> and voltage drop of less than half a volt in normal operation.


Yup, thought about that, but the total cost gets up there. Components
plus placement cost.
It seems like such a part should exist, it's not too far removed from
a linear voltage regulator, and I may end up implementing it that
way. A 317 or similar set up as a constant current source, using its
thermal protection to emulate the polyswitch current foldback.

Linear tech has a part that's close to what I want, but they want more
than $3 for it, which is too much by far.
From: Fred Bartoli on
dbvanhorn a �crit :
> I'm looking at an existing design that has a polyswitch device to
> limit our exposure to shorts on a cable.
>
> The existing part will hold 0.050A and trip at 0.10A which is right
> for us, but it's thruhole. All the SMD devices I've seen are 0.150
> trip, and they tend to have larger untripped resistance. If I can
> find a suitable SMD PTC, then I'm all for that.
>
>
> I need to emulate the original part, with low on resistance, with some
> minimum parts count circuit.
>
>
> I looked at setting up a CCS with an LDO, but couldn't find anything
> that had the right parameters.
> Vin is <=14V, Iout nom is 0.050A. I can tolerate up to 5V loss, no
> more. I need to survive Vout > Vin by a volt or two.
>
> Ideas?

How low a resistance do you want?

Hmmm, 5V max at 50/100mA?
Murata has some SMD PTCs that should fit the bill.

PRG21BB220MB1RK : 0805 size, 22R nom.
trip at 90mA(a)25�C, 110mA@-10�C, 70mA(a)60�C


--
Thanks,
Fred.