Polyfuse

On 2020-03-18 11:30 a.m., Rene Lamontagne wrote:
Does anyone know how long it takes for a USB1 (500 ma) tripped polyfuse
to reset after the load is removed?

Rene

Reason I'm asking is I was using a UM24 USB tester and have now
purchased a UM25C tester
https://www.amazon.ca/dp/B0826492XM/..._TE_item_image
with Bluetooth and want to do some USB current testing and kinda would
like to know beforehand what pitfalls to expect.

Rene

On Wed, 18 Mar 2020 13:16:42 -0500, Rene Lamontagne
wrote:

On 2020-03-18 11:30 a.m., Rene Lamontagne wrote:
Does anyone know how long it takes for a USB1 (500 ma) tripped polyfuse
to reset after the load is removed?

Rene

Reason I'm asking is I was using a UM24 USB tester and have now
purchased a UM25C tester
https://www.amazon.ca/dp/B0826492XM/..._TE_item_image
with Bluetooth and want to do some USB current testing and kinda would
like to know beforehand what pitfalls to expect.

Rene


I ain't no expert but I did fool around with a polyfuse intending to
apply it in a project. I was disappointed because the ones I had
would trip, but to reset, the current had to be well below the trip
point. (If it tripped at an amp it may take a current lower than 1/2
amp to reset it)

I ended up kludging together a current relay (Magnetic Reed Switch,
with a few turns of wire around it) operating a conventional relay
controlling the power. There was much less hysteresis.

I've got several of those USB volt amp testers, and one dummy load (as
shown). They are very handy for troubleshooting charging problems.

On 2020-03-18 5:09 p.m., default wrote:
On Wed, 18 Mar 2020 13:16:42 -0500, Rene Lamontagne
wrote:

On 2020-03-18 11:30 a.m., Rene Lamontagne wrote:
Does anyone know how long it takes for a USB1 (500 ma) tripped polyfuse
to reset after the load is removed?

Rene

Reason I'm asking is I was using a UM24 USB tester and have now
purchased a UM25C tester
https://www.amazon.ca/dp/B0826492XM/..._TE_item_image
with Bluetooth and want to do some USB current testing and kinda would
like to know beforehand what pitfalls to expect.

Rene


I ain't no expert but I did fool around with a polyfuse intending to
apply it in a project. I was disappointed because the ones I had
would trip, but to reset, the current had to be well below the trip
point. (If it tripped at an amp it may take a current lower than 1/2
amp to reset it)

I ended up kludging together a current relay (Magnetic Reed Switch,
with a few turns of wire around it) operating a conventional relay
controlling the power. There was much less hysteresis.

I've got several of those USB volt amp testers, and one dummy load (as
shown). They are very handy for troubleshooting charging problems.


From what I have read the load must drop considerably for the
temperature to drop for it to reset.
The dummy load was a good add on and helps to figure what to expect.

Rene

On 18/03/2020 22:09, default wrote:
I ain't no expert but I did fool around with a polyfuse intending to
apply it in a project. I was disappointed because the ones I had
would trip, but to reset, the current had to be well below the trip
point. (If it tripped at an amp it may take a current lower than 1/2
amp to reset it)

I ended up kludging together a current relay (Magnetic Reed Switch,
with a few turns of wire around it) operating a conventional relay
controlling the power. There was much less hysteresis.

I've got several of those USB volt amp testers, and one dummy load (as
shown). They are very handy for troubleshooting charging problems.

Remember it's temperature that really trips them, not current.

When not tripped they have a low resistance but once they have tripped
their resistance increases to a much higher value so that the current is
limited to a safer value.

If we say that, for instance, the resistance increases to 25 times the
original un-tripped value then it will only take one 25th of current to
keep it n the tripped state.

I have no idea if 25 is a typical factor by which the resistance
increases but you get the idea.

--
Brian Gregory (in England).

Rene Lamontagne wrote:
On 2020-03-18 5:09 p.m., default wrote:
On Wed, 18 Mar 2020 13:16:42 -0500, Rene Lamontagne
wrote:

On 2020-03-18 11:30 a.m., Rene Lamontagne wrote:
Does anyone know how long it takes for a USB1 (500 ma) tripped polyfuse
to reset after the load is removed?

Rene

Reason I'm asking is I was using a UM24 USB tester and have now
purchased a UM25C tester
https://www.amazon.ca/dp/B0826492XM/..._TE_item_image

with Bluetooth and want to do some USB current testing and kinda would
like to know beforehand what pitfalls to expect.

Rene


I ain't no expert but I did fool around with a polyfuse intending to
apply it in a project. I was disappointed because the ones I had
would trip, but to reset, the current had to be well below the trip
point. (If it tripped at an amp it may take a current lower than 1/2
amp to reset it)

I ended up kludging together a current relay (Magnetic Reed Switch,
with a few turns of wire around it) operating a conventional relay
controlling the power. There was much less hysteresis.

I've got several of those USB volt amp testers, and one dummy load (as
shown). They are very handy for troubleshooting charging problems.


From what I have read the load must drop considerably for the
temperature to drop for it to reset.
The dummy load was a good add on and helps to figure what to expect.

Rene

Recovery is asymptotic. It takes a while for the fuse to
approach it's original state. Bulk recrystallization might take
a time measured in seconds, while "low resistance" might take
minutes. If tripped enough times, a Polyfuse can also reach
a state where it's better to just replace it. The cycle count
probably isn't infinity.

https://en.wikipedia.org/wiki/Resettable_fuse

On some of my motherboards, a 1.1A Polyfuse powers two USB
connectors in a USB stack. If one port remains unoccupied,
then a relatively generous current can be drawn from
the remaining port (for example, to run a "USB light").

USB3 ports use an even larger fuse. And that fuse is
sometimes physically one or two sizes larger than the
physical size of a USB2 port fuse solution.

*******

Modern motherboards run all the USB ports from +5VSB. The
ATX supply has a 3 amp limit or so on that rail. Check the
label on your supply, to see if it's 2.5A or 3A.

Older motherboards, maybe 2005 - 2008 or so, you might find
a 1x3 or 2x3 pin header next to the fuse, and the pin header
jumper selects to run the port from +5V or from +5VSB.

+5V @ 20A
+5VSB @ 3A

If a header selector is set to +5V, then you'd be in
serious trouble if you bypassed the fuse to prevent
it from popping. At 20A of fault current flow level,
that's sufficient current to burn the "average"
track in a motherboard. Only the main distribution
planes handle more current than that (~100 amps in
VCore plane). The fuses used on the motherboard, provide
partial protection against silly stuff. As another example,
the fan headers have *no* fuse, and users posting to USENET
regularly report they've managed to burn out the
+12V to the three pin fan headers because there's no fuse.

"Know your limits" when working inside a PC, for best results :-)

Paul

On 2020-03-18 8:15 p.m., Paul wrote:
Rene Lamontagne wrote:
On 2020-03-18 5:09 p.m., default wrote:
On Wed, 18 Mar 2020 13:16:42 -0500, Rene Lamontagne
wrote:

On 2020-03-18 11:30 a.m., Rene Lamontagne wrote:
Does anyone know how long it takes for a USB1 (500 ma) tripped
polyfuse
to reset after the load is removed?

Rene

Reason I'm asking is I was using a UM24 USB tester and have now
purchased a UM25C tester
https://www.amazon.ca/dp/B0826492XM/..._TE_item_image

Â* with Bluetooth and want to do some USB current testing and kinda
would
like to know beforehand what pitfalls to expect.

Rene


I ain't no expert but I did fool around with a polyfuse intending to
apply it in a project.Â* I was disappointed because the ones I had
would trip, but to reset, the current had to be well below the trip
point.Â* (If it tripped at an amp it may take a current lower than 1/2
amp to reset it)

I ended up kludging together a current relay (Magnetic Reed Switch,
with a few turns of wire around it) operating a conventional relay
controlling the power.Â* There was much less hysteresis.

I've got several of those USB volt amp testers, and one dummy load (as
shown).Â* They are very handy for troubleshooting charging problems.


Â*From what I have read the load must drop considerably for the
temperature to drop for it to reset.
The dummy load was a good add on and helps to figure what to expect.

Rene

Recovery is asymptotic. It takes a while for the fuse to
approach it's original state. Bulk recrystallization might take
a time measured in seconds, while "low resistance" might take
minutes. If tripped enough times, a Polyfuse can also reach
a state where it's better to just replace it. The cycle count
probably isn't infinity.

https://en.wikipedia.org/wiki/Resettable_fuse

On some of my motherboards, a 1.1A Polyfuse powers two USB
connectors in a USB stack. If one port remains unoccupied,
then a relatively generous current can be drawn from
the remaining port (for example, to run a "USB light").

USB3 ports use an even larger fuse. And that fuse is
sometimes physically one or two sizes larger than the
physical size of a USB2 port fuse solution.

*******

Modern motherboards run all the USB ports from +5VSB. The
ATX supply has a 3 amp limit or so on that rail. Check the
label on your supply, to see if it's 2.5A or 3A.

Older motherboards, maybe 2005 - 2008 or so, you might find
a 1x3 or 2x3 pin header next to the fuse, and the pin header
jumper selects to run the port from +5V or from +5VSB.

Â*Â* +5VÂ*Â* @ 20A
Â*Â* +5VSB @Â* 3A

If a header selector is set to +5V, then you'd be in
serious trouble if you bypassed the fuse to prevent
it from popping. At 20A of fault current flow level,
that's sufficient current to burn the "average"
track in a motherboard. Only the main distribution
planes handle more current than that (~100 amps in
VCore plane). The fuses used on the motherboard, provide
partial protection against silly stuff. As another example,
the fan headers have *no* fuse, and users posting to USENET
regularly report they've managed to burn out the
+12V to the three pin fan headers because there's no fuse.

"Know your limits" when working inside a PC, for best results :-)

Â*Â* Paul


Thanks to all for the replies, And thanks Paul for the Wikipedia
pointer, that explains it's function quite well, Now I know it can take
seconds or more with load removed.


Rene

On Thu, 19 Mar 2020 00:28:29 +0000, Brian Gregory
wrote:

On 18/03/2020 22:09, default wrote:
I ain't no expert but I did fool around with a polyfuse intending to
apply it in a project. I was disappointed because the ones I had
would trip, but to reset, the current had to be well below the trip
point. (If it tripped at an amp it may take a current lower than 1/2
amp to reset it)

I ended up kludging together a current relay (Magnetic Reed Switch,
with a few turns of wire around it) operating a conventional relay
controlling the power. There was much less hysteresis.

I've got several of those USB volt amp testers, and one dummy load (as
shown). They are very handy for troubleshooting charging problems.

Remember it's temperature that really trips them, not current.

When not tripped they have a low resistance but once they have tripped
their resistance increases to a much higher value so that the current is
limited to a safer value.

If we say that, for instance, the resistance increases to 25 times the
original un-tripped value then it will only take one 25th of current to
keep it n the tripped state.

I have no idea if 25 is a typical factor by which the resistance
increases but you get the idea.

for being temperature activated they seem remarkably immune to ambient
temperature (some susceptibility, to be sure but the immunity was
impressive IMO)

On 19/03/2020 10:38, default wrote:
for being temperature activated they seem remarkably immune to ambient
temperature (some susceptibility, to be sure but the immunity was
impressive IMO)


I can't disagree with that.

Maybe it isn't as simple as I thought.

--
Brian Gregory (in England).