Hello, I see lots of PA designs here that use glass fuses on the rails and some that have elaborate SSR based protection on the output. Don't think I've seen one that uses Polyswitch's? Is there a reason they are not used more widely?
Perceived sound quality? Reliability? Trip current accuracy?
Back in the day when I was more fanatical about SQ, fuses and relays were a no no so cant imagine the poly would be worse?
Perceived sound quality? Reliability? Trip current accuracy?
Back in the day when I was more fanatical about SQ, fuses and relays were a no no so cant imagine the poly would be worse?
You obviously read the big debate on EW about the rising and falling current in relation to music being played causing a heating and cooling of the output "safety " fuse which in turn caused a type of distortion to be introduced .
Yes its a fact but it depends then on how significant this was in relation to the overall fidelity at higher volumes.
Poly switches aren't exactly the same but are "re-settable " ( due to a set overload figure ) so cut off the output when a safe level is exceeded ( of current )
It did contribute to more copies being sold and more readers letters to EW over quite a space in time as views were very fixed, but again interesting reading.
Again looking at a manufacturing graph Poly switches aren't "instantaneous " but have a rounded rise and according to the manufacturers contain graphite a well known noise producer under heavy load.
I cannot comment on the actual practical fidelity of those but you would think they are superior to a heating & cooling piece of wire like a fuse.
Yes its a fact but it depends then on how significant this was in relation to the overall fidelity at higher volumes.
Poly switches aren't exactly the same but are "re-settable " ( due to a set overload figure ) so cut off the output when a safe level is exceeded ( of current )
It did contribute to more copies being sold and more readers letters to EW over quite a space in time as views were very fixed, but again interesting reading.
Again looking at a manufacturing graph Poly switches aren't "instantaneous " but have a rounded rise and according to the manufacturers contain graphite a well known noise producer under heavy load.
I cannot comment on the actual practical fidelity of those but you would think they are superior to a heating & cooling piece of wire like a fuse.
I have used Polyswitches in several low power supplies. They work well to protect from MAJOR failure. They increase resistance as temperature increases and it is not linear function. Using them in an audio signal location like protection of a loudspeaker will increase distortion as the RMS power increases the levels go lower. Polyswitches in the DC supply for a power amplifier as it will not fully protect the amplifier if the output stage has a MAJOR failure.
The use of a fuse fully disconnects the power to the circuit and improves the LIFE SAFETY of the user.
Duke
The use of a fuse fully disconnects the power to the circuit and improves the LIFE SAFETY of the user.
Duke
Poly Switches work well, but they increase in internal resistance when they are tripped and reset repeatedly.
Thanks all. I was thinking of using them on the rails so any noise would be dealt with by PSRR. Les Sage used to use them between the 0v tap of the transformer and the star earth. The thinking being significant current only flows through that point under DC fault conditions. Not sure that's a great idea as at best you'll discharge the 1/2 capacitor bank into the fault once its tripped.
IMHO the most frequent fault condition is a short of the output, even only momentary.
Neither fuses will protect, power transistors are gone in an instant.
Some fail in short condition and now the fuse blows and protects eventually the woofer, the tweeter might be gone already
Only a properly designed current limit taking the SOA into account combined with a DC-limit will keep the semis alive.
Thermal protection can be added for pro-equipment
A separate relays can protect the speaker.
Neither fuses will protect, power transistors are gone in an instant.
Some fail in short condition and now the fuse blows and protects eventually the woofer, the tweeter might be gone already
Only a properly designed current limit taking the SOA into account combined with a DC-limit will keep the semis alive.
Thermal protection can be added for pro-equipment
A separate relays can protect the speaker.
Last edited:
Probably the size they'd have to be for handling many amps at dozens of volts...
From the laws of physics you can figure out the polyfuse volume must be related to the power being fused (assuming the same current density and resistivity at activation).
One designed for 0.5A at 5V is a reasonable size, but 8A at 65V is likely
to be larger and considerably more expensive than a wire fuse.
I've been thinking of using one to protect the 60 W rated tweeter in my 400 W rated 2 way speaker. I seriously don't want to disassemble a speaker case to replace a wire fuse. Will parallel the tweeter with a 22 v clamp of semiconductors to drive the PTCR fuse into open. Say about 2.7 A trip, which is about an 800 ma nominal fuse.
An ideal use case. I saw a design that put a globe / resistor in parallel with the poly so if tripped the tweeter would go into limp mode. Simple and effective for public address systems.
Maybe what you should do is get a polyfuse, measure it's resistance, trip it and measure the recovered resistance. The datasheet shows the resistance is increased, but maybe that's a worst case and they would be suitable.
-Chris
-Chris
over time when operated near the trip point they will deteriorate. even at half the trip point it gets worse
The delts between Ih and It is x2. They also slow to trip in comparison. Im using them alot lately on 5V and 3V3 DC and protection for low voltage transformers secondaries. Used in this application is basic protection, since most regulators and transformers can handle an overload for the time it takes to trip these devices. Also the transformers I use have integrated thermal fuses as additional protection; these are a pain in the a## to replace.
But i guess we'll never know till someone tries it over an extended time period and reports back.
But i guess we'll never know till someone tries it over an extended time period and reports back.