The Standex/Meder relay I looked at is rated for 1000 V switching and 2500 V breakdown:
https://www.mouser.ch/datasheet/2/262/8805285100e-1633635.pdf
Shouldn't that be more than enough for a <500 V estat headphone amp?
https://www.mouser.ch/datasheet/2/262/8805285100e-1633635.pdf
Shouldn't that be more than enough for a <500 V estat headphone amp?
If the voltage remains at <500 V, the 5.1k "safety resistors" will limit the current to <100 mA. Looking at the AC diagram in post 54, this current value is in the AC-2 range for <50 ms, and in the AC-3 range for <500 ms. The relay switching time (1.1 ms) is much faster than this. The time limit needs to be configured to the desired value by adjusting the time constant of the RC low pass (R7/R8 and C1 as described in post 79).
Contacts are on a 2.54mm raster, how many mm actual creepage distance do you expect?
And how much distance is needed to call it safe?
I would look at 230Vac creepage distances that are considered safe, instead of considering an adverticed 1kV MAX! rating as safe.
And how much distance is needed to call it safe?
I would look at 230Vac creepage distances that are considered safe, instead of considering an adverticed 1kV MAX! rating as safe.
I'd also rather be safe than sorry, but in principle, you shouldn't get the nasty voltage peaks that can occur on 230 V mains when a big load is switched off or when lightning strikes a high-voltage line.
According to https://cirris.com/high-voltage-arc-gap-calculator
Max. arc distance is 0.05 mm for 500 V, or 0.15 mm for 1000 V
The relay is specified for use at up to 1 kV in the datasheet. The distance of the external relay pins is 3.6 mm. That's more than 50x more than the arc distance at 500 V. I may be on the wrong track, but I don't see why this relay would not be safe to use at this voltage.
A slightly dirty and moist PCB doesn't necessarily insulate as well as dry air. The usual creepage distance is 3 mm for single-insulated equipment for 230 V AC mains, but such equipment has insulation designed to handle 1.5 kV RMS for a short time because of short voltage peaks on the mains. Those should not occur in your circuit, I think.
I once had a major switching spike when i was hoovering the floor. Power came off and on again, killed instantly my hoover.
No idea how much overvoltage there was, but the power company knew about it and payed for the damage.
So, yes, domestic wiring will have to cope with much more, thats why wiring is rated for 1kVp and minimum required creepage distance 4mm.
But those where offcourse MIN ratings that could be relied on, and the relays given MAX ratings are in no way comparable to that, actually, they arent much good for anything, sexcept for ads.
Anyway, actually needed minimum creepage distance depends much on conditions, like moist, dust a.s.o., so the only thing that i can say is, more is better.
Given the fact, that the soldering pads need space also, a 2.54mm raster really does not leave much creepage between the contactpair, where one carries the positive voltage peak whilst the other the negative.
At least i would not feel comfy putting 2.54mm center spaced traces on a pcb and have them carry any kind of hv between them, would you?
No idea how much overvoltage there was, but the power company knew about it and payed for the damage.
So, yes, domestic wiring will have to cope with much more, thats why wiring is rated for 1kVp and minimum required creepage distance 4mm.
But those where offcourse MIN ratings that could be relied on, and the relays given MAX ratings are in no way comparable to that, actually, they arent much good for anything, sexcept for ads.
Anyway, actually needed minimum creepage distance depends much on conditions, like moist, dust a.s.o., so the only thing that i can say is, more is better.
Given the fact, that the soldering pads need space also, a 2.54mm raster really does not leave much creepage between the contactpair, where one carries the positive voltage peak whilst the other the negative.
At least i would not feel comfy putting 2.54mm center spaced traces on a pcb and have them carry any kind of hv between them, would you?
Yes, pcb traces can be dealt with, but you have still to solder the relay pin to the board.
In industrial envirement I used 0.1mm/V uncoated creepage surface
But under more relaxed domestic conditions half of that is ok, confirms also quite close to IPC2221B.
Google for pcb clearance calculator for high voltage 😉
In industrial envirement I used 0.1mm/V uncoated creepage surface
But under more relaxed domestic conditions half of that is ok, confirms also quite close to IPC2221B.
Google for pcb clearance calculator for high voltage 😉
Last edited:
Sure, slits (airgaps) are a very good idea, if they are easy to implement, as in massproduction.
For a diy-pcb, putting the solderpads off center, and maybe drilling a hole in between, could work to get 3mm creepage between the relay terminations.
A 5mm, or more, pcb trace seperation should not be much of a problem i would thinck.
So, with some tricks, 2 relays would be enough
B.t.w., correction to #91, not 0.1mm/V, that would be silly, it should read 0.01mm/V instead.
My bad
For a diy-pcb, putting the solderpads off center, and maybe drilling a hole in between, could work to get 3mm creepage between the relay terminations.
A 5mm, or more, pcb trace seperation should not be much of a problem i would thinck.
So, with some tricks, 2 relays would be enough
B.t.w., correction to #91, not 0.1mm/V, that would be silly, it should read 0.01mm/V instead.
My bad
Last edited:
A metal screen between your ear and the driver, connected to both L & R, so no voltage across your head during a fault would be a good start. Depending on this screen being earthed is not acceptable in many countries. Headphones are like bathroom circuits - jumping out of the way is not easy, so you end up with 10 mA limits
Let's not mess with the headphones. I am assuming the headphones are properly produced and can be assumed to be safe product.
The goal is to put a limit to any potential fault current from the amplifier into the headphones (or whatever gets connected to the amp output), and to entirely stop the fault current from the amp within a reasonable delay.
It would be cool if there's a way to come up with a little add-on board that could be retrofitted to existing amps.
According to IPC2221B up to 500V 1.5mm, A6 = uncovered external component lead/terminations
Above 500V 0,0035mm/V, but that would be 3,5mm for 1kV.
So i recommend the max you can practically get with those relays, something around 3mm, yes.
Between uncovered traces IPC2221B up to 500V 2.5mm, B2 = uncovered traces
Above 500V 0,005mm/V, so 5mm for 1kV
When I was a kid, electric fires had visible heating elements and a very coarse protective grille. Interpretation of safety rules has got stricter over the years.
A pair of headphones probably never gets a safety inspection until an accident happens.
How about this (pins 13 and 23 are the relay contacts to the bipolar amplifier outputs, A1 and A2 are the relay coil):
Do you really think that Sennheiser, Stax, HiFi Man, etc. bring their headphones into circulation without going through the relevant safety assessments?
Yes, those old fires had no thermal cutout either, so many lives and buildings were lost to them... The legal protections for electrical/electronic machine safety and so forth are very much different these days, in fact its a major cost in manufacturing to comply with regulations and standards. Far fewer people die from preventable incidents. I am ignoring knock-off / counterfeit equipment though - that stuff can be shonky as anything.