Hi!
sorry for yet another...
My toroidy 2 x 18V 500VA
has a earth-wire (yellow-green) accompanying the 2 primaries.
Looking through the build-guides as well as sound-au's Power Supply Wiring Guidelines doesn't show this detail.
Do I connect it to the power-connection, to the chassis before the cl60, or after (together with the PSU's ground?)
My instinct tells me to hook it up as soon after the power-inlet/switch/fuse as possible...
thanks for confirmation!
sorry for yet another...
My toroidy 2 x 18V 500VA
has a earth-wire (yellow-green) accompanying the 2 primaries.
Looking through the build-guides as well as sound-au's Power Supply Wiring Guidelines doesn't show this detail.
Do I connect it to the power-connection, to the chassis before the cl60, or after (together with the PSU's ground?)
My instinct tells me to hook it up as soon after the power-inlet/switch/fuse as possible...
thanks for confirmation!
Attachments
Last edited:
The wire is likely a shield between primary and secondary, like the purple wire in the diagram for this transformer:
AS-3224 - 300VA 24V Transformer - AnTek Products Corp
Connect the wire directly to earth ground.
AS-3224 - 300VA 24V Transformer - AnTek Products Corp
Connect the wire directly to earth ground.
As the purpose of a shield winding is minimized primary-secondary coupling, it generally connects to secondary-side ground so it can form an electrostatic shield for that. As that in turn usually connects to chassis and chassis would connect to PE in an IEC Class I device, that makes all of the above just about equivalent.
That said, the shield winding is primarily of importance in an IEC Class II application where it greatly reduces mains leakage current - with big transformers leakage capacitance can reach around 1 nF or more otherwise, a shield tends to reduce that by more than an order of magnitude. If you can just dump leakage current into PE the shield winding is basically wasted.
How critical this is will depend heavily on the kind of inputs used and how they are implemented. For a traditional hi-fi amplifier with straightforward unbalanced inputs, going Class II is highly recommended. (If gilding the lily is your thing, you can additionally bring in PE just to connect a smallish Y2 rated capacitor from chassis to that - with a shield winding, this could be as small as 1-2.2 nF while still reducing any remaining mains leakage substantially. PE should probably be treated as just another mains wire in terms of insulation then.)
If you want balanced inputs, going IEC Class I with AES48-2005 compliant input wiring is generally encouraged.
Now if you want both, then it gets more tricky, with multiple possible options...
That said, the shield winding is primarily of importance in an IEC Class II application where it greatly reduces mains leakage current - with big transformers leakage capacitance can reach around 1 nF or more otherwise, a shield tends to reduce that by more than an order of magnitude. If you can just dump leakage current into PE the shield winding is basically wasted.
How critical this is will depend heavily on the kind of inputs used and how they are implemented. For a traditional hi-fi amplifier with straightforward unbalanced inputs, going Class II is highly recommended. (If gilding the lily is your thing, you can additionally bring in PE just to connect a smallish Y2 rated capacitor from chassis to that - with a shield winding, this could be as small as 1-2.2 nF while still reducing any remaining mains leakage substantially. PE should probably be treated as just another mains wire in terms of insulation then.)
If you want balanced inputs, going IEC Class I with AES48-2005 compliant input wiring is generally encouraged.
Now if you want both, then it gets more tricky, with multiple possible options...
I'd connect the chassis as well as the xformr's shield winding with wires as short as possible to the PE lug, primarily for safety reasons.
Best regards!
+1 on Ben Mah's post.
However a quote from a reddit article a few years back:
"In general, you always want the fuse to be the first thing. You want the fuse to be able to catch any faults. Every wire crimp, every switch's possibly bent lug approaching chassis ground, etc., is a potential fault that that fuse should be first in line to handle.
level 2bradn
2 points ·
Yeah this is definitely a good policy. Even if there's no reasonable chance of something before the fuse shorting out within the scope of the device itself, someone could have opened it up and left some loose metal inside. So yeah, putting the fuse first is really hard to argue against.
level 1bradn
2 points ·
If it's just a fuse, then having it before or after the power switch shouldn't matter much (you might want it after for the 2nd reason in your list - the fuse doesn't handle any faults in the switch, assuming the switch doesn't have the other leg of the AC wired to it). Of course, power outlets are often wired wrong, so don't rely on the power switch cutting off the hot connection.
But, if there are other power protection components, you may want them (and the fuse) before the switch - that way the switch isn't part of a high current event (eg, let's say you put some MOVs or TVS diodes at the input to clamp overvoltage - they might conduct a fair bit of current during one of those events, and you might not want the switch being part of that). But if you do use those things, put them after the fuse so that the fuse has a chance of blowing before your other hardwired components."
P.S. All commercially sold fused equipment clearly states that "Ensure to disconnect the equipment from the wall socket prior to replacing fuses!" or words to that effect. All the equipment I have ever purchased always had the fuse installed directly after the power inlet, whether captive or with an IEC or similar socket.
However a quote from a reddit article a few years back:
"In general, you always want the fuse to be the first thing. You want the fuse to be able to catch any faults. Every wire crimp, every switch's possibly bent lug approaching chassis ground, etc., is a potential fault that that fuse should be first in line to handle.
level 2bradn
2 points ·
Yeah this is definitely a good policy. Even if there's no reasonable chance of something before the fuse shorting out within the scope of the device itself, someone could have opened it up and left some loose metal inside. So yeah, putting the fuse first is really hard to argue against.
level 1bradn
2 points ·
If it's just a fuse, then having it before or after the power switch shouldn't matter much (you might want it after for the 2nd reason in your list - the fuse doesn't handle any faults in the switch, assuming the switch doesn't have the other leg of the AC wired to it). Of course, power outlets are often wired wrong, so don't rely on the power switch cutting off the hot connection.
But, if there are other power protection components, you may want them (and the fuse) before the switch - that way the switch isn't part of a high current event (eg, let's say you put some MOVs or TVS diodes at the input to clamp overvoltage - they might conduct a fair bit of current during one of those events, and you might not want the switch being part of that). But if you do use those things, put them after the fuse so that the fuse has a chance of blowing before your other hardwired components."
P.S. All commercially sold fused equipment clearly states that "Ensure to disconnect the equipment from the wall socket prior to replacing fuses!" or words to that effect. All the equipment I have ever purchased always had the fuse installed directly after the power inlet, whether captive or with an IEC or similar socket.
Last edited:
