I think I need a little help. AC IN switch circuit.

[mkiijam]

For the 1st time I'm deciding to power a device by having the front panel ON SWITCH actually just turn on the transformer to the power supplies. In the attached schematic I think I've done everything I know, but thoughts? Should there be any relay contact protection? etc?

[madires]

Is there any reason why you don't simply switch AC IN directly?

[mkiijam]

Yes.

Doing so would require the AC from the AC INLET on the back to travel up to the front panel switch and cross over (or be carried on the PCB) sensitive circuits.

[bostonman]

Why not just use a relay. Low voltage to the front panel and you can turn off the mains where it enters the unit.

[mkiijam]

Why not just use a relay. Low voltage to the front panel and you can turn off the mains where it enters the unit.

Ummm... that's what I'm doing? I think my circuit in the attached PDF schematic does that, yes?

[madires]

And the small 5V PSU is an SMPSU?

[tunk]

If you're using a push button, then you could place the
switch at the back and use a long rod to the front panel.

[bostonman]

Yes, I meant put the relay in the back where the mains are instead of the front.

[mkiijam]

And the small 5V PSU is an SMPSU?

Yes. A little "Stand by" idea.

If you're using a push button, then you could place the
switch at the back and use a long rod to the front panel.

Design style of the gear necessitates a ROCKER switch

Yes, I meant put the relay in the back where the mains are instead of the front.

Yes, it might not be clear put the AC in from the FUSED AC INLET drops right to that PCB. The 5V Standby is what runs up to the front panel to a ROCKER switch. The return of that switch activates the RELAY which is switching the AC MAINS to the transformer primary.

I'm also concerned about Flyback issues around the relay / transformer when the fields collapse.

[Benta]

If you're using a push button, then you could place the
switch at the back and use a long rod to the front panel.

That's what I'd do as well. Typical solution also seen in audio and T&M equipment.

[madires]

I don't know how sensitive the circuit is that you're concerned about running wires for mains AC to the front panel. Would the 5V SMPSU also create a problem for the circuit (long wires to the front panel)? Regarding the relay switching an inductive load it would be a good idea put something like a snubber on the load side.

[mkiijam]

I don't know how sensitive the circuit is that you're concerned about running wires for mains AC to the front panel. Would the 5V SMPSU also create a problem for the circuit (long wires to the front panel)? Regarding the relay switching an inductive load it would be a good idea put something like a snubber on the load side.

Pretty sensitive. The 5V will not be an issue.

Is the diode a good enough "snubber"?

[madires]

The SMPSU creates switching noise and comes usually with an EMI suppression cap between primary and secondary. The diode is meant as a flyback diode for the relay coil. But there's another switched inductive load: the transformer. In many cases switching the transformer works fine without a snubber (check the relay's rating for inductive loads).

[Zero999]

It will work of course.

I did a similar thing, but it so a device with volt free contacts which weren't mains rated to control a mains load. I used a transformer, bridge rectifier and filter capacitor, to power the relay, rather than a switched mode power supply. I don't remember whether I used the usual back-EMF diode, or switched the AC and allowed the bridge rectifier to act as the freewheeling diode.

[mkiijam]

I used a transformer, bridge rectifier and filter capacitor, to power the relay, rather than a switched mode power supply. I don't remember whether I used the usual back-EMF diode, or switched the AC and allowed the bridge rectifier to act as the freewheeling diode.

I thought about that too. I've seen a ton of those from back in the day working on TV's and Hi-Fi stereos. They are seemed to have little "standby" transformers. I honestly don't know which is better nowadays.

[Retirednerd2020]

Depending on your sensitive circuits, SMPS noise may be more prone to be problematic and more difficult to get rid of than 50/60 Hz. noise.  Not a slam dunk from a susceptibility stand-point, given the limited circuit details provided.  If you do have specific low frequency concerns, make sure you choose a good location and orientation for your transformer.  Toroid?  EI?  If toroid, there can be more inrush current to contend with depending on the design.

[Zero999]

I used a transformer, bridge rectifier and filter capacitor, to power the relay, rather than a switched mode power supply. I don't remember whether I used the usual back-EMF diode, or switched the AC and allowed the bridge rectifier to act as the freewheeling diode.

I thought about that too. I've seen a ton of those from back in the day working on TV's and Hi-Fi stereos. They are seemed to have little "standby" transformers. I honestly don't know which is better nowadays.

A mains transformer has a higher standby power consumption, but is fairly noise-free.

A modern switched mode power supply has a much lower standby power consumption, but is more noisy.

The device I built simply turned on a compressor, when a pressure calibrator demanded it, via a dry contact. This was to save power, noise and ware on the compressor running continuously. Standby power consumption wasn't an issue, since it isn't left running all the time. The whole thing fitted into a small 150mmx75mm*50mm diecast aluminium enclosure, with a UK 13A socket on the side and a cable to a 37-D connector to the pressure calibrator.

[Benta]

A few comments:
Using a 12 V relay with a 5 V supply is not a good idea.
Your "diode snubber" is not a good solution. It will delay release and increase contact burn. Use a 12...18 V bidirectional tranzorb instead.

How big is your toroid? If over 300 VA, you'll likely need soft start.

[mkiijam]

A few comments:
Using a 12 V relay with a 5 V supply is not a good idea.
Your "diode snubber" is not a good solution. It will delay release and increase contact burn. Use a 12...18 V bidirectional tranzorb instead.

How big is your toroid? If over 300 VA, you'll likely need soft start.

Oops! Yeah... it's the same footprint for all the SMPSU's but it'll need to be 12V or 5V for sure. The transformer is 50VA.

[NiHaoMike]

One trick is to have the switching supply powered across the relay contacts, so that it turns off once the relay is turned on. Use a diode from a switched rail to keep the relay on.

[mkiijam]

One trick is to have the switching supply powered across the relay contacts, so that it turns off once the relay is turned on. Use a diode from a switched rail to keep the relay on.

This sounds interesting and I'm trying to visualize it

[tooki]

I don't know how sensitive the circuit is that you're concerned about running wires for mains AC to the front panel. Would the 5V SMPSU also create a problem for the circuit (long wires to the front panel)? Regarding the relay switching an inductive load it would be a good idea put something like a snubber on the load side.

Pretty sensitive. The 5V will not be an issue.

If it’s that sensitive, I’d imagine having a SMPS anywhere nearby to be FAR, far worse than mains wires. The 5V isn’t the issue. It’s the multi-kHz noise superimposed on it, as well as radiated out from the SMPS itself.

Surely you can just route the mains wire to the front panel far enough away? (I mean, that’s how countless pieces of audio gear do it.) You could also shield the mains wiring if necessary.

Not to mention that if your circuit is that sensitive, maybe you should shield it itself, since it may encounter other sources of noise when out in the field.

[mkiijam]

Here is a picture that may help in understanding my dilemma. The main PCB covers the entire area inside the box and I really don't want AC running around or over the other circuits.

[MarkF]

Here is a picture that may help in understanding my dilemma. The main PCB covers the entire area inside the box and I really don't want AC running around or over the other circuits.

That's why they make these:



Route your wires along the frame away from sensitive components and tie them in place with self-stick mounting pads.

[mkiijam]

The entire area is covered by PCB with components on it. There is area around the "sides" of the PCB up to the front panel, but not along the back. The IEC inlet is in the center with AUDIO IO's on either side.