Relay voltage and noise

[defaced]

I have a PT that was originally used in a hybrid SS/tube amp. It has a 20-0-20 winding that I'd like to use for a relay power supply. I'm thinking that if I take half of the winding, rectify, filter, and regulate it, that I could use 24v DC relays for switching in my amp instead of the heater supply. My question is would 24v DC relays possibly introduce more noise into my amp than 5v DC relays, or because it's DC is doesn't make a whole lot of difference, or it won't make any difference because (assuming the same brand/type of relay), that the magnetic field will be the same because I and n are proportional (assuming I remember Physics)? I'm specifically thinking in terms of relay coil to signal coupling; we'll assume lead dressing isn't an issue.

[Enzo]

Yoiur coil isn;t going to make much of a field, and DC doesn;t radiate itself much. If you share a ground return with the relay coil currents and the preamp circuits, the relay coil ripple might add some noise. SO don;t share grounds. Physics? It takes the same magnetic energy to pull in the relay armature no matter the voltage. Higher voltage coil will draw less current. Same magnetic field. The relay design won't intentionally waste much energy spreading magnetic fields all over anyway, they will tend to mainly be in the coil and the pole running through it.

Make sure to snub the coil with a diode and maybe a cap. Switching noise would be a more likely noise source than the coil.

How many relays you got? You can run pairs of 12v coils in series off 24v, or single coils plus a resistor. You can get relays in many voltages, the little pc mount types seem most common in 12v. Of course 5v and 24v are also available readily.

So you have a 20-0-20. Instead of half a winding, why not use two diodes and go full wave from the whole thing just as we do in a Fender high voltage supply? Either way you are rectifying 20v. Yields about 28VDC.

I wonder about the need to regulate the relay supply. Seems like an unecessary complication. A 12v relay is happy as a clam on 9-15v most of the time. So unless this system uses a ton of relays and the current draw off the supply is SO variable we don't know what it will settle to, just run unregulated. So it is roughly 28v instead of 24v. Add a resistor. If the voltage is way too low, the coil might not pull in the armature, but that is not going to happen when the voltage is off by one volt. Low voltage might slow the thing a hair, but the timing difference would be a laboratory measurement, not something you'd hear. And if you are a volt high, the thing will get a trifle hotter than before. These things are very flexible.

[Don Moose]

It's also worthwhile to make sure all the switching contacts (NO, NC, COM) are AC-coupled to their sources/targets and that they have a DC path to ground. A big resistor (22M) will do if there isn't already a pot or other path.

Hope this helps!

[defaced]

The thought behind regulation was that it'd help smooth out the voltage but that's an error in my thinking. The only thing I care is if the relay switches on and off and doesn't add EMF noise to the amp.

I've never used a full wave rectifier, just a bridge, what does that do to voltage and current compared to a full wave?

I'm currently running three SPDT relays and am going to simplify that to a single 3PDT when I transition to a real board (the amp is prototyped on wood and nails). I may decide to relay the power and stand by switch to simplify lead dressing and to remove 500v DC from the stand by switch. Haven't decided yet. But even then, it's still easy to design around because the only variable current draw is the 3PDT relay.

[Enzo]

Grounding a center tap and running a rectifier from each end of the winding IS full wave. It does it with two diodes instead of four.

[defaced]

I screwed that up. What I meant is compared V and I for full wave vs. bridge rectifier.

Here's what I'm thinking:
If full wave of a 20-0-20 gives me 28 volts after rectify and filter, and I would expect 28 volts a 20-0 winding after bridge rectifying and filtering, then I would expect to have more current available from the full wave rectified 20-0-20 winding, correct?

[Enzo]

You would have the winding's rated current either way.