Power supply for input selector relay.

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
This is my first question in this forum, a question about relay input selector power supply. The preamp will have +15 and -15 rail for the chips. I found a schematics (attached) which seems to be OK for the +12V relay power, but I'm not really sure that if it's OK just to draw some current from +15V line.

Any ideas?
Thanks!
 

Attachments

  • PSU.jpg
    PSU.jpg
    183.7 KB · Views: 442
Relays produce a lot of nasty high frequency noise (sharp edges) that couples into everything. This is true even if you use a back diode. People that design pro things like packaging machines put the computer and instrumentation on +5, +-15 supplies, and use an entirely different supply with a different transformer for the relays. Then they put the devices in separate steel boxes, and separate the two with optoisolators to drive the relays. You should at least use a separate power transformer and rectifier for the relay. Don't see a relay in your schematic. People that design SS front ends for Leslie motorized speakers go through this all the time. Cheap causes problems. A cheap source of separate transformer for the op-amp circuits is wall transformers from the charity resale shop. A suplus HP printer 30V @ 4A wall transformer puts the nasty transformer outside your box. The relay and its 12 or 24v supply should have its own steel box. Don't know if the LM317 works okay without a center tap, I used two 1n5344 zener diodes producing a fake "ground' in the middle of 18 VDC for the four op amps in my disco mixer.
 
Last edited:
Administrator
Joined 2007
Paid Member
This is my first question in this forum, a question about relay input selector power supply. The preamp will have +15 and -15 rail for the chips. I found a schematics (attached) which seems to be OK for the +12V relay power, but I'm not really sure that if it's OK just to draw some current from +15V line.

Any ideas?
Thanks!

I think it will be fine... just be sensible how much current you draw (which applies no matter what it feeds). I assume you are talking about small signal relays.

Make sure the "zero" return for the relay coils goes back to the PSU and not anywhere that current spikes could contaminate a signal ground. In other words couple to TP14 as shown in your diagram.
 
You can use the supply as shown above. A 3V dropdown (15V to 12V) can simply be achieved by using a resistor (calculate or measure). It is only for 1 relay at a time I presume? An 7812 or 7912 can also be used to make a separate supply from the rectifier part.

What indianajo writes is true, but I would not be so concerned of noise switching for a small relay in the preamp.
 
Relays produce a lot of nasty high frequency noise (sharp edges) that couples into everything. This is true even if you use a back diode. People that design pro things like packaging machines put the computer and instrumentation on +5, +-15 supplies, and use an entirely different supply with a different transformer for the relays. Then they put the devices in separate steel boxes, and separate the two with optoisolators to drive the relays. You should at least use a separate power transformer and rectifier for the relay. Don't see a relay in your schematic. People that design SS front ends for Leslie motorized speakers go through this all the time. Cheap causes problems. A cheap source of separate transformer for the op-amp circuits is wall transformers from the charity resale shop. A suplus HP printer 30V @ 4A wall transformer puts the nasty transformer outside your box. The relay and its 12 or 24v supply should have its own steel box. Don't know if the LM317 works okay without a center tap, I used two 1n5344 zener diodes producing a fake "ground' in the middle of 18 VDC for the four op amps in my disco mixer.

I never heard this, but it might explain why I didn't like the relay attenuator I used for a while, although it only switched when changing the volume.
 
Thank you for the quick replies. I'm really surprised so many different opinions about this. I expected there is some kind of common way for the relay PSU, but it seem like there isn't.

The unit will have 5 relays daisy chained for balanced 5.1. Each relay will have a back diode.

Would it be better taking 30V directly from 2x15V transformer tap (1 and 4 in the picture) , then bridge and +24V regulator to drive 24V relay? The relay coil is not grounded, though. 24V relay seems to draw less current then 12V, but it may do not matter at all...

The unit will have 5V PSU with separate transformer and regulator for digital control chips, so I don't really want to see another transformer in the unit. I can order custom wound transformer (5V 12V 15Vx2), but I'm afraid if it would be overkill.
 
JimmyWilliams just went through this on a Leslie interface, see Op-Amp question Separate transformers means to me a transformer with a different core and different steel box for the nasty relays and fans. I'm building a SS organ power amp, preamp power (100 VDC @ 100 ma) with stop switching relays and fans. I'm putting the unshielded 100 VDC transformer and the 24VCT transformer in a PCAT box this week. The three surplus steel encased power amp transformers, and something to produce +-8V for op amps, will be going in an aluminum master chassis, since I have about 10X as many aluminum and stainless steel boxes as ferous steel. 24 VCT transformers are easy to come by in factory surplus, since it is the standard DC relay power supply for factory equipment. +5 +-15 packaged supplies like from Power1 are standard for factory instrumentation like scales, but usually leave the factory blown up, so I'm buying power transformers surplus. Not compact, but more so than the +450 VDC @ 2A power supply I'm replacing.
 
... let me ask a n evil, evil, really evil question:

Who the **** cares about switching noise "coupling into audio circuits" from a relay that switches only to select a different source? :scratch1:

To be honest, I do not really judge the sound quality of my setup in the instant it switches to another source.... :rolleyes:

Use a separate transformer if you have a small one available, use the existing +-15V if not.

Greetings,
Rundmaus

EDIT: I just read that you have a +5V supply for digital circuitry already planned. Why don't you use this for relays with a 5V coil, as they are readily available?
 
Last edited:
I just read that you have a +5V supply for digital circuitry already planned. Why don't you use this for relays with a 5V coil, as they are readily available?
I'm afraid that relay switching spike would make digital processor unstable. Also I'm trying to separate analogue and digital as much as possible, because I have never built digital and analogue circuit in one box kind of unit. Everything is so unpredictable for me...
 
Last edited:
Rundmaus may enjoy big "pow" noises out of his speakers when he changes inputs. I don't.He may enjoy turning off his power amp before he switches inputs on his preamp. I don't. When I am playing an organ piece, when I switch stops "POW" is not part of the music JSBach wrote. The Hammond H182 makes no noise when you switch manualy from "trumpet" to "theater solo 16". I hope to achieve the same silence when I change from "principal" to "mixture" stop by kicking a "piston" (momentary pushbutton) that actuates relays in the back on the preset wires.
Using 5V winding instead of the op amp winding as suggested for the relays may help inhibit switching transients on the audio. Agreed coupling in relays is non-standard on digital logic power supply.
Frankly, with the disco mixer RA88a, I've got all 3 inputs (LP, radio, CD) on at the same time with no audible increase in hiss or hum. But it took me a long time to get there, several modifications. No relays in it, no relay power supplies. But good Slider pots are kind of expensive, if I hadn't gotten them as a package with the mixer at the flea market. Other brands of disco mixer have notoriously noisy slider pots.
 
Last edited:
Rundmaus may enjoy big "pow" noises out of his speakers when he changes inputs. I don't.He may enjoy turning off his power amp before he switches inputs on his preamp. I don't. When I am playing an organ piece, when I switch stops "POW" is not part of the music JSBach wrote. The Hammond H182 makes no noise when you switch manualy from "trumpet" to "theater solo 16". I hope to achieve the same silence when I change from "principal" to "mixture" stop by kicking a "piston" (momentary pushbutton) that actuates relays in the back on the preset wires.
Using 5V winding instead of the op amp winding as suggested for the relays may help inhibit switching transients on the audio.

I have a DA converter that makes huge pop noises whenever relay latches, so I opened the box and checked circuit. What made me really surprised was the relay coil power is not regulated, they are connected to directly to the pre-regulator filter cap of analog power supply.
 
I was going to say, your original circuit with the LM317 regulators did not look like a relay power supply. Relay power supplies in packaging machines (both Keyclox/Woodman, Apex & Ishida) have a circuit breaker, lightning arrestor, transformer, rectifier, & filter capacitors. No regulators. Relays are so slow that a little 60 hz ripple doesn't bother them.
 
Last edited:
Ummm... just a thought, why not using simple differential amplifier as voltage follower which will have high output impedance when its common current become zero, or simply using amplifier with mute ability if you need simpler one but not as good as first one.

May I ask? is there will be remote control with it? If yes, is the volume also controlled with remote?
 
Hi guys, thank you for your suggestions. I decided to perform following tests and will post the result.

1. I have an Acopian 12V linear PSU somewhere in te basement, so I will test it first. This would be similar to indianajo's suggestion, not in the steel box, though.

2. I also will try the first method I posted (12V from +15 line, regulated) and compare it with Acopian method. .

3. Also I may try adding another bridge and draw it from the transformer directly.
 
Last edited:
Experiments are good. If you don't have a problem, why worry. Also wiring can suppress noise. GE Appliance Div. control panel standards required electromechanical wiring (motors, relays) to be bundled separately from control wiring (low voltage, sensitive) and no parallel runs. The two wire bundles were supposed to cross at 90 degrees. Similarly bundling fan & relay coil wiring separate from signal wiring would also limit crosstalk. In fighting off the hum in my disco mixer, I came to really appreciate the construction of the Dynakit PAS2, that has the signal wiring in it's own steel enclosure, separate from the power switch and AC. The power transformer was steel encased also, on all 3 dynaco devices I have. Ferrous steel is a useful noise control device against electromagnetism.
 
Hi there,

Rundmaus may enjoy big "pow" noises out of his speakers when he changes inputs. I don't.He may enjoy turning off his power amp before he switches inputs on his preamp. I don't.

If you really get a big "pow" on switching inputs, you are doing something seriously wrong on the signal side of the relay! Loud popping is definitely not caused by "noise" from the relay coil - and can therefore not be cured by making an overly complicated relay power supply. :cool:

I have build a few amplifiers with very simple relay input switching, consisting of a rotary switch, a few wires and good-quality small-signal relays. I never heard any audible "pow" from the speakers in years. :)

One thing is very important, see below.

Using 5V winding instead of the op amp winding as suggested for the relays may help inhibit switching transients on the audio. Agreed coupling in relays is non-standard on digital logic power supply.

My suggestion to use the existing +5V supply depends on one important prerequisite: You have to take care properly of the relay coil switch-off spikes, namely by a capacitor and a diode of opposite polarity parallel to the coil.

Forgetting about this will inevitably fry any digital or analog circuitry on the same supply by short HV spikes on switch-off , I have seen this kill LEDs which were to indicate the selected input.

Summarized, loud (meaning audible :D) switching "pops" from input relays indicate a problem on the signal side, not on the relay power side - if correct snubbing of the relay coil (see above) is provided. Noise induced from relay switching may be important in industrial applications where 10s oder 100s of relays fire with high frequency, but not in audio applications.

Best regards,
Rundmaus
 
As I said, I support the experiment planned. I don't argue with data. If all the "preamp" inputs are "line level", that is about 1.5 VAC, and output is also about 1.5 VAC, then relay switching may not be much of a problem. If one or more inputs are gain 50 magnetic phono cartridge, or tape head, or dynamic microphone, (10000 ohm source impedence), then pops will be more of a problem.
Back diode snubbing of industrial DC relays is so common they are built into the ones stocked by most industrial distributors. To get one without a back diode requires buying 100 from the factory, or from am amateurish electronics supplier. The rotary selector switch I have, the PAS2 preamp, shorts all the inputs except the one selected, which eliminates the chance of pops or stray emf from the unconnected input jacks. When I was a young technician I built a 2P relay switching circuit for one differential input channel of a data acquisition system, a TI DFS2 . It was supposed to switch between the "gun" geophone near the emmision device, and a field geophone that usually monitored the channel. It popped so loudly that the DFS would cut gain on all the channels reflexivly, ruining the data. I put a multi phase timer circuit on it to switch on a shorting FET (very high tech in 1975) , switch the relay to NO,FET off, delay for gun input, FET on, switch the relay back to NC, FET off. The FET's weren't low enough in on resistance to keep the DFS from halving the gain. With today's lower resistance FET's, that might work, but the signal has to be biased not to go below 0. IC Analog switches available now seem to bottom out at 3 ohms, which is not enough for my proposed Hammond organ circuit that emulates harmonic sliders that have 0.5 ohms between detents, so I'll probably be doing the shorting dance with my relay device. Hammond got away with their primitive "preset' switching, because their inputs are .5 ohm source or lower, going into mixing transformers, for high current low voltage switching that was insensitive to RF noise. I'm going to leave bare PCB on my control circuit to install the multiphase silencing circuit if I have to, but will not implement it in the beginning as an experiment in limiting effort.
 
Last edited:
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.