Enable 12V trigger with audio signal

[jmbulg]

I would like to trigger the power-on of a SMPS based amplifier (Hypex SMPS+UCd) using the audio signal. I have seen that it is possible to swith on the SMPS module by applying a 3-12 V DC. Searching the forum I found only references to external 12 signals which could be used to power on the SMPS module.

But is there a simple way, or commercial module, which takes the audio signal (coming from a preamp with possibly low voltage at low listening volume) and when detecting a signal, provides the necessary max 12 V and, after some time without input, allows to switch off the SMPS?

(PS: I enjoyed this features on some commercial amps I used for central/surround channels in an AV system, only turning them on when using actual surround)

Thanks in advance

[AndrewT]

ESP shows a few "auto" switches triggered by audio.

[jmbulg]

Thanks,
Found Signal Detecting Auto Power-On Unit but it goes beyond my present skills in electronics.

I have in the meanwhile found Xantech Audio Sensor Module - SMAUD01 - Smarthome which can be used together with
Xantech Current Sensor Connecting Block - CSM1 - Smarthome, but I suspect it is rather expensive with regard to the simple thing I expect it to do ?

[Mooly]

I don't know of any commercial ones.

Easy to DIY one.
A low power FET opamp configured as mixer with 10meg or more input impedance to "detect" L/R. Use high gain so the signal clips to the rails and feed to a retriggerable delay so that power stays on for a predetermined time in absense of audio.

Can be elaborated on to provide opto isolation to trigger a PSU etc.

[jmbulg]

I guess I could also use Audio Signal Controlled Relay Switch Module (CPM011) adding a control on the maximum input voltage (rectifier+zender ?) with the relay switching on/off the 12 V controlling the amp.
As the module itself needs 12 V all I would need to add is a sufficient stable 12V DC source.
Does this sound ok for you ? Or did I miss something (i'm a real novice )

[Mooly]

The pdf on that link leaves a lot of question marks (literally ) For example, input impedance 1k ????

If they don't know who does...

Yes in theory it's OK but if it does have 1k input impedance then that is too low really. You don't want that across a line level audio feed. Also how long does the output stay active when audio removed ? You don't want the SMPS going off between tracks or changing a CD. You need a delay... perhaps as long as 5 minutes... but only you can answer that.

[jmbulg]

The hypex can be put into standby by applying a voltage between 3 and 12 V
Does the following circuit ouline makes any sense (sorry, I’m a total novice).
The source voltage could be 12 V dc if there is an opamp which can work with it.
A few questions (if the first step, does it make sense, is passed) :
-are there opamps which present a sufficiently high impedance to the line in ?
-which kind of delayed relay should I look for (my badly drawn box on the right part)
-how can I be sure the line noise will not trigger the thing ?
-what kind of resistence I should put (needed because otherwise shortcut when signal)

[BobEllis]

You're close, but the refinements needed to make it work take you the Rod Elliot circuit you linked to. Issues are undefined off state of comparator, comparator can't deliver the current needed to drive a relay.

Question - you apply voltage to get the hypex into standby or remove a voltage?

A caveat with the ESP circuit - either drive it from a single channel or convert the input buffer to an inverting configuration. As drawn there is a lot of crosstalk between the channels.

[jmbulg]

Originally Posted by BobEllis

You're close, but the refinements needed to make it work take you the Rod Elliot circuit you linked to. Issues are undefined off state of comparator, comparator can't deliver the current needed to drive a relay.

Question - you apply voltage to get the hypex into standby or remove a voltage?

A caveat with the ESP circuit - either drive it from a single channel or convert the input buffer to an inverting configuration. As drawn there is a lot of crosstalk between the channels.

The voltage applied to the hypex gets it into standby. Simply not connecting the cable means he will be on when the main power is on.

In my case I plan a mono installation, so single channel in any case.

The problem of the comparator not able to deliver the current to the relay, is this always the case or can it avoided by selecting matching opamps and relays ?

If the latter is possible and with the mono system, could the ESP circuit be simplified to an extend where I can try, or conversely can my own naive tentative be corrected without too much complications?

Thanks already for all the feedbacks.

[BobEllis]

Do you know how much current the module needs to be driven into standby? You might be able to drive the standby pin directly from a comparator, if you make it go high when no audio is present. You'd want to add a bit of delay to hold it on and prevent amp cycling during quiet moments. Although if the ESP circuit scares you we probably shouldn't be talking about morphing it into something else.

The issue with the comparator driving a relay directly in this circuit is twofold. One, a typical NE5532 can deliver ~35 mA but the output current could be as low as 10 mA. That may not be enough to reliably switch your relay. Secondly, without D1, C4, R11 in the ESP circuit the relay would cycle when the signal drops below a certain threshold.

The ESP circuit works fine with only one input connected. I built it, then clipped one input when I discovered the crosstalk. Works as intended.

I can't see it being simplified and working the way you want. U1A and surrounding parts amplify the signal so that it is high enough to operate U1B's comparator even at low levels. The comparator is the switch. When it goes high, it charges C4 and turns on the mosfet to trigger the relay. C4 and R11 form a hold on delay so that the relay stays tripped when the signal falls below the threshold for a moment. Without it the relay would operate with every cycle of the signal. D1 prevents C4 from discharging through U1B when its output is low.