advertisement


Automatic line level detection

B

Bemused

What's this all about then?
Wondering about automatic line level detection, as for instance automatically switching on power amps when the a source is playing music, and off when not.
Balanced audio signal, yes the amps are nCore :D
Not interested in conventional 12 volt triggers right now ad would like automatic detection.
Any pointers gents?
 
Thel Audio does sell one of these:
http://www.thel-audioworld.de/module/nm25/nm25.htm

It is a softstart module which can be configured as
1) conventional softstart with hardware switch
2) standby module with relay switch
3) standby module with signal sensing

Option number three is maybe what you are looking for.
I built two hackernaps with those modules and it works quite good.
The only micro issue might be, that it needs maybe a few seconds until switch-on if the music is on low volumes.
 
Thanks it looks a nice bit of kit but overkill for just the part of it I would need.
As the nCores are not really DIY I thought to make up for it by making a automatic amp switch, unless a real cheap kit is on the market.

Could an op-amp sitting across the xlr signal not be the basis of such a bit of kit?

As a starter, and I am woolly here, I think 0dB balanced is 4V so -50dB as a switching threshold is around 10mV, so I would need an op-amp to detect 10mV differential and swing its output to a rail and enable a driver stage?

Then make the driver stage latching in the on state with some delay to knock the latch off.

Looking here I then noticed the zero level detector just below and that looks a better idea.

Detect a zero signal, run a timer, then shut the amp down. The timer would reset whenever the level rises above say 50dB.
 
Bemused said:
Thanks it looks a nice bit of kit but overkill for just the part of it I would need.
As the nCores are not really DIY I thought to make up for it by making a automatic amp switch, unless a real cheap kit is on the market.

Could an op-amp sitting across the xlr signal not be the basis of such a bit of kit?

As a starter, and I am woolly here, I think 0dB balanced is 4V so -50dB as a switching threshold is around 10mV, so I would need an op-amp to detect 10mV differential and swing its output to a rail and enable a driver stage?

Then make the driver stage latching in the on state with some delay to knock the latch off.
Click to expand...

There is a bit more to it than that - the signal crosses through zero.

My guess would be something like:
* Op -amp wired up as differential amp to balanced input, with some extra gain, converts to single ended.
* Active full wave rectifier circuit, so both halves of wave form are detected.
* Threshold circuit
* 1 minute or so latch (say 555). These can often drive a relay directly.

You then need to arrange permanently on stand-by power for this detector circuit - can't be power by main supply, otherwise it would never start.
 
Thanks PD.
Yes I would be building a small standby supply to run the detector.
Not thought of the zero crossing, maybe rectification feeding a capacitor if it was single ended, which it isn't.

I think I edited the post whilst you were posting, my edit considered actually detecting the signal going low rather than high which gets around any zero crossing problems?

Op-amp detects low signal,
On low it enables a 555,
The 555 times out and shuts down the power amp.
Whenever the op-amp detects an amp on signal it resets the 555 and hence the delay to shut down timer resets.
At listening levels the 555 gets triggered every zero crossing but the 555 can not time out and do an amp shutdown.
 
Actually the SMPS in my amps have a standby function which will make life a little easier.

I just need to lift its standby input to between 3v3 and 12v0 and it drops to standby so I wont need to bother with relays, just needing a delayed logic output from a zero detector?
 
Detect a zero signal, run a timer, then shut the amp down
Click to expand...

If i understand you right, can´t you detect a signal instead, charge a capacitor during signal detecting, if the signal is off: capacitor discharges (timer function) and finally switches a relay?

Thinking about it twice, will the amp switch on again?
 
Your input is an electronically balanced line. If you use a pair of comparators referenced to some adjustable voltage above 0v one or other will trigger each time the signal is above the threshold. So the -ve line will be positive for negative swings and the +ve line positive for positive swings.

You then OR these to reset a 555. Diodes might remove the need for logic. And removing the need for logic is a sine qua non of the DIY room and Bemused Buffalos.

Paul
 
Thanks all, I will order some 741s and 555s and do some bug build experiments.

Thinking about it twice, will the amp switch on again?
Click to expand...

I am detecting the signal feed to the power amps so yes. Would not switch on if I detected amp output though.
 
And removing the need for logic is a sine qua non of the DIY room and Bemused Buffalos.
Click to expand...

Quite true, I will try and not disapoint.
Seems there are many ways to achieve auto shut down.
 
Back here again following the discovery of true energy costs for leaving stuff switched on.

Schematic for balanced line level detection and a power relay to shutdown the nCores.

BallancedlineAutoShutdownSchematic_zps6ee6e27d.png


I have a breadboard prototype thats more or less works. Just waiting on a FET for the output stage.

IMG_1985_zps73bb7733.jpg


Once the prototype works I have a board layout ready and waiting.
Its all a little rushed and scruffy but I think the board is OK.

BallancedlineAutoShutdownBoard_zps60135596.png
 
Something like a TL061 has a lower supply current (250uA), and will still work fine here. As this circuit is always on, it is worth making it low power.
 
Lovely, I was also wondering how to use the standby switch on my UCD SMPS's. I guess this would work there the same as for your ncore psu's. I'd like to order one please :)

Stefan
 
Many thanks as always PD I will swap to TL061s.

The three op-amps are Bal / Se, x100 gain and comparitor but no need to tell you that :D

I am thinking of changing the gain of the Bal / Se to x100 and deleting the existing gain stage.

Is this a wise thing to do?
 
orangeart said:
Lovely, I was also wondering how to use the standby switch on my UCD SMPS's. I guess this would work there the same as for your ncore psu's. I'd like to order one please :)

Stefan
Click to expand...

The SMP600s in my nCore have a standby function selected by pulling up a control pin. (applying a voltage to enter standby) Pulling the standby pin up was my first thought but that just adds the standby power of the SMPS to the power of this circuit. So I opted for a cleaner and simpler aproach of just switching the SMPS primary supply off.

To use standby the output mode would need inverting but again there is no point in having two standby circuits.

Also this is designed to detect a balanced signal and would need bodging for use with single ended so for the odd occasion when I might take the nCores for a visit to a friend with single ended kit I added the pair of terminals to permanently energise the output relay.

The relay is a Finder 40.54 series and rated at 8 amp AC1, both poles are in parallel so it's good for around 14 amp. I don't think the SMPS is a very inductive load but it might not last so long switching large conventional transformers.

The transformer on the board is perhaps a little large at 4 VA but it fits the board profile, in any case I lost the will to live when looking for decent smaller transformers that also had an Eagle part.

Should it actually work you would be most welcome to a board.
 
Ah I see. Sounds like a plan. I presume the signal passes through, any room for it to be degraded? I'm balanced so just the ticket.
 
No the signal piggy backs of the nCore / XLR or whatever and passes to the shutdown board and just ends there. Could not see the point in passing it through as it only extends the length in a noisy enviroment.
Just two terminals for the signal, hot and cold. Of course the signal could be looped around XLR / amp / Shutdown board in any manner to suit. Signal only extends into a small corner of the board and a quiet corner at that. The signal only sees a pair of 10K resistors and an op-amp diferential input, its buffered from there on in. So the signal is effectively open circuit at the board.
 
Did some more prototype trials last night and the sensitivity is all or nothing so I did a spice simulation and the penny dropped.
I have no biasing on the balanced to single ended converter, and its compounded by the fact that the op-amp running on a single rail needs has a minimum output of around 1v5.
Geez I can be dim at times :D
So I will have a little redesign and try to save an op-amp by using the BAL/SE stage with x100 gain and delete the gain stage.
Also had a look at board houses and the current size of 100x54mm is not economical so have made some changes to reduce size. Using 2 X 9V 2 x 9v secondary saves two rectifier diodes and reducing the transformer from 4VA to 2VA also helps there again I always knew the transformer was too large.
 
You must log in or register to reply here.


advertisement


Back
Top