LM358 circuit help needed ...

Thread Starter

daba1955

Joined Apr 27, 2019
84
Hi guys, new here, and very much a digital guy, not much experience on analog circuits.

OK, so here goes with my problem...

I have a DMX stage lighting controller (far-eastern brand) that has a microphone built-in so that you can sync your chases to sound. However it is WAY too sensitive, and I want to slug it down, and preferably make it adjustable.

The circuit uses an EPE microphone, feeding pin 6 of the LM358, and there's an RC network from there to pin 5. In addition, there's pads for a VR off pin5 going somewhere under the LM358, but they've put a fixed resistor of 47K. The photos I've attached should explain it more, and I'm sure some of you will have seen this type of arrangement many times. I'm assuming the output of the LM358 is used to trigger the chase change.

So the question is, If I replace that 47K with a 50K pot, (I'll use a 20-turn), will I be able to adjust the sensitivity downwards, and take control of the gain.

2019-04-27_135902.jpg
 

Attachments

Thread Starter

daba1955

Joined Apr 27, 2019
84
It wasn't easy, but I've traced the whole circuit out to the controller processor.
The 50K pot shown is a fixed 47K resistor on the board, but that is how the pot would be if it were fitted.

2019-04-28 05.48.23.jpg 2019-04-28 05.48.23.jpg
 

Thread Starter

daba1955

Joined Apr 27, 2019
84
I'm thinking now that the fixed 47K (where the pot is) may be "mid-range", so considering a 100K pot would be a better choice...
 

crutschow

Joined Mar 14, 2008
23,506
That circuit can't be right as the capacitor blocks the op amp input bias current so it would saturate at one of the rails.
Are you sure the 50k feedback resistor isn't connected directly to the (-) input along with the input resistor and capacitor?
 

Thread Starter

daba1955

Joined Apr 27, 2019
84
You are correct crutschow, I goofed...

Here is the correct circuit ...

CORRECTION : That 10K should be 51K, don't know what the capacitor value is ...
 

Attachments

Last edited:

Thread Starter

daba1955

Joined Apr 27, 2019
84
I can't change the picture on the last post, so I've attached it here...

Am I correct in thinking that the feedback resistance should decrease to de-sensitise the op-amp ? In other words, more feedback, the harder you have to hit the input ?
 

Attachments

Thread Starter

daba1955

Joined Apr 27, 2019
84
One more question.....

Looking at the spec of the LM358, the output is capable of sourcing a minimum of 20mA, typical 40mA.

Can I put an LED on it as per the attached ? It will help when setting the level ....
 

Attachments

LesJones

Joined Jan 8, 2017
2,349
I think your schematic is still wrong. There is positive feedback that would probably make toe op amp latch up. I don't see how an LED on the op amp output will help setting the signal output level. All it will show is the DC level on the output pin.

Les.
 

Thread Starter

daba1955

Joined Apr 27, 2019
84
I think your schematic is still wrong. There is positive feedback that would probably make toe op amp latch up. I don't see how an LED on the op amp output will help setting the signal output level. All it will show is the DC level on the output pin.

Les.
Hi Les, thanks for your input.

Remember this is a Sound-to-Light circuit, and I'm guessing (because I can't get at the code in the processor) that it feeds a digital input to trigger chases. In that respect, I think it is intended to swing between the supply rails, so latching on and off would be intentional. Or maybe I've got it all wrong....

I'm going to try a 100K 20-turn pot, with the LED as well, and see how it goes. I can always cut the LED idea if it doesn't work. Just about to wire it to the board ....
 

Attachments

Thread Starter

daba1955

Joined Apr 27, 2019
84
Actually, even with it disassembled I can still power it up and see what the output voltage is sitting at, I suspect it will be at one of the rails, and the MIC input just makes it jump to the other for a short time on receipt of peak sound levels. If I were coding the controller I'd probably drive an interrupt off the input pin rising or falling to initiate chase changes - conjecture I know, but that's all it needs to do ....

EDIT : Sitting at +4.1V solid (power is +5V). Doesn't move with normal level sound, but a loud sound and it shoots down briefly ....
 

LesJones

Joined Jan 8, 2017
2,349
I was assuming the board was just a pre amp and the audio level detection was done in the unit that it connects to. From your test in post #12 it looks like your assumption is the correct one.

Les.
 

Thread Starter

daba1955

Joined Apr 27, 2019
84
I was assuming the board was just a pre amp and the audio level detection was done in the unit that it connects to. From your test in post #12 it looks like your assumption is the correct one.

Les.
As I said, the circuit just needs to detect high volume peaks and trigger a change, looking at the circuit, is this what is called a charge pump ?

I'll get it ready for a test, but will have to take the DMX controller to the venue as I have no lights here to test it with.

EDIT : Looks like I'm going to have to connect the LED to +V, not GND, otherwise it will be on all the time there's no sound, or the output pulse may be too short to see the LED, in which case I will ditch the idea ....

We shall see ....
 

Attachments

Thread Starter

daba1955

Joined Apr 27, 2019
84
Actually works !

I remembered I have a faulty light here, and used it just to test the idea.

However, I still think it is too sensitive (although greatly reduced) even with the pot fully clockwise, and that is minimum resistance, giving maximum feedback.

So what do I do now, increase, or decrease, the value of the 51K resistor between pins 5 and 6 ? I think increase, but would like it confirmed if poss.

And the LED is a waste of time, it obviously only goes off briefly, too fast to see....

EDIT : I've cut the LED off, and it is perfect now. I can set the sensitivity such that it needs a very loud sound to trigger the chase scene change.

YAY SUCCESS !! I'm happy to drill the controller front panel now to give access to the potentiometer slot...
 
Last edited:

Thread Starter

daba1955

Joined Apr 27, 2019
84
Ok gurus, well out my depth here.....

Sensitivity is well reduced, and adjustable as I wanted, but I still feel this circuit could be improved on.

It appears to be triggering off all sound, no matter what the frequency, and I would have thought it ought to trigger chases off bass notes only.

Anyone got any ideas how I can do this ? I'm prepared to remove the microphone from the board and use it to feed a low-pass pre-amplifier, so that it only gives output when bass notes are detected. There are loads of google images of low-pass filters using op-amps, anyone got any favourites, any tried and tested solutions I can work with.

I have 5V and 12V supplies available on the board, and the existing LM358 runs off 5V if that matters. There is tons of room inside the console for a sub-board ....

TIA
 

LesJones

Joined Jan 8, 2017
2,349
You could try putting a 10K pot in series with a capacitor in parallel with the microphone. I would suggest a capacitor between 100 nF and 1 uF to start with. (I do not know the nature of the sound that you are trying to detect to trigger the lights.) The capacitor is to avoid changing the DC conditions of the circuit. (A lower value of resistance will give more attenuation.)

Les.
 

Thread Starter

daba1955

Joined Apr 27, 2019
84
You could try putting a 10K pot in series with a capacitor in parallel with the microphone. I would suggest a capacitor between 100 nF and 1 uF to start with. (I do not know the nature of the sound that you are trying to detect to trigger the lights.) The capacitor is to avoid changing the DC conditions of the circuit. (A lower value of resistance will give more attenuation.)

Les.
The lights are used in a place where they have live bands about every two weeks. Mostly rock music, so we want the lights to respond to the bass drum mostly. I've even thought of putting a socket in the console for an external microphone we can put in close proximity to the bass drum. No doubt I'd have to totally kill the sensitivity of this circuit if we do that .... The existing circuit triggers the chase sequencer off just about any loud sound, no matter what the frequency, so the idea of a low-pass pre-amp ought to work, its just a question of making one work off a single supply. Does your suggestion lower the MIC performance at higher frequencies ?

Do you mean like this ?
 

Attachments

Thread Starter

daba1955

Joined Apr 27, 2019
84
The LM358 data sheet gives this as an example of a low-pass filter we could use as a pre-amp, but they don't give any detail on how to change the cut-off frequency f0 ....
 

Attachments

Top