Hello,
I’ve a simple circuit for an instrument I’m building but I can’t seem to get it working correctly. Basically I’m sending an audio signal to a pre-amp; the signal is then pulsed by an analogue switch before being sent out to the power amplifier.
The problem I can’t overcome is the ‘clicks’ that come from the analogue switch. I’ve tried several different models; even special audio ‘clickless’ switches but it seems to make little difference. I have followed all the recommendations in the various datasheets but without success. Is the click due to the control signal or due to the switch turning on? The audio switches are supposed to turn on gradually to eliminate this noise but it comes out loud and clear on my system.
I have tried using a SPDT version to switch the amplifier input from signal to ground; the output was silent between pulses but the switch clicks were not eliminated.
I tried putting the analogue switch on the power amp speaker leads, but that only resulted in melting the IC. I would only suppose that if the chip was placed immediately after the pick-up that any click would be well amplified by the time it reached the speaker. Even without any audio signal through the circuit the switching clicks were still coming out loud and clear.
I’ve been using MAX4514 and SSM2412 switches recently (I have to use DIP packages, the others types are too small to handle).
So where am I going wrong? Is there such a thing as a ‘clickless’ analogue switch and if so how do I wire it up to keep the clicks at bay?
All suggestions welcome. Thanks.
I’ve a simple circuit for an instrument I’m building but I can’t seem to get it working correctly. Basically I’m sending an audio signal to a pre-amp; the signal is then pulsed by an analogue switch before being sent out to the power amplifier.
The problem I can’t overcome is the ‘clicks’ that come from the analogue switch. I’ve tried several different models; even special audio ‘clickless’ switches but it seems to make little difference. I have followed all the recommendations in the various datasheets but without success. Is the click due to the control signal or due to the switch turning on? The audio switches are supposed to turn on gradually to eliminate this noise but it comes out loud and clear on my system.
I have tried using a SPDT version to switch the amplifier input from signal to ground; the output was silent between pulses but the switch clicks were not eliminated.
I tried putting the analogue switch on the power amp speaker leads, but that only resulted in melting the IC. I would only suppose that if the chip was placed immediately after the pick-up that any click would be well amplified by the time it reached the speaker. Even without any audio signal through the circuit the switching clicks were still coming out loud and clear.
I’ve been using MAX4514 and SSM2412 switches recently (I have to use DIP packages, the others types are too small to handle).
So where am I going wrong? Is there such a thing as a ‘clickless’ analogue switch and if so how do I wire it up to keep the clicks at bay?
All suggestions welcome. Thanks.
There are a number of reasons for this. If there is any DC differential between the switch elements then you are going to hear that as a noise/click/thump when the switch closes/opens. So with the switch open there should be identical DC conditions on the in/out pins... to within microvolts really. So a lot depends on the way the circuit is implemented. You mention that your is noisy with no signal so check this.
When you switch an audio signal, unless it is done at a zero crossing point (which is impractical and difficult) then again there is an abrubt step change in "DC" conditions and you hear a noise.
Breakthrough of the control signal can be an issue. A lot depends on the specific devices and the circuit impedances as to how much of a problem this is.
When you switch an audio signal, unless it is done at a zero crossing point (which is impractical and difficult) then again there is an abrubt step change in "DC" conditions and you hear a noise.
Breakthrough of the control signal can be an issue. A lot depends on the specific devices and the circuit impedances as to how much of a problem this is.
If I tied both the input and output pins to ground with a capacitor would this be sufficient to eliminate the voltage differential that you mentioned? I suppose the only way to find out is to experiment.
As mentioned previously, the clicks are present even when there's no input signal connected to the analogue switch. I've tried reducing the voltage of the triggering signal and adding a decoupling cap which seems to have a slight effect but isn't the solution.
AndrewT, could you please explain a little more about your suggestion of the voltage controlled FET mute? Thanks.
I've attached a schematic of the section of the circuit involving the analogue switch; there could be a major error in the design that I haven't noticed (I'm an ME not an EE). Basically the signal from the preamp is split in two through a unity gain signal buffer/splitter; one signal passes directly to the main amp and the other is pulsed by the analogue switch. There's a panpot in the line out to control the two signals.
Vielle568
As mentioned previously, the clicks are present even when there's no input signal connected to the analogue switch. I've tried reducing the voltage of the triggering signal and adding a decoupling cap which seems to have a slight effect but isn't the solution.
AndrewT, could you please explain a little more about your suggestion of the voltage controlled FET mute? Thanks.
I've attached a schematic of the section of the circuit involving the analogue switch; there could be a major error in the design that I haven't noticed (I'm an ME not an EE). Basically the signal from the preamp is split in two through a unity gain signal buffer/splitter; one signal passes directly to the main amp and the other is pulsed by the analogue switch. There's a panpot in the line out to control the two signals.
Vielle568
Attachments
caps leak current when a bias exists between the plates.
If you use a DC blocking capacitor and leave the connection end unterminated you will find that the unterminated connection develops a voltage that approaches the voltage on the biased side of the capacitor.
Your switch could be behaving exactly like an unterminated connection. Over a short period of time the switch contact voltage changes from 0V ref signal ground.
Clicking the switch to the terminated connection allows that voltage to dissipate. That is your click.
Do not leave any connections unterminated.
If you use a DC blocking capacitor and leave the connection end unterminated you will find that the unterminated connection develops a voltage that approaches the voltage on the biased side of the capacitor.
Your switch could be behaving exactly like an unterminated connection. Over a short period of time the switch contact voltage changes from 0V ref signal ground.
Clicking the switch to the terminated connection allows that voltage to dissipate. That is your click.
Do not leave any connections unterminated.
I have found a solution; I don't know if it's a good solution but it works well enough for my needs.
I added two 104 caps going between ground and the analogue switch input and output pins. I can only assume that these act as decoupling caps and bring the DC voltage levels on these two pins at more or less the same level (0V) and as a result the 'click' from switching has now become almost inaudible.
Thanks everyone for your help and suggestions. They were greatly appreciated.
Vielle568
I added two 104 caps going between ground and the analogue switch input and output pins. I can only assume that these act as decoupling caps and bring the DC voltage levels on these two pins at more or less the same level (0V) and as a result the 'click' from switching has now become almost inaudible.
Thanks everyone for your help and suggestions. They were greatly appreciated.
Vielle568
That's how I have connected it up; with a cap to ground on the audio signal line either side of the analogue switch. Initially I tried 1uF caps but there was an obvious attenuation in the signal getting through to the amplifier. The caps were then reduced to 100nF (0.0000001F) and the signal recovered its strength and the clicks from the analogue switch remained almost inaudible.
Am I effectively shorting out the audio signal by adding these caps? I'd like to know what I have done incorrectly. It seems to be working OK at the moment.
Vielle568
Am I effectively shorting out the audio signal by adding these caps? I'd like to know what I have done incorrectly. It seems to be working OK at the moment.
Vielle568
Last edited:
Yes, you are shorting the audio with these caps. At 10Khz the 0.1uF cap has a reactance (or "resistance") of only 160Ω or so. Add another on the other side of the switch and it's not good.
You need to find the reason why it clicks.
With no signal you should have 0.000 volts DC on top of that 470K resistor. With the switch open you should also read 0.000 volts DC on the output of the switch even though it is "open". If those conditions are not met then there will be a click when the switch operates.
If you have the patience to read through all this then it might help... similar problem on a Rotel amp.
http://www.diyaudio.com/forums/solid-state/202909-help-dc-volume-pot.html
You need to find the reason why it clicks.
With no signal you should have 0.000 volts DC on top of that 470K resistor. With the switch open you should also read 0.000 volts DC on the output of the switch even though it is "open". If those conditions are not met then there will be a click when the switch operates.
If you have the patience to read through all this then it might help... similar problem on a Rotel amp.
http://www.diyaudio.com/forums/solid-state/202909-help-dc-volume-pot.html
Thank you Mooly for your explanation. I'll take the time and go through the information you have provided, however, if the reactance of the 0.1uF cap is only 160 ohms at 10kHz then would a lower value cap have a higher reactance? Or perhaps this is just not the way to solve the problem.
I'll go through the data you sent me and see what I can figure out. Thanks for your help with this project.
Vielle568
I'll go through the data you sent me and see what I can figure out. Thanks for your help with this project.
Vielle568
Connecting a cap directly across an audio output is normally a "no no".
The major concern with connecting a small capacitive load across an output is that it can cause instability and oscillation. Normally that's a problem even for low values of C. Some opamps are oscillation prone even driving the capacitance of a few metres of cable which would be in the 10's to 100's of Pf. Many opamp data sheets show the recommended method of driving capacitive loads which usually involves a low value resistor after the opamp output. But that's for small low values of C
0.1uf to an opamp output is just swamping it. If you had a scope you could see for yourself what it really does. The opamp would swing less and less voltage at the output as the frequency increased and the transient performance would be non existent.
The major concern with connecting a small capacitive load across an output is that it can cause instability and oscillation. Normally that's a problem even for low values of C. Some opamps are oscillation prone even driving the capacitance of a few metres of cable which would be in the 10's to 100's of Pf. Many opamp data sheets show the recommended method of driving capacitive loads which usually involves a low value resistor after the opamp output. But that's for small low values of C
0.1uf to an opamp output is just swamping it. If you had a scope you could see for yourself what it really does. The opamp would swing less and less voltage at the output as the frequency increased and the transient performance would be non existent.
Ok, others has already said what I would have been saying. my suggestion would be, remove the switch IC. physically. replace it with a tactile switch (mechanical switch). connect 10M resistor parallel to the switch. Now try switching it (manually).
If this works probably we can use different switching ICs. do listen it properly when switch is in off position and the click noise too.
OK Mooly, I understand what you're saying and I'll take the caps out. I've tried out the circuit a few times and the output signal doesn't have the loud clicks but there appears to be a mild effect like a phaser that is detectable. I do have a scope but it isn't capable of doing spectrum analysis so I'll take your word for the effects on my signal output.
I'll read through that info you left me yesterday and see what I can learn from that.
Meanwhile, Acosticraft has suggested a 10M resistor in parallel with the switch. I assume that this is to suppress the audio signal from passing but to allow the DC levels either side to be at the same level. Do you have any comments on this approach Mooly? It certainly won't take long to wire it up and see what happens. I'll let you know....
Vielle568
I'll read through that info you left me yesterday and see what I can learn from that.
Meanwhile, Acosticraft has suggested a 10M resistor in parallel with the switch. I assume that this is to suppress the audio signal from passing but to allow the DC levels either side to be at the same level. Do you have any comments on this approach Mooly? It certainly won't take long to wire it up and see what happens. I'll let you know....
Vielle568
You would see the effect on your scope if you looked at say a 10 or 20khz sine wave (square would show it even better) and then just dabbed the cap across the output. It will reduce the amplitude dramatically as the opamp current limits. Try it at say 10 volts output
If you haven't a signal generator then a CD with test tones is good too.
The 10M across the switch ! I wouldn't like to say really. It shouldn't make any difference, the only real effect being a tiny amount of signal being allowed to pass. It all depends on what the switch actually feeds into (impedance wise) and what the static conditions of that input are.
If you haven't a signal generator then a CD with test tones is good too.
The 10M across the switch ! I wouldn't like to say really. It shouldn't make any difference, the only real effect being a tiny amount of signal being allowed to pass. It all depends on what the switch actually feeds into (impedance wise) and what the static conditions of that input are.
This shows an opamp driving a 0.2uf cap at 10 khz. The green trace is the input signal and is also how the output of the opamp should be driving any normal load within its limits. The blue trace shows the actual output of the opamp. You can see how distorted the signal is as the opamp can not drive the cap.
Although the opamp here is a different type it is representative of what happens.
Although the opamp here is a different type it is representative of what happens.
Attachments
Have you tried to connect a resistor from the output of the switch to ground? This may help to keep the same DC level on both sides of the switch. Perhaps a 470kohm or perhaps smaller. Since we don't know exactly what is connected to the output of the switch is is difficult to give exact guidance.
Maybe replace the switch with a jumper. When you open and close the jumper, do you still have the click? That should narrow things down. If you don't. try softening the edge of the switch pulse with a small resistor and cap, maybe 50 ohms in series and .01uF to ground. Also, disconnect the switch pulse and pulse it to be sure you don't have a cross coupling issue elsewhere, like through ground.
I can understand what you're telling me Mooly. No, I don't have a signal generator but I can see clearly from your signal plot the deformation you've been describing. Those caps have now been removed from my circuit.
I did try putting the 10M resistor across the switch terminals. I used a mechanical switch initially as Aucosticraft suggested. There was no audible click when the switch was toggled; just a change in hiss (probably due to all the temporary cables wiring up the circuit; nothing shielded at the moment). However, when the analogue switch was put back in place the click was still audible, especially at low signal levels.
In reply to JensH the analogue switches in this circuit are placed after a preamp and signal splitter. They are controlled by a microprocessor; the pulse output passes through a Schmitt trigger inverting buffer to ensure a clean signal to control the switch. After the switch the pulsed audio signal passes to an output connector and is then cabled to a mixing table and power amp. There are three similar circuits in this design that run in parallel. And yes, I have actually tried connecting resistors of various values between the switch terminals and ground, but it didn't seem to have much effect on the clicks being generated.
In reply to Conrad Hoffman I have just replaced the analogue switch with a simple push type switch and tried pulsing the signal manually. The output is actually very clean and there are hardly any clicks when the button is pressed; the ones that can be heard from time to time I would assume are due to switch bounce and can be ignored. If this is so then the problem narrows down to the control signal. The manufacturers state that the switch is a special model made for audio signal with a gradual ramp on ramp off break before make contact; it still clicks loudly however. I'll try Conrad's suggestions and see what happens......
Vielle568
I did try putting the 10M resistor across the switch terminals. I used a mechanical switch initially as Aucosticraft suggested. There was no audible click when the switch was toggled; just a change in hiss (probably due to all the temporary cables wiring up the circuit; nothing shielded at the moment). However, when the analogue switch was put back in place the click was still audible, especially at low signal levels.
In reply to JensH the analogue switches in this circuit are placed after a preamp and signal splitter. They are controlled by a microprocessor; the pulse output passes through a Schmitt trigger inverting buffer to ensure a clean signal to control the switch. After the switch the pulsed audio signal passes to an output connector and is then cabled to a mixing table and power amp. There are three similar circuits in this design that run in parallel. And yes, I have actually tried connecting resistors of various values between the switch terminals and ground, but it didn't seem to have much effect on the clicks being generated.
In reply to Conrad Hoffman I have just replaced the analogue switch with a simple push type switch and tried pulsing the signal manually. The output is actually very clean and there are hardly any clicks when the button is pressed; the ones that can be heard from time to time I would assume are due to switch bounce and can be ignored. If this is so then the problem narrows down to the control signal. The manufacturers state that the switch is a special model made for audio signal with a gradual ramp on ramp off break before make contact; it still clicks loudly however. I'll try Conrad's suggestions and see what happens......
Vielle568
Last edited:
This is definitely because of charging of capacitor.
Can you add a 10E resistor in parallel with the coupling capacitor (the one at the output of NE5532 ? See what happens?
If the above solution did not work would you be ok replacing existing switch with CD4016 (its not pin compatible). If you intent to go with CD4016 I can suggest you how to configure it to avoid click.
Last edited:
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.