I am interested in building a audio inputs switch (audio selector). I have experience in digital circuits, but not much in analog circuits, but I am passionate about audio. So I need a little help.
Had thought of using a pair of 4051 with 8 inputs each, and also a small micro controller (atmel avr or microchip pic). It would be better to use the 4052? (I prefer to have 8 stereo inputs with two 4051, one for each channel)
Someone can recommend me any design or schematic?
I have to put a "audio buffer" or preamp? in the output? in the input? in two places? (if so, could use a NE5532?)
Any information is welcome: schematics, designs, tips, ideas... (I need help especially in the analog audio part (input or output preamplifier, buffer...)
Thanks you very much for everything!
PS: The 4051 has an acceptable quality? or make noise in the audio?
Thanks! bye!
Had thought of using a pair of 4051 with 8 inputs each, and also a small micro controller (atmel avr or microchip pic). It would be better to use the 4052? (I prefer to have 8 stereo inputs with two 4051, one for each channel)
Someone can recommend me any design or schematic?
I have to put a "audio buffer" or preamp? in the output? in the input? in two places? (if so, could use a NE5532?)
Any information is welcome: schematics, designs, tips, ideas... (I need help especially in the analog audio part (input or output preamplifier, buffer...)
Thanks you very much for everything!
PS: The 4051 has an acceptable quality? or make noise in the audio?
Thanks! bye!
I'm not sure about the quality of the old CMOS muxes. There's newer and more expensive muxes available from mfgrs like Maxim and Analog Devices.
ADG508FBN from Analog Devices is one that I've used in data acquisition applications. The specs are pretty good, and it has the sometimes useful feature of going high-impedance on the inputs when unpowered. The dual 4:1 version would be more useful, the ADG509F, since one chip is all you need for a stereo four-input switch.
Maxim's versions are the MAX4708 and MAX4709, if those are more available. Just beware of chips from vendors in China that are used non-faultprotected MUX chips that have been removed from boards and re-marked. I nearly got stung with 100s of those, buying through a broker.
ADG508FBN from Analog Devices is one that I've used in data acquisition applications. The specs are pretty good, and it has the sometimes useful feature of going high-impedance on the inputs when unpowered. The dual 4:1 version would be more useful, the ADG509F, since one chip is all you need for a stereo four-input switch.
Maxim's versions are the MAX4708 and MAX4709, if those are more available. Just beware of chips from vendors in China that are used non-faultprotected MUX chips that have been removed from boards and re-marked. I nearly got stung with 100s of those, buying through a broker.
I am interested in the classical CMOS 4051 or 4052, for two reasons: they are very cheap and easily found at any electronics shop. Thanks for your answer, all help are welcome!
Regards!
Regards!
go to the hafler web site and look for the manual for the 915 preamp.
this design will have some useful tips for you.
mlloyd1
this design will have some useful tips for you.
mlloyd1
What is the best way to buffering the in and out of 4051 (and 4052). Using the NE5532 operational (or other)? discrete design with transistors?
You have a drawing of the buffering circuit?
One question, could I use the output buffer to send the signal to more than one output (i.e: amplifier, REC output, VCR ...)?
You have a drawing of the buffering circuit?
One question, could I use the output buffer to send the signal to more than one output (i.e: amplifier, REC output, VCR ...)?
4051 and the likes have a resistance that varies with the applied signal voltage. If you don't object to phase inverson, - using an opamp as an inverting amp- try to use the mux part inside a summing/current amp - mux output directly couples to the - input, and choose series and feedback resistors to match your impedance and gain demands.
The SSM-2402 is a fundamentally better part for the task - albeit quite more expensive. This app.note might give some ideas....
http://www.analog.com/static/imported-files/application_notes/35650192AN121.pdf
The SSM-2402 is a fundamentally better part for the task - albeit quite more expensive. This app.note might give some ideas....
http://www.analog.com/static/imported-files/application_notes/35650192AN121.pdf
Hello, like AuroraB said, I've been looking schematic of pre-amplifier Hafler.
I have seen the 4051, has connected the audio input directly, without any input buffer. In the output it has a buffer, a FET Follower (is the correct name?).
I modified the schematic to suit what I want to do, is attached. Is that correct? If I build like this, works ok?
As I said I do not know much about analog signals. I can connect "Audio amp output to" to more than one input line of various device at the same time (without loss of signal)?
Another question, a 2SK163 replacement easy to find? I asked in my usual electronics but they don't have it.
Thanks a lot!
I have seen the 4051, has connected the audio input directly, without any input buffer. In the output it has a buffer, a FET Follower (is the correct name?).
I modified the schematic to suit what I want to do, is attached. Is that correct? If I build like this, works ok?
As I said I do not know much about analog signals. I can connect "Audio amp output to" to more than one input line of various device at the same time (without loss of signal)?
Another question, a 2SK163 replacement easy to find? I asked in my usual electronics but they don't have it.
Thanks a lot!
Attachments
Steerpike said:
Steerpike, can you give me more info about how to buffering the input and outputs of 4051 (or 4052)? Thanks!
4051 muxes
Please bear in mind that these chips are NOT silent switching. By their very nature, there is an artifact at the time of switching. The Analog Devices purpose built mux/switches are better.
Also, put the switch at the virtual earth point of the output buffer amp so you are switching current, not volts.
Please bear in mind that these chips are NOT silent switching. By their very nature, there is an artifact at the time of switching. The Analog Devices purpose built mux/switches are better.
Also, put the switch at the virtual earth point of the output buffer amp so you are switching current, not volts.
You could always control relays (you know, those old-fashioned clunky things) digitally and avoid all the problems of analogue switches.
I'd use an NE5534A or 5532A configured as a unity gain amp on both the input and output for each switch. Op amp can be inverting - which supposedly sounds better, and will maintain absolute phase since there are two.
this will likely do what you want assuming:
1. you are ok with unity gain (otherwise, you will need a gain stage following)
2. you are not driving low impedance loads. i wouldn't use the jfet buffer alone to drive loads less than about 3K, but that's my choice.
it has been my experience that 2SK163 are easy to find but a detailed data sheet is not.
😉
good luck!
mlloyd1
1. you are ok with unity gain (otherwise, you will need a gain stage following)
2. you are not driving low impedance loads. i wouldn't use the jfet buffer alone to drive loads less than about 3K, but that's my choice.
it has been my experience that 2SK163 are easy to find but a detailed data sheet is not.
😉
good luck!
mlloyd1
pekinin said:
I would really love to build a 4-channel switch for my audio project using the previously mentioned MAX4709. I'm interested in this if it is less noise/distortion than simply using small DPDT switches. I would also like to have a switch that uses a single button to select each source.
I found this:
http://www.electronicsforu.com/efylinux/circuit/feb2003/Add on stereo channel selector.pdf
which looks very interesting but I'd like to use the much easier to find LM3914 instead of the KA2281 in that shem. Or better yet, omit it completely for a simpler circuit. I'm also curious what kind of distortion it might have. I've searched around the net a LOT but haven't found much (very little on ESP pages too). Anyone know a good schem for this kind of a circuit?
I found this:
http://www.electronicsforu.com/efylinux/circuit/feb2003/Add on stereo channel selector.pdf
which looks very interesting but I'd like to use the much easier to find LM3914 instead of the KA2281 in that shem. Or better yet, omit it completely for a simpler circuit. I'm also curious what kind of distortion it might have. I've searched around the net a LOT but haven't found much (very little on ESP pages too). Anyone know a good schem for this kind of a circuit?
For analogue switching you can't beat this. The FET's are at the input to a virtual earth amp and so see no signal across them. The isolation is far better than any switch or relay... and that is something those 4066 are poor for, due to that stray capacitance etc. The second opamp is just to maintain correct phase.
And this lends itself to full remote control too.
And this lends itself to full remote control too.
Attachments
After try all type of audio multiplexer I have only one thing to say : 4051 and or 4052 do not meet my standards in audio !
Relay are the best .
Relay are the best .
Hi danzup,
Solid state and relays have pros and cons.
The dual series shunt FET arrangement gives far superior isolation than a single relay, it's cheaper, it's easier to place exactly where needed on the PCB, consumes zero power and is more reliable long term. The secret is to use them correctly so they never see any voltage... hence the virtual earth input stage... and to drive the gate slightly positive which substantially reduces the on resistance, but increase by no more than 0.3v above ground 🙂
Solid state and relays have pros and cons.
The dual series shunt FET arrangement gives far superior isolation than a single relay, it's cheaper, it's easier to place exactly where needed on the PCB, consumes zero power and is more reliable long term. The secret is to use them correctly so they never see any voltage... hence the virtual earth input stage... and to drive the gate slightly positive which substantially reduces the on resistance, but increase by no more than 0.3v above ground 🙂
Ok , you can do what you want ...... but :
- I do not think that FET can give decent isolation !!!
- I think you cannot beat 3..5 mm which is the distance between the relay armature in air nevertheless in vacuum (!!!) since now I am using reed vacuum relay with Ag armature !!!!
This days relay are not like 50 years ago but FET are !
Also relay you can put wherever on pcb you want .
Hi... with regard to the isolation... a good test is to apply a squarewave at around 3 khz or so and with the amp on FULL volume and your ear up to the speaker see if any breakthrough occurs. I can apply 20 volts peak to peak (max my function generator gives) with zero breakthrough. You certainly couldn't measure it as it is below the noise floor. That is an extremely severe test.
Depending on the relay design you may still experience breakthrough at HF due to capacitive effects.
Depending on the relay design you may still experience breakthrough at HF due to capacitive effects.