Topic: Circuit help needed for battery line mixer

[Chuck]

I want to build a nice, quiet, full bandwidth, 12v DC powered 4 input, 2 output line mixer. I've seen the passive designs, but there is a lot of signal loss there.
Does anyone have a good circuit for this? I know it's easy to do with op-amps, but op-amps need a bi-polar power supply, unless you do single supply biasing, which adds noise and complexity. A simple transistor circuit would be best, as long as it's 12v powered and very clean.

I saw this:
http://www.aaroncake.net/circuits/mixer1.htm

Parts:

R1, R3 2 10K Pot 
R2, R4 2 100K 1/4 W Resistor 
R5 1 6.8K 1/4 W Resistor 
C1, C2, C3 3 0.1uF Capacitor 
Q1 1 2N3819 Junction FET 
MISC 1 Wire, Shielded (Metal) Case, Phono Or Other Plug For Output
 
Is this a good circuit? Is the transistor the best one for the task?

Can anyone help?

[Chuck]

I just found this one...
Does anyone see any problems with this one? Would this be a better choice?
I'd rather run it on 12v DC. Would that change anything?
Maybe use a different op-amp?
I could do without the pan-pot too.

Parts:
P1,___________100K   Linear Potentiometer
P2_____________10K   Linear Potentiometer

R1,R2,_________15K   1/4W Resistors
R3,R4,R11,R12_100K   1/4W Resistors
R5,R6__________22K   1/4W Resistors
R7,R8_________390K   1/4W Resistors
R9,R10________560R   1/4W Resistors
R13___________220R   1/4W Resistor

C1,C2_________330nF   63V Polyester Capacitors
C3,C8_________100µF   25V Electrolytic Capacitors
C4,C5__________10pF   63V Ceramic Capacitors
C6,C7___________4µ7   63V Electrolytic Capacitors

IC1___________TL062   Low current BIFET Dual Op-Amp





--------------------------------------------------------------------------------

Circuit description:

The schematic of this circuit is drawn as a stereo unit to better show the input Main Fader and Pan-Pot connections. The TL062 chip contains two TL061 op-amps into the same 8 pin case and is wired as two virtual-earth mixer amplifiers having a voltage gain of about 4, to compensate for losses introduced in the passive Pan-Pot circuitry. Therefore, total voltage-gain is 1.
Each channel added to the mixer must include the following additional parts:
P1, P2, R1, R2, R3, R4, C1 and C2.
These parts must be wired as shown in the above circuit diagram, connecting R3 and R4 to pin #2 and pin #6 of IC1 for Right and Left channel respectively. These IC1 pins are the "virtual-earth mixing points" and can sum together a great number of channels.

Notes:

Current drawing for one stereo Main Mixer Amplifier Module is 800µA.
Frequency response is 20Hz to 20KHz - 0.5dB.
Total Harmonic Distortion measured @ 2V RMS output = < 0.008% @ 1KHz; < 0.017% @ 10KHz.
THD is 0.005% @ 1V RMS output.
Maximum undistorted output voltage: 2.8V RMS.

[Church-Audio]

Looks good, I would lose the pan pots and replace it with a 1% resistor metal film. As a mater of fact, I would replace all resistors with 1% metal film. The caps c6-7 I would replace with 10uf 3uf is too low for an output coupling cap. I would check out also c1-c2. So for P2, if that is the pan I would replace it with a 5k 1% resistor, and I would hand match all of the resistors used to substitute the pan control so they are even closer then 1%. That way your stereo image is 100% correct. The distortion measurements mean nothing to me, because they did not spec at what input voltage this distortion is at. This is a critical factor, its no good having a preamp or mixer with a distortion of 0.004% if its at a input signal of say -30db This is a common trick among circuit designers, I measure my distortion figures at the real input level the preamp in question is likely to face day to day. I then I add 6 db to that input and then I measure the distortion. Build the circuit first do you have a breadboard? just wondering.

[Chuck]

Looks good, I would lose the pan pots and replace it with a 1% resistor metal film. As a mater of fact, I would replace all resistors with 1% metal film. The caps c6-7 I would replace with 10uf 3uf is too low for an output coupling cap. I would check out also c1-c2. So for P2, if that is the pan I would replace it with a 5k 1% resistor, and I would hand match all of the resistors used to substitute the pan control so they are even closer then 1%. That way your stereo image is 100% correct. The distortion measurements mean nothing to me, because they did not spec at what input voltage this distortion is at. This is a critical factor, its no good having a preamp or mixer with a distortion of 0.004% if its at a input signal of say -30db This is a common trick among circuit designers, I measure my distortion figures at the real input level the preamp in question is likely to face day to day. I then I add 6 db to that input and then I measure the distortion. Build the circuit first do you have a breadboard? just wondering.

Chris:

Thank you very much for your response.
Basically, what I want is two stereo inputs with one stereo output. Both inputs have gain control. Does the single supply biasing for this circuit comprimise it in any way? I just want the cleanest way to combine two line level sources. I'm not entirely sure how to remove the pan pot.
The designer also noted that the pan pot drops the gain, so there is gain added by the op-amp to compensate.
I do not have a bread board. I usually just use a perfboard, and twist the leads togther as needed before soldering to test.

Would using 12v DC make a difference in this circuit? Would I have to change any of the components?

[Church-Audio]

Looks good, I would lose the pan pots and replace it with a 1% resistor metal film. As a mater of fact, I would replace all resistors with 1% metal film. The caps c6-7 I would replace with 10uf 3uf is too low for an output coupling cap. I would check out also c1-c2. So for P2, if that is the pan I would replace it with a 5k 1% resistor, and I would hand match all of the resistors used to substitute the pan control so they are even closer then 1%. That way your stereo image is 100% correct. The distortion measurements mean nothing to me, because they did not spec at what input voltage this distortion is at. This is a critical factor, its no good having a preamp or mixer with a distortion of 0.004% if its at a input signal of say -30db This is a common trick among circuit designers, I measure my distortion figures at the real input level the preamp in question is likely to face day to day. I then I add 6 db to that input and then I measure the distortion. Build the circuit first do you have a breadboard? just wondering.

Chris:

Thank you very much for your response.
Basically, what I want is two stereo inputs with one stereo output. Both inputs have gain control. Does the single supply biasing for this circuit comprimise it in any way? I just want the cleanest way to combine two line level sources. I'm not entirely sure how to remove the pan pot.
The designer also noted that the pan pot drops the gain, so there is gain added by the op-amp to compensate.
I do not have a bread board. I usually just use a perfboard, and twist the leads togther as needed before soldering to test.

Would using 12v DC make a difference in this circuit? Would I have to change any of the components?

Disregard what I said about  the 5 k resistor, and simply cut the pan pot out of the circuit. put nothing between the two resistors after the input level pot. I should warn you that the gain on this mixer is all 100% all the time this might increase your noise floor. But if the design is good it should be OK. It would be simple to change the feedback loop resistor to change the gain. If more or less is needed. It all depends on the type of input level this mixer will see what are you feeding it? Last but not least GET A BREAD BOARD they only cost $10 to $50 for a good one and they will save hours of reworking with a soldering iron, besides you might burn your fingers Go out and get one dammit don't be a cheap skate.

[Chuck]

Looks good, I would lose the pan pots and replace it with a 1% resistor metal film. As a mater of fact, I would replace all resistors with 1% metal film. The caps c6-7 I would replace with 10uf 3uf is too low for an output coupling cap. I would check out also c1-c2. So for P2, if that is the pan I would replace it with a 5k 1% resistor, and I would hand match all of the resistors used to substitute the pan control so they are even closer then 1%. That way your stereo image is 100% correct. The distortion measurements mean nothing to me, because they did not spec at what input voltage this distortion is at. This is a critical factor, its no good having a preamp or mixer with a distortion of 0.004% if its at a input signal of say -30db This is a common trick among circuit designers, I measure my distortion figures at the real input level the preamp in question is likely to face day to day. I then I add 6 db to that input and then I measure the distortion. Build the circuit first do you have a breadboard? just wondering.

Chris:

Thank you very much for your response.
Basically, what I want is two stereo inputs with one stereo output. Both inputs have gain control. Does the single supply biasing for this circuit comprimise it in any way? I just want the cleanest way to combine two line level sources. I'm not entirely sure how to remove the pan pot.
The designer also noted that the pan pot drops the gain, so there is gain added by the op-amp to compensate.
I do not have a bread board. I usually just use a perfboard, and twist the leads togther as needed before soldering to test.

Would using 12v DC make a difference in this circuit? Would I have to change any of the components?

Disregard what I said about  the 5 k resistor, and simply cut the pan pot out of the circuit. put nothing between the two resistors after the input level pot. I should warn you that the gain on this mixer is all 100% all the time this might increase your noise floor. But if the design is good it should be OK. It would be simple to change the feedback loop resistor to change the gain. If more or less is needed. It all depends on the type of input level this mixer will see what are you feeding it? Last but not least GET A BREAD BOARD they only cost $10 to $50 for a good one and they will save hours of reworking with a soldering iron, besides you might burn your fingers Go out and get one dammit don't be a cheap skate.

Thanks again Chris,

No, money isn't why I don't have a breadboard. It just never occured to me that it would be useful...
I will take your advice and get one. 

Yes, I realized that the circuit has a fix gain. I will probably lower it some. So, if I remove the pan pot, then I can connect the other stereo input essentially at the same place the pan-pot is now, right? I'm going to re-draw it and re-post later, to make sure I have it all figured out correctly...

As for the capacitors, should I replace the ceramics with something else? Tantalum maybe? This part of it is like the "black magic", I'm a bit lost as to what substitutions to make for the very best sound/ performance.

[Chuck]

Isn't 100k a little high for the input pot? and 100k for R3 

Is this passive input section going to be OK? It seems like a brute force way to combine the inputs.

I wish there was just an easy to follow (already thought out) schematic for this. It can't be that difficult right?

I guess I just don't understand the part of the circuit that makes the op-amp single supply rather than bi-polar supply. I wonder if it compromises the circuit in some way?

[Chuck]

Circuit reworked to remove the pan pots and simplify a bit. I removed the coupling capacitors in an attempt to avoid the distortions generated from them. I also lowered the input pots value from 100k to 10k.

Any help, advice is appreciated.

[ghellquist]

You have to insert the coupling capacitors again (C1 and C2) in original circuit. These are necessary to remove any DC bias you get from the inputs. And no need to worry, they do not add any significant amount of distortion.

Gunnar

[Chuck]

Yeah, I realized that. I'll post the revised circuit soon.

[Church-Audio]

Yeah, I realized that. I'll post the revised circuit soon.

Use 10uf 63 volt Poly caps in the input section.

[Chuck]

Chris, is there a brand and model number that you prefer? Also, wat are the best caps for bypassing at the + and - side of the op-amps? I also know that people bypass around those DC blocking caps too.

[kuuan]

sorry for plunging in.
I want to build a mic mixer and wonder if I could use Chuck's circuit.

Would it work mic level, or would I have go mic > preamp > mixer?
To mix mics I'd again need the pan pots, but should I keep the other mods of Chris?

Is my drawing correct?
If coming from my preamps could it be a good idea to omit the 100k input pots? Chuck, in your drawing you had made them 10k, but according to the original circuit the input pots are 100k, the pan pots are 10k. Was this on purpose, also that you omited R 9 and 10 and C 6 and 7 at the output section?

Are pan pots regular pots?

[Chuck]

I'm still working this. I'm debating now, whether to add two more channels for microphone inputs. I have been looking at the INA217 and the microphone pre-amp circuit TI has in the literature for it. I am going to do more research, before I build this. I want to get it right and use good layout and probably point to point wiring, as those home-made circuit boards are not for me.

That circuit you re-drew is for a line mixer.
I figured the 100k input pots were a bit high, and may cause exsesive noise. 10k log pots should work. I'm looking at the Panasonic EVJ-Y10F03A14, dual 10k log pot, that the headphone amp builders use. It's relatively inexpensive and easily bought at Digi-Key. As far as the pan pots, you can buy some that have the center nesting, which may be useful. Again Panasonic makes some. I didn't want to include the panning, so I omitted it. Unfortunately, you really do need the output capacitor, as with the single supply biasing there will be DC offset at the output.

If I go with making this a microphone pre-amp too, I may power it with those small SLA's, so I can get clean phantom power. Some of those 6v cells can be split off and used to power pre-amp/ mixer section.

[kuuan]

I shall observe with interest how your final project will look like.
Would it work as a mic mixer though if each channel gets amplified by my DIY preamps? ( http://taperssection.com/index.php?topic=71782.0;all )

I have problems finding good pots here in austria. ( Digi-Key runs seperate sites for different countries but at their austria site the panasonic pot you mentioned is not sold, though it is at their US site, even though the order will be sent from the US and a charge of 31 Euros, over 40 US!! is added for the service. )
Found some at a conrad site which has shops in austria and will be much cheaper for me to buy from, but did not find any metal film rotary switch as recommended by Chris for my preamp.

pots, probably the best I found are:

'Vishay' branded socalled 'Drahtpotentiometer' in german, translated 'wire-potentiometer' with 10 steps, can be had in mono only, costing 8 euro each
stereo pots branded 'Alps', e.g. model number: RK27112 50KBX2CC, costing 14 Euro each.

Do you know these brands? Are they recommendable? They are expensive...wonder if I can find chaeper pots which do a good enough job ( specially for me, a beginner )

[Church-Audio]

Isn't 100k a little high for the input pot? and 100k for R3 

Is this passive input section going to be OK? It seems like a brute force way to combine the inputs.

I wish there was just an easy to follow (already thought out) schematic for this. It can't be that difficult right?

I guess I just don't understand the part of the circuit that makes the op-amp single supply rather than bi-polar supply. I wonder if it compromises the circuit in some way?

This depends on the input impedance of the circuit in question 100k seems high I would say 10k to 20k would be fine. But again you have to take a close look at the impedance of the input stage of the opamp.

[Church-Audio]

Chris, is there a brand and model number that you prefer? Also, wat are the best caps for bypassing at the + and - side of the op-amps? I also know that people bypass around those DC blocking caps too.

To be very honest I get most of my caps from eBay, the surplus market is great as long as you test them before you use them they are on average 50% cheaper then retail. For coupling caps I use anything from Tantalum to Polystyrene and even monolithic. Take a look at bc components they make some sweet caps at a very small size go to e-sonic.ca also check out digikey.com for bc components.
Check out the Cornell Dubilier type DME, or surface mount tantalum if you want really small cheap high capacitance values. Kermit type t197a106k050as.

[poorlyconditioned]

Chris, is there a brand and model number that you prefer? Also, wat are the best caps for bypassing at the + and - side of the op-amps? I also know that people bypass around those DC blocking caps too.

To be very honest I get most of my caps from eBay, the surplus market is great as long as you test them before you use them they are on average 50% cheaper then retail. For coupling caps I use anything from Tantalum to Polystyrene and even monolithic. Take a look at bc components they make some sweet caps at a very small size go to e-sonic.ca also check out digikey.com for bc components.
Check out the Cornell Dubilier type DME, or surface mount tantalum if you want really small cheap high capacitance values. Kermit type t197a106k050as.

If you're ever in Toronto, visit "Supremetronic" on College Street.  This is a small "hole in the wall" store, but they have tons of components, particularly caps and resistors, all values/types.  I can get 2.2uF film caps for something like 60 cents each!  Much cheaper than buying for Digikey or other sources.

  Richard

[kuuan]

notes on the circuit, mostly to high for me, possibly nothing new to you, taken from answers in a german electronic forum

not all apply to Chuck's version of the circuit

The resistors in the loop of the level pots are relatively small. Consequentely,with the level pots being log., level will first rise very slowly, but very steep at the end. Remedy: choose lin. level pots. or: level pot only 1/10 ( 10k ) and in exchange following ( pre) -resistores and pan pot 10x larger ( 150k and 100k )
log. pan pot doesn't work, medium position will not mean equal distribution to both channels.
in mid position of a pan pot, if lin., level diminishes 5/5+15 0 1/4. The amplification of the opamps is 390/100 = 3.9, just equalling.

how to optimize amplification I was told: one can make the resistores on the opams bigger than 390k. E.g. 1MOhm results in (1M/100k)/4 = 2.5 x amplification. Higher ohm however means higher sensitivity to interferences.

btw: I had drawn the caps in the input the wrong way around, if using 10uF instead of the 330 ploy caps.
he says: the coppeling condenser together with the following 100k res. build a 6db bass cut. With the 330uF the 'corner' frequence is 4,8 Hz. with 10uF it makes a sensational 0,159Hz. The improovment with the condenser's value therefore is academic only.

[Chuck]

I've given up on the original schematic too. I am in the process of changing my two channel recording rig, so my attention is not on this project for now. I may eventually build a mixer with two microphone inputs and 2-4 line inputs. I have been slowly buying the hardware (XLR & RCA panel mounts, teflon coated wire, Hammond box etc...) so when I find a circuit I like I will be all set.

Watch out for those higher value resistors. Resistors can add noise and some of those values you mentioned will add lots of gain. The original circuit is for a line mixer, so lots of gain was not the goal.

Richard (poorlyconditioned) made a remark in a thread that is slowly sinking in with me. He said, essentially...  you can go out and buy good sounding, well designed pre-amps. Designing and building them may not be worth the time. Well, when I finally do build another one, I want it to be good, so I'm going to proceed slowly.

[kuuan]

I have not given up on this schematic.
I enjoyed building the small preamp, I'd enjoy building this mixer.
However I am afraid that my mics don't make for mixing:
 
I have 2 stereo mics, an AT822 which is an X/Y stereo mic in a single housing, and a pair of CMC4s, two speparate mics but joined by one wire.
I ask myself: Does panning of any of these four channels make any sense at all? Would I not better leave clean left and right channels? The mixer also would not have level pots as I'd adjust their levels at their respective preamps.
If there are no pan pots and no level pots all that's left is simply an 'adder', but I was told that 'adding' 2 stereo mics is not done.

My question:
Do you think that 'adding' my two stereo mics, their left and right channels respectively, makes any sense?
Or could even panning indiividual channels of my mics make sense?

As I see it now the only possible config. to actually 'mix' would be if I had a mono mic, a lav on a person ( I have two Giant Squid omni lavs ) and mix it with ambient sound of a stereo mic.
For this again all I'd need is an 'adder' as I'd better make an adapter to feed the mono signal fully to left and right channles instead of panning it.

So instead of the mixer all I may be able to use is an 'adder', adding the left and right channels of two stereo mics, which also can be used to add a mono signal to a stereo mic fully to both channels.

well, I hope my questions are not too stupid.
I enjoy this hobby, finding out, solering asf.,  my experience recording equals zero though. I wanted to make better audio for my videos, found this site and see how a vast and fascinating field this is. I will never be at a paar with you guys, and I am very grateful that you kind folks bear with me, even teach me.

In case I learn that as adder can be useful for me I'll have to enquire further.

thanks,
a.