Hey, folks.
I currently use a small, active sub-mixer in my live-performance rig. this combines the output from three sources (two computer DAWs and one iPod) then sends this to my powered speakers. (and / or power amps)
It sounds pretty crappy, even after extensive mods, so I would like to replace it with a passive mixer + a simple line amp for gain makeup.
I would like to make my own, both to save space and to optimize it for my specific needs. For instance, my source impedance will always be fairly low, and I don't need pan pots. (I'd also like to have very good volume pots, which are rarely found on simply commercial products)
My question today is what resistors to use, in the passive section:
-----------------------------------------
1: Regarding the input resistors: I read on one site that "The value of the mixer’s input resistors must be equal to, or greater than, the highest source impedance." - And a corresponding schematic shows 1K per leg.
I find this surprising, since I've always been told that an ideal input impedance is ten times the source. Can someone confirm which is correct, and why?
Obviously, more resistance means more volume loss, so ....
My highest source impedance would be from a Lynx Aurora DAC, which is 50 ohms unbalanced, 100 ohms balanced. Since the schematic referred to is for an unbalanced source, that implies I only need 50 ohm resistors. Can that possibly be right? Using the "ten to one" rule, it's still only 500 ohms. How much loss would that give? (it can't be much, since 5K resistors supposedly give only a 6 db loss.)
2: If I add volume pots (typically 10K) these would probably almost always stay "close" to maximum volume. Thus they would not contribute much to the input impedance of the mixer, correct?
# 3: In some schematics, the various inputs simply pass through a set (+ / - ) of resistors, then simply combine with the other inputs, before going to the next device or makeup gain amp. In other designs, such as NYDave, there is a resistor at the output, connected BETWEEN each of the two final + and - lines. (shunt resistors?)
In this case, a 16 channel mixer using 6.8K inputs, had 182 ohm "output" resistors. (or whatever you call them) but he recommends a different "output" value with less channels or a higher input impedance.
I don't understand the purpose of that resistor. (or why some designs don't use it at all.) Can someone explain this, and recommend a value (if needed) for my situation? I see the values listed, but no explanation of what formula was uses, or why. (evidently it affects output impedance. Maybe that's ALL it does?)
=====================
BALANCED INPUTS?:
I don't really NEED balanced inputs, but that would be nice. Can such a mixer easily be made balanced? If so, how?
I currently use a small, active sub-mixer in my live-performance rig. this combines the output from three sources (two computer DAWs and one iPod) then sends this to my powered speakers. (and / or power amps)
It sounds pretty crappy, even after extensive mods, so I would like to replace it with a passive mixer + a simple line amp for gain makeup.
I would like to make my own, both to save space and to optimize it for my specific needs. For instance, my source impedance will always be fairly low, and I don't need pan pots. (I'd also like to have very good volume pots, which are rarely found on simply commercial products)
My question today is what resistors to use, in the passive section:
-----------------------------------------
1: Regarding the input resistors: I read on one site that "The value of the mixer’s input resistors must be equal to, or greater than, the highest source impedance." - And a corresponding schematic shows 1K per leg.
I find this surprising, since I've always been told that an ideal input impedance is ten times the source. Can someone confirm which is correct, and why?
Obviously, more resistance means more volume loss, so ....
My highest source impedance would be from a Lynx Aurora DAC, which is 50 ohms unbalanced, 100 ohms balanced. Since the schematic referred to is for an unbalanced source, that implies I only need 50 ohm resistors. Can that possibly be right? Using the "ten to one" rule, it's still only 500 ohms. How much loss would that give? (it can't be much, since 5K resistors supposedly give only a 6 db loss.)
2: If I add volume pots (typically 10K) these would probably almost always stay "close" to maximum volume. Thus they would not contribute much to the input impedance of the mixer, correct?
# 3: In some schematics, the various inputs simply pass through a set (+ / - ) of resistors, then simply combine with the other inputs, before going to the next device or makeup gain amp. In other designs, such as NYDave, there is a resistor at the output, connected BETWEEN each of the two final + and - lines. (shunt resistors?)
In this case, a 16 channel mixer using 6.8K inputs, had 182 ohm "output" resistors. (or whatever you call them) but he recommends a different "output" value with less channels or a higher input impedance.
I don't understand the purpose of that resistor. (or why some designs don't use it at all.) Can someone explain this, and recommend a value (if needed) for my situation? I see the values listed, but no explanation of what formula was uses, or why. (evidently it affects output impedance. Maybe that's ALL it does?)
=====================
BALANCED INPUTS?:
I don't really NEED balanced inputs, but that would be nice. Can such a mixer easily be made balanced? If so, how?
Last edited:
Related question:
I just read that with such a passive mixer, any volume adjustment on one input affects the volume of all three. Is this true?
Does using higher value input resistors minimize this effect?
I just read that with such a passive mixer, any volume adjustment on one input affects the volume of all three. Is this true?
Does using higher value input resistors minimize this effect?
Yes having higher value input mixer resistors cuts the feedback to the sources which means changing value on one doesn't affect the others.
Really if two inputs are 8 ohms out and one 50, having the input resistors at 8 would be somewhat dangerous for the active devices at the outputs of the various sources. I'd say 10x is much safer. 10k input resistors would be a whole lot safer.
Really, considering how silent active stages (op amps) can be, and the possible dangers of the interactions of sources if you get a plug wrong in setup, I'd go for an op amp at the end of the sum resistors of 10k. Not a 4558, a low noise MC33078, NJM2068, NE5532, LM4572 or something more expensive. Cost, a 1 dip socket, a $.38 op amp, some $.10 resistors and caps, another wall transformer and some RF filtering and capacitance as the power enters the steel box.
There are ways to blow up op amps or discrete transistors, that is getting the base and emitter more than 5 v apart backwards. Most mixers might encounter a guitar amp output at 17 vac or so, so pro mixers have a way to protect themselves from such a situation. A resistor between the high and low of the inputs is one way, low enough to blow up sacrificial 1/10 W resistors (or more expensive fuses) in the input lines. These are usually coupled with diodes that conduct large currents if the input gets outside the rails of the power supply. I've bought a set of gear from an amateur band that apparently caused this problem a lot, all the input resistors or op amps were blown up.
Balenced input mixer just costs 2 op amp sections per input plus stereo pots per input, so they cost more. See various commercial schematic diagrams of high end mixers. My Peavey Unity is not that expensive, it is single ended input. Schematics are often on eserviceinfo.com
Long cables are sources of RF interference into mixers, which makes them howl and whizz, so even the outputs should have some resistance or capacitance across the wires to short that stuff out. I use disk capacitors myself across, plus free coils out of dead switcher power supplies in series with output wires. The world is full of RF these days; bands are using radio mikes & pickups on the business band, and everybody has a cell phone transmitting even when they are not talking. I reworked a $15 disco mixer to have all this stuff, stop hissing & humming & picking up the lamp dimmer & the mad CBer driving by, an RA-88a, which see, but I don't have the bookmark handy on this computer. Cost about $20 in parts.
As far as pots, my Peavey Unity 12 is 1998 and has good pots. They are rotary so probably the money is saved on the configuration, not the thin carbon track. Numark and Behringer, used mixers on the craigslist may not be so fortunate. From the number of questions about the input pots on here, I guess replacement pots for latter brands are expensive and weird sized (probably patented) too. The way Dodge Neon fuel hoses are $80 and have to be flown in from Atlanta because they are patented & sole source. Sell low, make money on the *****y patented parts, is the service model introduced by some very successful import brands. Whereas Ford & Chevy used standard generic parts as much as possible until at least the 80's.
Really if two inputs are 8 ohms out and one 50, having the input resistors at 8 would be somewhat dangerous for the active devices at the outputs of the various sources. I'd say 10x is much safer. 10k input resistors would be a whole lot safer.
Really, considering how silent active stages (op amps) can be, and the possible dangers of the interactions of sources if you get a plug wrong in setup, I'd go for an op amp at the end of the sum resistors of 10k. Not a 4558, a low noise MC33078, NJM2068, NE5532, LM4572 or something more expensive. Cost, a 1 dip socket, a $.38 op amp, some $.10 resistors and caps, another wall transformer and some RF filtering and capacitance as the power enters the steel box.
There are ways to blow up op amps or discrete transistors, that is getting the base and emitter more than 5 v apart backwards. Most mixers might encounter a guitar amp output at 17 vac or so, so pro mixers have a way to protect themselves from such a situation. A resistor between the high and low of the inputs is one way, low enough to blow up sacrificial 1/10 W resistors (or more expensive fuses) in the input lines. These are usually coupled with diodes that conduct large currents if the input gets outside the rails of the power supply. I've bought a set of gear from an amateur band that apparently caused this problem a lot, all the input resistors or op amps were blown up.
Balenced input mixer just costs 2 op amp sections per input plus stereo pots per input, so they cost more. See various commercial schematic diagrams of high end mixers. My Peavey Unity is not that expensive, it is single ended input. Schematics are often on eserviceinfo.com
Long cables are sources of RF interference into mixers, which makes them howl and whizz, so even the outputs should have some resistance or capacitance across the wires to short that stuff out. I use disk capacitors myself across, plus free coils out of dead switcher power supplies in series with output wires. The world is full of RF these days; bands are using radio mikes & pickups on the business band, and everybody has a cell phone transmitting even when they are not talking. I reworked a $15 disco mixer to have all this stuff, stop hissing & humming & picking up the lamp dimmer & the mad CBer driving by, an RA-88a, which see, but I don't have the bookmark handy on this computer. Cost about $20 in parts.
As far as pots, my Peavey Unity 12 is 1998 and has good pots. They are rotary so probably the money is saved on the configuration, not the thin carbon track. Numark and Behringer, used mixers on the craigslist may not be so fortunate. From the number of questions about the input pots on here, I guess replacement pots for latter brands are expensive and weird sized (probably patented) too. The way Dodge Neon fuel hoses are $80 and have to be flown in from Atlanta because they are patented & sole source. Sell low, make money on the *****y patented parts, is the service model introduced by some very successful import brands. Whereas Ford & Chevy used standard generic parts as much as possible until at least the 80's.
Last edited:
Thanks, Joe.
That's some REALLY helpful info.
I'm mostly on the same page now. Passive mixers seem like such a great idea, until you really understand the shortcomings. Plus, if you need a high-gain makeup stage, then you are battling noise, distortion & possibly slew issues. Eesh.....
That's some REALLY helpful info.
I'm mostly on the same page now. Passive mixers seem like such a great idea, until you really understand the shortcomings. Plus, if you need a high-gain makeup stage, then you are battling noise, distortion & possibly slew issues. Eesh.....
Last edited:
- Status
- Not open for further replies.