Many people begin their home recording dreams believing that an inexpensive program to go with the sound card in their computer will put them on the road to stardom. They learn soon enough that they need microphones, preamps, monitors, a CD recorder, and all too soon own far more things to connect than they have places to which to connect them. Sometimes, though, a little low-tech is what the doctor ordered.
“Do I need a mixer?” is frequently a question early on. Though a mixer can often resolve the no-connections problem, it’s much more than a signal router.
When the user needs mic preamps as well as signal routing capability, a small mixer is frequently the best choice. It can pass signals from mics, keyboards, a CD player, or guitar processor into the sound card’s single pair of inputs, and you’ll usually get a listening volume control (so you won’t have to disturb the computer’s settings) to boot. While a mixer is a necessity when recording to two input channels from several mics (for instance on a drum kit), most track mixing in the computer-based project studio is performed in the workstation software. In a two-track recording system, a mixer rarely gets put to full use.
A patchbay is another approach to solving the spaghetti problem. With all the sources and destinations appearing at a single, easily accessible set of jacks, it’s easy to connect the mic preamp to the sound card input when recording vocals, the guitar processor when recording guitar parts, the sound card output to the monitor speakers when you’re overdubbing and mixing, and to the DAT when making a safety dub or copy to send to the duplication house. This is a very flexible approach, but it requires organization and good labeling as well as a clear understanding of where things need to go.
Still, trying to figure out how to get what you want to hear to your headphones when overdubbing without getting feedback from the speakers can be puzzling (gee, now I have to buy a headphone amplifier too?). For those who have been following my sprinklings of phase and polarity discussions in the erstwhile ‘Oops Wrong Button’ series, you might also recognize that you’re bothered by the comb filtering due to digital throughput delays when monitoring your own voice on headphones.
There should be a simple, non-intrusive solution to using a simple recording setup.
A homey solution
One solution is to use a home stereo receiver as the router for the gozintas and gozoutas. Nearly all receivers have an input selector switch for one or two line-level sources, a “tape monitor loop” that sends the input directly to the recorder and the recorder’s output to the power amplifier section (and then on to the speakers). There’s a switch that allows listening to either the playback or the source, most are equipped with a headphone jack and a switch to mute the speaker output (if it’s not automatic when plugging in the phones), and some even have inputs and outputs for two recorders so you can copy directly from one to the other, monitoring either.
It’s easy to knock this approach as being “non-professional,” but it’s a good and inexpensive way to get things to work. There could be one in the garage replaced by the kilowatt surround system in the home theater, or a yard sale can often yield one for a few bucks.
Home audio equipment nearly always uses RCA jacks, its operating level is close to the low-level standard of -10 dBV, and of course it’s all unbalanced, but this matches up with a lot of project-level gear. If you have a mix of balanced and unbalanced, high and low level gear, my Studio Problem Solvers (5/2000) and Pin 2 Hot (7/1999) ‘Oops Wrong Button’ articles will help you deal with those issues.
Home stereo receivers have integrated power amplifiers, so they don’t mate well with the powered monitors that have become much more popular for studio use today. Some have a pair of jacks on the back that allow you to take an output after the volume control and before the power amplifier, but that’s rare on the newer units. There are “stereo preamps,” the front end of the receiver without a power amplifier, but they tend to be audiophile items (read “expensive”). It’s just the ticket if your project is to transfer vinyl records to disk, but that’s not the problem we’re solving here.
A routing problem solver
You don’t need a receiver or sophisticated electronics to select inputs and outputs, you need a switch. The problem, like with most of the problem solvers I’ve written about in past issues, is that you can’t always buy just what you want. Why not build it? And why not toss a volume control for your monitors in the same box?
The first thing to do is decide what you need. The example I’ve shown here will suit many small project studios based around a computer with a simple sound card, but you can expand it to incorporate more sources and more destinations.
It’s not a mixer, nor is it a mic preamp; you’ll have to take care of those things outboard. It’s completely passive, other than the optional headphone amplifier, so if you’re careful with grounding and shielding when you build it, it won’t stomp on your signal. If you have a fine mic preamp, its signal won’t go through some unknown amplifier or buffer on the way to the sound card input. I’ve shown an unbalanced design, but if your system is balanced, all it takes, with a couple of caveats, is to double up on the wiring.
Since it’s passive, it doesn’t care about operating levels. What goes in is what comes out. If you need to deal with a mix of balanced, unbalanced, +4 dBu, and -10 dBV devices, that can all be done externally. You could build level converters into the box, but that’s getting beyond the scope of this straightforward design.
Here’s where we define what we want it to do. This is the first phase of product development, whether it’s a simple passive switch box or a 96-input automated console. I’ve chosen a set of requirements based on a modest single-room studio and the things that a solo artist might need to do.
The box provides inputs for two recorders and two other line-level sources, lets you switch an input (which can be the sound card’s output) to the monitor speakers or amplifier, and lets you record from any input source to either (or both) of the two recorders. Follow through the diagram in Figure 1 as we go along.
The primary recorder is usually the computer with a stereo sound card. The secondary recorder in this basic studio is marked as a DAT. It could be a CD recorder or even a cassette deck, a MiniDisk, a reel-to-reel recorder...or it could also be entirely unnecessary if you have a CD-R drive in your computer and dump the mix from hard disk directly to CD. With two recorders, it’s useful to be able to copy from one to the other, so both the DAT and sound card also appear on the input switch.
There are two other line-level inputs. One is general purpose, and it’s where you’d connect a keyboard, a guitar signal processor, a MiniDisk or mp3 player. The other is targeted here for a pair of mic preamps. There are a couple of reasons, both related to monitoring, for the mic preamps to get special treatment.
Each of the two recorders has a switch to select whether its input comes from the mic preamp, the auxiliary input, or the other recorder. The monitor selector switch sends the output of either the sound card or auxiliary recorder to the monitor speakers through a volume control so you can adjust the listening level without disturbing any level adjustments in the computer.
The auxiliary input can be switched through the volume control straight to the monitors so you can hear your synth without going through the computer. If you have a DAW interface such as the Digi 001 or SeaSound Solo that includes a hardware volume control but is tied to the computer with a fixed length cable, this feature will become very useful when you move the computer and interface into another room to get the fan noise out of the studio.
Since the output is at line level, it’s equally at home feeding powered monitors or a separate monitor amplifier. In addition to the monitor volume control, there’s a mute switch allowing you to mute the speakers without having to readjust the level. This is useful when tracking on headphones.
In the situation where you’re likely to be needing this switchbox, the mics and the speakers will probably be located in the same room. To avoid feedback, you don’t want it to be too easy to route the mic preamps directly to the monitors.
In this design that’s accomplished simply by not providing a position on the monitor selector switch for the mic preamps. It’s possible to feed the mic signal to the monitors by monitoring the sound card output, routing the mic to the sound card input, and having the software set so that the mic track is in “input monitor,” but you have to work at it. If you use your mic preamp as a direct input for a guitar or bass and want to monitor on speakers while you’re playing, that’s the way to do it.
Feeding the cans
So how do you hear your voice or acoustic instrument, other than acoustically, when recording with a mic? That’s where the optional headphone monitor section comes in. Build it in if you want it, or leave it out if you don’t, or just build that part of the project if that’s all you need. The headphone amplifier has a simple mixer on its front end to mix the mic(s) with the playback of the rest of the tracks coming from the sound card.
There are a couple of features in the headphone amplifier section that resolve a few annoyances when recording with a basic sound card setup. First off, it has a separate volume control from the main monitor output, so you can set the headphones to a comfortable level without having to readjust when listening to the playback on speakers. You’ll mix better if you listen at a consistent level.
Since the mic inputs can’t be routed directly to the monitor outputs, they get their own inputs to the headphone monitor. There are individual headphone level controls for the sound card mix and the mic preamp output so you can easily adjust the balance between the backing tracks and the track you’re currently recording.
You can also set a different balance in the phones than in your real mix. Often you’ll sing to a track better if your voice is louder in the phones than it would be in the final mix, or maybe quieter. This gives you independent control.
The two mic preamps are handled as two mono inputs with a pan pot and level control for each. This allows you to center-pan a single mic, or when using two mics to pan them logically in the phones (left and right if you’re recording stereo, or blend and center them if you’re recording mono). Of course, panning in the playback is determined by how you’ve set it in the mixing section of your recording program.
Every digital device has a minimum delay on the order of 1 to 3 milliseconds from input to output. Some singers notice that when monitoring their own voice from mic to sound card output—it doesn’t sound quite right in the headphones. This is caused by the comb filtering that results from combining the direct acoustic sound with the slightly delayed sound right at the eardrum. By providing a direct analog headphone monitor path, this effect is eliminated.
The mixer section of the headphone amplifier (Figure 2) consists of a dual op-amp IC1, one section per channel, set up for unity gain on each input. One mixer input for each channel is picked off from the monitor selector switch ahead of the master monitor volume control R1 (points A and B), with a level control R2 for the sound card mix. Each mic preamp output (points C and D) is split into left and right outputs through pan pots R11 and R12. Dual pots R3 and R4 adjusts the level of the panned outputs to the left and right mixer op-amps.
Since sensitivity between different headphone models is all over the map, a booster stage consisting of a dual op-amp per channel (IC2 and IC3) assures that there’s enough current to drive any headphone to a reasonable volume. The headphone driver behaves like two op-amps in parallel to double the amount of current that can be sourced by a single amplifier.
Since this project includes some parts that aren’t completely familiar, here are some tips. The three selector switches S1, S2, and S3 are 2-pole multi-position wafer switches. These switches have wiper contacts that swing around on a shaft, making a connection to each of several fixed contacts arranged around the circumference.
The Radio Shack switch in the parts list actually has six positions, more than you need if you build exactly what I described, but it leaves room for expansion. If you want to include inputs for a few keyboards, just add more wires. If you examine the switch closely, you can see the wiper and the fixed contacts, but it’s a bit difficult sometimes to relate them to the switch terminals. Figure 3 shows the switch with the terminals identified.
The volume controls R1–R4 are dual pots, two potentiometers ganged together on the same shaft. One pot controls each channel of the stereo pair (Figure 4). It’s important for their two resistances to “track” together accurately when rotating the control so that the center of the stereo image doesn’t wander as you adjust the volume. Some are better than others, though the ones that Radio Shack was selling when I wrote this article are fairly decent.
Selecting your pot
The master volume is the most critical, since that’s always going to be in your listening path. As you need four dual pots, test them all before you start assembly and use the best one for the master monitor level control R1.
You’ll need a voltmeter that can read millivolts and a 9 volt battery. A digital multimeter is handy because the voltage that you’ll be measuring can switch from positive to negative as you rotate the pot shaft.
Temporarily connect top terminals of the two sections together and connect it to one terminal of the battery (it doesn’t matter which), then connect the bottom terminals together and connect them to the other terminal of the battery. Now, connect the meter between the two wiper terminals. See Figure 5.
If you’re using the Radio Shack pot in the parts list, ignore the extra terminal opposite the three main terminals. It’s a tap for turning the pot into a “loudness” control, and we don’t use it.
Looking at the diagram, you’ll see that you have two voltage dividers connected to the same voltage source. Ideally the voltage at the wiper should be the same on both pots for any position along the resistive element, causing the voltage difference between them to be zero.
Nothing is perfect, however. Note how the meter reading changes as you slowly rotate the pot from one end to the other. Record the highest and lowest voltage over the rotational range, noting that the lowest voltage may be negative. The pot with the smallest spread of voltage differences is the one that tracks best. Of the five I tested, the best one ranged from -0.055 to +0.128 volts, a spread of 0.183 volts. Relative to the 9V source, that’s about a 2% imbalance between the two sections. Want more math? Okay: this translates to slightly less than 0.2 dB tracking inaccuracy, less of a change in balance than most of us are able to hear under most circumstances.
All the ICs in the headphone amplifier are Burr-Brown OPA2604. This is an excellent general purpose dual op-amp for audio applications. The 0.1 uF bypass capacitors on the power leads of the ICs are polyester film type, while the larger value capacitors are electrolytic types.
Since the headphone amplifier isn’t highly critical, 5% tolerance is fine for the 100K and 5.1K Ohm fixed resistors. But feel free to splurge on 1% tolerance resistors.
Powering it up
You can of course use a pair of 9V batteries to provide the positive and negative voltages for the headphone amplifier section, but since this is likely to be turned on all the time you’re working, it’s sensible to use an AC power supply. Besides, I promised a couple of articles ago that I’d offer up a suitable power supply for your studio problem solvers.
Back in the 11/1998 and 12/1998 ‘Oops’ columns we discussed power supplies and how they work. The power supply in Figure 6 is one I’ve used for several little studio devices. It uses a wall wart for its input; buy a new one at Radio Shack, find one at a hamfest or electronics surplus store for a buck, or scrounge one from your junk box. Find one with an AC output of 15 to 20 volts. A DC wall wart won’t work. The circuit employs a full wave voltage doubler, supplying about 20 volts into the two voltage regulator ICs, which in turn regulate the outputs to 15 VDC.
Build it up
Here’s where you can get creative. If you have rack-mounted equipment, it’s convenient to build the switcher as a rack unit. You’ll need a blank rack panel as well as a chassis box. If you have a tabletop system, then a suitable sized chassis box without the rack panel is appropriate. While this unit operates with line-level signals, to prevent hum it’s a good idea to use a fully enclosed metal box rather than plastic.
Lay out the panel so the switches and pots are located conveniently, and mount the input and output connectors on the back panel of the chassis. Use whatever type of connector is most appropriate for your setup, probably 1/4" phone or RCA jacks. They don’t all have to be the same. If your sound card has mini stereo phone jacks, you may find it convenient to use similar jacks on the switchbox so that you can connect it with a mini-phone cable.
It’s convenient in most installations for signal in and out jacks to be on the rear of the box and the headphone jack to be on the front. You may opt for putting the line input jacks on the front, however, and use them like half a patchbay if you have several keyboards or an assortment of mic preamps. Or you might choose to use more positions on the selector switches and add more input jacks. Whatever design you settle on, use a square or a straightedge and ruler to lay out the holes and drill them accurately so the knobs will line up neatly.
Ever want to take a hacksaw to your electronics? Well, you’ll need to if you want this project to look neat. Switch and pot shafts are always too long, and you must cut them to length to fit your chassis and the knobs you select. Mark the desired length, clamp the free end (not the component!) in a vise, and saw away. Clean up the rough edges with a file.
Wiring for the switchbox portion is all straightforward point-to-point. There’s no need to use shielded wire internally, since the box will provide adequate shielding.
If you use jacks with metal mounting bushings, the cable shields for the inputs and outputs will be adequately grounded. Connect the ground terminals of the pots and mute switch together by jumpering from one to the next, and connect the string to the ground terminal of one of the main monitor output jacks. That will become the system ground point for the power supply, and also for the headphone grounds.
The headphone amplifier and the power supply can be constructed on perf board, using a separate board for each. Be sure to buy the type that has a hole spacing of 0.010" so the ICs will fit properly.
Lay the components out loosely on the board to get a sense of the size, then remove the parts and cut the board to size, leaving room around the border for the input, output, and power terminals, and for some mounting spacers. Note the polarity of the electrolytic capacitors. On small electrolytic capacitors, often it’s the negative lead that’s prominently marked.
Construction of the power supply isn’t critical, other than observing the same cautions as above for the capacitors. Diodes are also polarized and must be installed in the proper orientation. The band around one end of a diode corresponds to the perpendicular line on the schematic toward which the arrow points.
The IC regulator pins are numbered from left to right with the heat sink down and the pins pointing toward you; the metal tab with the mounting hole is electrically “hot” as well as being the heat sink. Since this is a pretty low current device, you can get away without heat sinking it to the chassis and just letting it dissipate heat into the air. If you choose to mount the regulators to the chassis, however, you’ll need insulating mounting hardware so that you don’t ground the heat sink.
It’s your choice whether to build the power supply into the switch box or make it external so that you can use it to power other problem solvers. This one will supply about 100 mA, adequate for the appetite of a few small signal devices.
Electrolytic capacitors (as used in the power supply) are made either with both leads coming out the bottom or with one lead coming out each end. Take your pick depending on how you choose to lay out the power supply. You can use a capacitor with a higher voltage rating than what’s specified. It won’t be any better, but it’ll be larger. Don’t use a lower voltage rating one, though. Whichever type you choose, don’t put them in backwards or you’ll let out the smoke.
Operation should be pretty obvious once you’ve studied the diagrams. Connect your monitor amplifier or powered monitors to the monitor output jacks, sound card to its input and output jacks, and anything else that you have to the remaining jacks.
For starters, bring up a recorded sound file, start it playing, switch the monitor selector to the Sound Card position, and bring up the volume control. If you hear nothing, make sure the mute switch isn’t in the “mute” position. Also, be sure you’ve wired the volume control so that the volume increases when you turn it clockwise (unless you’d prefer it the other way).
If you included the headphone amplifier section, plug in the phones and check it out. Now try out all the other routings and enjoy.
When recording, the normal mode is to monitor the sound card output. The other monitor sources are for convenience—if you just want to listen to the playback of some other sound source or want to work out a keyboard part without fooling with the computer.
When recording with the microphone and monitoring on headphones, remember to mute the “input monitor” of the sound card using the software control panel. Otherwise you’ll hear it twice, with comb filtering guaranteed. If you’re recording a keyboard part and monitoring on speakers, then you’ll want the sound card’s input monitor unmuted so you can hear that along with the playback of the backing tracks.
There’s no reason why you can’t double up everything for balanced inputs and outputs if that’s what you have. You’ll need 4-pole switches and a quad pot for the main monitor volume control.
The switches will be easier to find than the pot. Noble makes quality quad pots, but they’re expensive (www.lartronics.com is a distributor). For a balanced version it’s okay to keep the headphone amplifier unbalanced and make connections accordingly. I rationalized the executive decision to illustrate an unbalanced unit on the basis that if you’re serious enough to be using all balanced equipment, you’re probably due for a patchbay and a more sophisticated headphone cue system anyway.
This is a simple project on the surface, but there’s a lot of utility to be had and you can customize it to suit your needs—or just build the parts that you need. If this sounds like a great idea but more work than you are able to do, there are a couple of commercial products that might fit your needs. The Furman Sound SRM-80 has most of the features of this project, plus adds switching for multiple monitor speakers and includes a meter. The Vergence Audio PVC (Passive Volume Control) is a volume control only, but it works with both balanced and unbalanced equipment.
Build or buy, I hope this has stimulated your thinking about monitoring and cleaning up your act.
Mike Rivers may be contacted via email@example.com.
Hard-to-find parts list
S1, S2, S3 Two pole, three (or more) position rotary wafer switch Radio Shack 275-1386A
R1 - R4 100 K Ohm dual audio taper potentiometer Radio Shack 271-1732C
S4 Double pole double throw (DPDT) toggle switch Digi-Key CKN1032-ND
IC1 - IC3 OPA2604 DIP Operational Amplifier Digi-Key OPA2604AP-ND
IC101 7915 Negative 15V regulator Digi-Key LM7915CT-ND
IC102 340T Positive 15V regulator Digi0Key LM340T-15-ND
7815 is an equivalent part, but older part number. It’s OK to use that instead of the 304T-15
Generic parts (Radio Shack will do)
All fixed resistors are 1/4 Watt metal film, 5% or 1% tolerance
R5-10, 21, 26 4.7K Ohm
R13-20, 23, 28 100 K Ohm
(Buy two Radio Shack 271-309A 1% metal film resistor assortments for 4.7K and 100K resistors)
R22, 27 22 K Ohm Radio Shack 271-1339
R24-25, 29-30 82 Ohm Radio Shack 271-1107
R11, R12 10K Ohm linear pot Radio Shack 271-1715
C1-2, 4-5 22 uF 25 V electrolytic Radio Shack 272-1014