by Jason M. Neal
|Preamp kits are no longer available. As much as I hated to do it, I was unable to devote the time necessary to support the project. This page is here for reference purposes. I would like to thank everyone who has sent kind words. I will return to loudspeakers again someday.|
Around the time Eric Wallin announced his design of a DIY microphone preamplifier, I began to think about developing a printed circuit board to complement his work. So I contacted Eric and he was very receptive to the idea. I figured there were quite a few DIYers who were apprehensive about building the preamp using point-to-point wiring. Eric has done a wonderful job providing layout diagrams for building it on a prototype board. However, a PCB effectively eliminates the wiring guesswork, and perhaps represents a more elegant implementation for those who wish to build several for evaluating loudspeaker polar response and the like. Assembly is quick and easy. Anyone who can solder a few connections and drill a couple of holes in a project box should have no problem whatsoever. Moreover, commercial units such as the Mitey Mike cost well over $100. Need several? Then be prepared for a significant financial outlay. This mic preamp can be built for around $30, and maybe less if you happen to have some of the parts lying around.
The project took longer than expected, but it is finally here. The primary design goals were compactness and simplicity. The PCB measures 1.7" x 2.3" (43 x 58mm), or about half the size of a credit card. A great deal of effort went into getting all copper traces onto a single-sided board. No jumper wires are used; the only wires exiting the board are those to the panel-mounted components such as the switch and input/output jacks. A single-sided board not only simplifies assembly but also makes it easy for those of you who would like to etch your own boards. All copper traces are a fairly heavy 30mils, so you don't have to be a pro to pull off etching your own.
I have decided to offer several different options to potential builders. Some may only need the PCB, whereas others may wish to get a full preamp kit. Or buy nothing at all, my feelings won't be hurt :). The PCB artwork is free for the taking. I've already expressed to Eric that I do not wish to make money on this project. Doing so would be in poor taste considering it is based on Eric's design. I do this because I enjoy it (pretty sick, huh?). The kit price is merely intended to cover my costs.
Parts for the preamplifier are available separately including the pcb, components, case, connectors, and even the microphone capsule. In many cases, it may make more sense economically to obtain all the parts from a single source. Shipping is often expensive, so the cost to build a preamp could easily double if your must purchase your parts from several different sources. Alternatively, you can purchase parts at your local electronics supplier such as Radio Shack, although the typical selection and quality leave something to be desired. In an effort to obtain quality parts and keep costs down, I am sourcing parts from approximately 5 suppliers.
A couple of different operational amplifiers have been specified for the preamp. The original plans call for the popular TL074, which is an excellent low-cost amplifier in terms of distortion and slew rate. However, a few builders have expressed interest in a higher performance op-amp. Therefore, the LM6134 has been made available with the kit, which offers wider bandwidth and lower power consumption (approx. 75% less than the TL074). This op-amp is naturally more expensive, though under regular use it can pay for itself with increased battery life. It is offered as an optional upgrade along with a precision voltage reference for the mic's bias supply. Some relevant specifications for the TL074 and LM6134 op-amps are given below:
Table 1. Comparison of operational amplifiers
(nV / Hz^0.5)
If you are sourcing parts yourself, you can undoubtedly find other suitable quad op-amps of varying price and performance. Simply check to be sure your choice is pin-for-pin compatible with those mentioned above. Just as an example, the LM6144 (brother of the LM6134) exhibits wider bandwidth and slew rate, while power consumption falls roughly between the TL074 and LM6134.
Additionally, the LM4040 voltage reference included with the op-amp upgrade improves regulation of the microphone bias voltage. It takes the place of diodes D1-D3, which exhibit a temperature-dependent voltage drop as diodes are wont to do. The voltage reference is spec'd at 2.5V +2%, and in conjunction with the first op-amp stage, maintains a very stable 4.0V bias over the useful life of the 9V battery.
Below is a revised copy of the schematic. Note that the components used in the upgrade are identified in parentheses. Also, a composite (layered) view of the printed circuit board is available here . When assembling the circuit board, it will be helpful to print out a copy of the schematic for cross-referencing the component values and designators. Please use this high resolution image of the schematic for printing to ensure the text is legible. It is best to save the file and print from an external viewer which allows you to fit the image to the page and print in landscape orientation.
Printed Circuit Board
Soldering the components to the circuit board is the first logical step. I recommend the smallest solder (~0.02" or 0.5mm) you can find. It flows quickly which helps prevent overheating the components. It is typically easier to solder the "shortest" parts first; that is, those which lie closest to the board. This way, when you turn the board over to solder, the parts aren't likely to fall out. Here is the recommended order:
The assembly is slightly different depending on which operational amplifier is used. Provisions have been made to the circuit board to accommodate either of the recommended op-amps. The basic schematic is based on the original TL074 design, and the changes necessary for the LM6144 are indicated in parentheses on the schematic, and are described below:
Along those same lines, exercise a little care when soldering in the IC socket to avoid melting the plastic body. You must make 14 solder joints in close proximity, so allow time for cooling after completing each joint. It also helps to skip around (more or less diametrically, like you would tighten the screws on a driver) as you solder to prevent heat buildup in a concentrated area.
The leads of the transistor (Q1) and voltage reference (Q2, if used) must be prepared before insertion into the board (see Figure 1). Using a small pair of needle nose pliers or even tweezers, bend the center lead toward the curved portion of the plastic case. The lead should have two distinct "knees", the first about 2mm from the device's body and the second about 2mm from the first. Take a quick look at the PCB to get an idea as to how far the lead should be bent. Note that there is no silkscreen outline on the PCB for the LM4040 (Q2), but its solder pads are integrated into the area normally occupied by diode D2.
Figure 1. Transistor lead preparation
If you are using a transistor other than the 2N3904, check the manufacturer's data sheet to ensure the pinout is compatible. Almost any small-signal transistor will suffice for this application, but may require that the pins be formed differently.
Figure 2. Panel drilling guide
I prefer to use a drill press and brad point drill bits to drill precise holes in the plastic mounting panel. Otherwise, the drill bit tends to "creep" and the resulting hole is offset from the desired location. However, you can get near-perfect results with a hand held drill and standard bits if you are patient. Use an awl or similar tool to mark the centers of the holes to help guide the bit. It pays to practice on a scrap piece of plastic to get the technique down and also to make sure the hole sizes are optimum. If something goes horribly wrong, don't panic. Since the switch panel is removable, you can easily make a replacement from a piece of scrap plastic or sheet metal. Once all the holes have been drilled, go ahead and mount all the components to the panel.
Take the bottom half of the project box and locate the two plastic dividers that separate the 9V battery compartment from the rest of the enclosure. Drill a hole approximately 1/8" in diameter in one of the dividers. Later, you'll be passing the battery clip wire through the hole and tying a knot in the wire to form a strain relief, but don't do it yet.
Solder short pieces (~3" long) of #24 or #26 solid wire to the labeled output pads on the circuit board, denoted as LED, OUT, SW, ATT, MIC, and PWR. Also, solder a slightly longer piece (~5") to one of the GND pads (there are two on the board). Carefully route the GND wire such that it passes through each of the ground tabs on the two panel-mounted connectors and the cathode of the LED (the shorter lead). Solder these three connections. Then solder the MIC and OUT wires to the center pins of the two panel-mounted connectors. Note which wire is soldered to each connector, and label them on the case so you'll know where to connect your mic probe and output cables later when you're busy testing loudspeakers.
Now is time to wire up the switch. Using Figure 3 as a guide, solder each of the corresponding wires (ATT, SW, and PWR) from the pcb to the switch. However, do not solder the battery clip's red wire to the "to battery +" connection just yet.
Figure 3. Switch wiring
It may be necessary to solder a short piece of red wire onto the 9V battery clip to make it long enough to reach the "to battery +" connection on the switch. A small piece of clear heat shrink is included in the kit to insulate this connection. Locate the unused GND pad on the pcb and take a look at the black lead on the battery clip. The object is to tie a knot in the wire to form a strain relief, while also leaving plenty of the clip extending outside the case for connecting the battery (see Figure 4). Once you have a good, tight knot in the desired location, solder the leads to the switch and GND pad on the pcb.
Figure 4. Forming strain relief
Install the op-amp into its socket, observing the proper orientation (match the notch on the op-amp with that on the socket--not that I've ever made that mistake, of course...). Then slip the pcb and the switch/connector panel into the case and screw the case together. If the case does not close completely at first, do not force it. There are four short plastic stubs located on the top half of the case that may be touching components mounted to the PCB. Simply cut them off with a pair of diagonal cutters. You may want to put a thin piece of foam in the battery compartment to prevent the battery from rattling around. I recommend a piece of acoustic foam which has the "egg crate" convolutions removed with a long boxing razor. You're a speaker builder, right? Now you know why you've been saving those little scraps.
Well, that's it. Now wasn't that fun? No chasing those nasty little jumper wires around or counting pin numbers on the IC. Now let's prepare the microphone wand.
For a really nifty appearance, sand the tubing with 400 or 600 grit sandpaper to give it a nice sheen. Lightly chuck the tube into a drill and hold the sandpaper around the tube as you depress the drill's trigger. Then clean it with a solvent such as mineral spirits to remove the sanding residue, and spray the tube with clear spray paint to preserve the finish. Or you may prefer a classic matte black finish, or whatever. To hold the wand while spraying, drive a nail in a piece of scrap wood so that it sticks up out of the wood by about 2" . Slip the tubing over the nail so it stands on end, and apply several light coats of spray paint. You don't really need a $4 can of paint. The $0.99 house brand from Wal-Mart works just fine.
Prepare a length of coaxial cable, the exact length of which will depend on your testing setup. I use a 10ft (3m) cable to connect my microphone to the preamplifier, and another 10ft cable connecting the output of the preamp to my computer. Fish the cable through the wand, leaving enough cable extending outside the straight end of the wand (the end opposite the 45 degree bend) to solder the mic. If you have trouble, squirting a little WD-40 or similar lubricant into the wand may help.
Twenty foot lengths of coaxial cable being offered with the kit. It is dual-shielded (foil plus braid) and has a flexible construction which makes it easy to manipulate when setting up loudspeaker tests. It is 3.7mm (0.15") in diameter so it can be easily routed through the recommended brass wand. My preference is BNC connectors because they virtually eliminate loose cables and intermittent connections when moving things around. On the other hand, RCA connectors form a forgiving break point in case the cable is suddenly pulled, so the preference is ultimately up to the user. Both RCA and BNC male connectors are available with the kit. The RCA plugs are the familiar solder type with cable strain-relief clamp, while the BNC plugs are of the crimp/compression variety...if you are unfamiliar with compression BNC plugs, your first encounter could send you into an anxiety attack--each connector consists of six pieces. Therefore, I've prepared a BNC assembly diagram to help explain things.
Continuing on with the wand assembly, coil the tiny microphone leads around a pen or pencil so they form the shape of a spring, as shown in Figure 5 (or see picture). This coil provides a little slack in positioning the mic and coaxial cable into the wand to prevent damaging the mic's delicate leads. Strip about 1" of the coaxial cable to reveal the two conductors and solder them to the microphone. Either tape or place heat shrink tubing on the connections to prevent a short circuit and also to protect the stiff wire from fracturing near the solder joint. Carefully push the microphone capsule and coaxial cable back into the wand while trying to maintain the coiled mic leads. Before gluing the capsule into place, perform a dry fit to ensure it fits snugly into the the end of the wand. If it fits very loosely, you can build up its diameter with a few turns of electrical tape. When you're happy with the fit, apply a dab of epoxy around the inside perimeter of the wand opening and the back side of the microphone capsule. Then push the capsule into the end of the wand, being careful not to get any epoxy on the microphone's protective black screen. While the epoxy cures, suspend the wand in a vertical position with the mic pointed at the floor so the epoxy will settle at the capsule's back and yield a strong bond.
Figure 5. Microphone wand construction
Form a cable strain relief by slipping a 3" piece of 3/8" diameter adhesive-lined heat shrink tubing over the wand, positioning it so approximately half of it covers the wand (near the 45 degree bend) and the other half covers the coaxial cable. Then shrink it into place with a heat gun. A hair drier will work in a pinch; just place it on its highest heat setting and hold it very close to the heat shrink. The heat shrink will collapse to approximately ½ of its original diameter, if you anticipate it will not shrink enough to properly grip the cable you're using (i.e. the cable is less than 3/16" in diameter), wrap a few turns of electrical tape around the cable near where it enters the wand. The tape will build up the cable's diameter so the heat shrink will hold it securely. When the tubing has properly shrunk, a small bead of adhesive will squeeze out of the ends of the heat shrink. You now have a durable and functional microphone wand which may be easily mounted to a tripod for loudspeaker testing.
An alternative method of constructing the mic wand involves mounting some sort of connector to the end of the wand. The advantage is that if for some reason the connecting cable is damaged, it may be easily replaced. An RCA jack seems like a logical choice, but be sure to use quality connectors or the strain exerted by the weight of the connecting cable can cause it pull apart unexpectedly. An XLR jack is a pretty good choice since the barrel locks onto its corresponding plug. However, the "bulk" of an XLR connector potentially represents a large reflecting surface, primarily at high frequencies where the plug's dimensions approach the wavelength of sound. Simply wrapping the connector in sound-absorbing material will do well in masking these reflections. The "direct cable" construction discussed previously does not exhibit this characteristic, and assembly is generally quicker and easier. The caveat is of course replacing the cable in the event it is damaged, so if you anticipate your dog chewing it in half or running over it with a forklift (?), you might want to use a removable connection.
Testing and Troubleshooting
Once the preamplifier and microphone wand are assembled, a few quick tests using a minimum of test equipment can help determine if it is working properly. When the preamp is first switched on to the high gain setting, the clipping LED should illuminate for approximately 1.5 seconds while all of the circuit voltages stabilize. If switched to the low gain setting, the LED will deliver a quick flash of about 1/4 second. In either case, the LED will deliver a very brief blip when the switch is returned to its off position.
Now check the voltage at the mic input jack with a basic multimeter. The bias voltage should measure approximately 4.0V. Similarly, with no mic plugged in, the voltage at the preamp's output jack should measure very nearly 0V. Now plug the mic into the mic input and switch the preamp to its high gain setting. Blow gently across the mic, gradually increasing the force until the clipping LED illuminates. Then throw the switch to the low gain setting and blow across the mic again. If all is working properly, note that it requires considerably more force to cause the clipping indicator to light.
If each of these tests check out OK, we can proceed to stage two. Here's where the fun begins! Plug your mic into the preamp, and run the preamp's output into one of your stereo system's line-level inputs. Be sure to activate the appropriate input on your equipment's selection panel. Now you're a vocal artist . . . if only in your own mind. Instant gratification!
You can now begin setting up your loudspeaker tests. Everyone
seems to be using a different software package, so I'm afraid I do not
have the resources or the experience to offer details about using each
system. I will try to provide a list of links to pages and/or
discussion forums which have information about the more popular
measurement systems. Eric has written a detailed introduction/tutorial
to Audua's Speaker Workshop,
prompted the design and development of this preamplifier project.
Audua maintains a discussion
board for its software with which new users should probably
familiarize themselves (also see the archived messages contained here.) In the
meantime, please email me
your suggestions and links to web resources which may prove useful to
DIY loudspeaker builders.
If for some reason the above tests did not turn out as planned, please check the following areas:
These simple inspections will reveal probably 98% of all problems. If you've performed these tests but still cannot identify the source of the problem, you can send your questions to me. If you need help, please include a detailed description of the problem and the results of the tests mentioned above. Perhaps even better (and usually more prompt :) is to have a friend who is experienced with electronics take a look at your project. A visual inspection will almost always identify problems quickly and accurately.