Mic Preamp Banner

Here's a balanced input microphone preamplifier I'm working on

 

Intro | Circuit Description | Schematics | Construction | To Do | Photos | Links

Introduction
This design has been assembled from various datasheets, application notes and existing designs. It has the following features:

  • Balanced low-impedance input
  • Balanced and unbalanced outputs
  • Low noise, low distortion
  • RFI filtering and input protection
  • Low-cut (rumble) filter (-18dB/oct @ 50Hz)
  • Low rolloff filter for close-micing (-6dB/oct @ 150Hz and 450Hz)
  • Gain variable in 6dB steps from 12 to 72dB (or optionally 0 - 60dB)
  • Output level control
  • Overload indicator
  • +48V phantom power (switchable)

The design is still being revised (suggestions welcome! - see email link below) and has finally been built. I'm fairly happy with the design, but already there are some improvements to be made. This is not designed to be an extremely high quality pre-amp but has pretty good performance. My budget doesn't stretch to top quality mics by any means, nor high-power PC DAW recording setups, but I figure a good mic preamp wouldn't go amiss, and I can learn a thing or two about preamp design along the way.

You can download the schematics and PCB layouts from the links below

Circuit Description
The main gain stage is based on the INA103 (Texas Instruments/Burr Brown) instrumentation op-amp, which is a high quality and fairly expensive device. I'm using it because I got hold of some free samples ;-) There are various other suitable devices but I haven't looked too hard. As for the NE5532s, I already have some of these so that's what I'm starting with, but OPA series op-amps (and many others) have better specs - I may substitute a better device if the NE5532 performance isn't good enough, but then again I'm not trying to build a near-theoretical-limit amplifier...

I've chosen to have a stepped gain control because I already have suitable rotary multi-position switches. A variable resistor can also be used, but a special reverse log taper potentiometer is required to give good control, and these are not so easy to get hold of and probably not worth it unless you're manufacturing loads of preamps.

I've used bipolar electrolytic capacitors for coupling the stages. These are a compromise between performance and cost - ideally plastic film caps such as polypropylene should be used, but these are rather expensive and large... I haven't done a head-to-head of plastic film and the bipolar electrolytics, but I don't think the performance hit will be anything to worry about. The same applies to the INA103 gain control decoupling caps. Rane uses a 470uF polarised electrolytic in their MS-1b, but I'm not sure it's such a good idea to use a polarised cap with zero DC bias. I've gone for 1000uF (bipolar electrolytic) or so for slightly better low freq response at higher gains. The -3dB point at 60dB gain in the INA103 is about 55Hz (theoretical - the practical value is probably a bit higher). My PCB layout has space for 5 x 220uF bipolar caps because I can get these cheaper than 1 x 1000uF for a single preamp. Inserting a capacitor in series with the gain set resistor ensures the gain drops off at low frequencies and is unity at DC, which reduces rumble and noise a fair bit, and is probably more stable.

The power supply is fairly conventional and provides +/- 15V and +48V regulated. The phantom supply is an interesting design based on the Rane MS-1b (incidentally, Rane's website has a treasure trove of application notes, schematics and audio papers). It triples the input voltage to give about 70V unregulated - only a single 18V transformer is needed. This is regulated down to 48V using a standard LM317 regulator.  A series pass transistor and zener diode form a pre-regulator to ensure the LM317's 40V input/output differential limit is not exceeded. Extra smoothing is provided by the R-C filter after the phantom switch. This also slows the rise and fall of the phantom power to soften the switching transients. The power supply will provide power for 2 preamps if clip-on heatsinks are fitted to the regulators. The supply transformer is external to the preamp enclosure in my setup, like a wall-plug unit, but could be built in. I don't have to worry about mains voltages inside the preamp and earth loops/hum aren't a problem (they shouldn't be a problem with balanced connections anyway :-). The mains earth connection can easily be removed if desired.

The overload indicator circuit is nothing special. Full-wave peak detection of the signal level at two points triggers an LED briefly if the level goes above a preset level. The first sample point (OL_1) is at the output of the main gain stage to detect the level of the raw audio before any possible filtering - if there is significant low frequency content in the signal and you engage the low cut and rolloff filters, the output level could be much smaller and would not trigger the overload indicator (as sensed at OL_2, the second sampling point, after the filters) despite the input stage being overdriven. The peak level detector also takes account of any possible op-amp offsets, which may be a couple of volts at high gains. I used TL074 and 741 op-amps here simply because I happen to have a bunch of them - almost any quad and single op-amps should do, but check the pinouts. The overload trigger point should be set so the LED illuminates when the signal at OL_1 or OL_2 exceeds about +/- 10V, which is well within the supply rails so no distortion should occur. It's unlikely you'll ever need this signal level at the output...

Schematics and PCB layout
The project files are Acrobat PDF format with small GIF previews.

Main schematic (28K PDF) (GIF preview)
This is the first properly drawn and most readable schematic I developed. It's up to date, but the PCB layout was done in a different EDA package (This schematic was drawn using Proteus, and the other schematics and PCBs were drawn with Eagle, in case you're interested). This schematic shows how everything connects up, but the component numbering doesn't match the later schematic and PCB files. There are also several minor changes in the later schematics.

All schematics and PCB layouts (1688K PDF) (GIF preview)
These schematics correspond to the preamp PCB layouts. This PDF includes the PSU and preamp schematics, an interconnections schematic, the PCB copper foil patterns (suitable for photo-etching) and the component location diagrams. Unfortunately, Eagle doesn't generate nearly as elegant or compact output as Proteus...

Eagle Schematic and PCB files (103K ZIP, Eagle 4.09 format)
Just in case you want to play around with the schematics and PCB layouts, here they are. You can use these with the freeware version of Eagle, which can be downloaded from Cadsoft for nothing! What nice people. The freeware version restricts the size of PCBs you can lay out, which is why there are separate boards for power supply and preamp.

Make sure you download the tutorial (PDF file) as well if you haven't used Eagle before, since the interface takes a bit of time to get used to, and the online help isn't particularly helpful.

Construction
I built my preamp into a salvaged enclosure that used to house the power supply/charger for a Branson 104 Caliper (for measuring the thickness of sheet metals, paint coatings etc.). This small enclosure has a carrying handle on top, rubber feet on the base, neatly fits the two PCBs and has just enough front panel space for the controls/indicators. See the photos below. Any small metal enclosure can be used. Shielding and grounding is important. If you plan to put the transformer in the preamp enclosure, make sure it's well screened and away from the preamp input, and preferably use a toroidal type since these have lower stray magnetic field than E-I stacked core types.

To-Do
There are several (many, really ;-) possible improvements on this design that I've identified since building the first revision. I'll cover a few things I've discovered here.

Electronically balanced (or servo-balanced as various manufacturers call it) inputs suffer from poor common-mode rejection unless the microphone output and cable are very well balanced (impedance in each source output and each cable conductor must be exactly equal). Small differences in the impedances can drastically reduce the CMRR. Unfortunately, such imbalances are not uncommon in typical studio situations. Jensen Transformers have a bunch of application notes on this and other topics, one of which suggests bootstrapping techniques for improving servo inputs without using a transformer (not that I have anything against transformers, I just want a low-cost preamp). The bootstrapping raises the common-mode impedance from a few kilohms to very high values (megohms) across the audio range and reduces the sensitivity to cable and source imbalances, although it's still not as good as using a transformer. A quick Spice simulation showed that adding the bootstrapping gives a noticeable improvement in CMRR and reduces sensitivity to cable/source imbalances.

The output balanced driver could be improved slightly by using a single-chip balanced output driver such as the SSM2142 from Analog Devices. This should give better balancing, and better drive capability.

It may be more sensible to use frequencies like 75Hz for the cutoff, and 100/200Hz for the rolloff filters, depending on the mics/instruments you'll be recording. It's easy enough the calculate the required filter capacitor values using the standard filter equations.

Photos
A selection of crappy-webcam and ancient-digital-camera pics of the preamp
Preamp front panel
 The mic preamp in its enclosure. The case is approx 180mm (W), 100mm (H), 130mm(D) (7x4x5 inches) and the PCBs just fit in. The panel controls are (left to right): Gain switch, filter switches, output level, power LED, phantom LED and overload LED, and phase switch.
Preamp rear panel
 The rear panel. From left: Power socket, phantom switch (adjacent to power socket), line-out and mic in. Obviously, I haven't repainted the case yet... The plate in the centre hides a nasty hole left by the previous occupant - a mains power connector. Maybe the preamp will get a proper home one day.
Preamp power supply transformer
 This is the external transformer, salvaged from a dead set of Christmas tree lights. I changed the output cable to a three-core mains cable so the mains earth could be connected to the preamp chassis.
Preamp power connector
 Close-up of the power socket. This is a 3-pin DIN plug/socket. The phantom power slide switch can be seen better in this pic. Obviously, I haven't gotten round to painting the enclosure yet...
Preamp with front panel open
 The front panel of my enclosure unscrews to allow easy access to the innards. The PCBs are mounted on a tray that screws to the front and rear panels. The front panel is an offcut of brushed stainless steel sheet.
Preamp gain control switch
 This is the stepped-gain rotary switch. I mounted the gain resistors directly onto the switch terminals, and the two white wires connect the assembly to the PCB
Preamp PCB
 Excuse the messy wiring - I'm still twiddling things! This is the preamp PCB, with the INA103 just visible at lower left. The large, dark blue caps are the bipolar electrolytics. I made a mistake when ordering - these are 47uF parts, and are about twice as big as the 22uF parts specified. Doh! The two wire links at bottom right should be ferrite chokes (mic input). I was in the middle of experimenting when this photo was taken...
Preamp power supply PCB
 And here's the power supply PCB. The linear regulators at lower left have small clip-on heatsinks, which are definitely needed if you run two preamps off the PSU PCB. The input choke is at lower right.

Comments, hints, tips and questions on this design welcome!
Email Matthew Crowley

Links

Check out CrowleyWeb and the Propeller Clock (with instructions to build your own)

Need a cheap DIY microphone to match your preamp? Check out this one.

Also check out the SPC electric fiddles!

CrowleyWeb | MattWeb | Balanced Mic Preamplifier

Groovy Divider

© Matthew Crowley 2003
Last updated 1 June, 2004