A look at stand alone microphone pre-amps - part one

by Bob Hodas and Paul Stubblebine

In the beginning Man said "Let there be recording" and he created mic pre-amps and they were good. Well, maybe okay would be more appropriate. And so man said let there be EQ, compressors, gates and auxiliary sends and thus consoles grew and were praised. And many years later some men yearned for the good old days and re-invented the pre-amp, and said "Let it be good" and it was.

With the advent of the console that has everything in line -- with the possible exception of a candy machine -- some members of the recording community felt that going as "direct-to-tape" as possible gave them a better signal than passing through a vast number of op-amps and circuits. Paul and I happen to belong to that school of thought.

Tom Flye taught me to go straight out of the console pre-amp outputs to tape machine inputs at the patchbay whenever possible. Thus we avoided bus circuits and EQ and whatever else was not needed for the signal on that particular track. This technique worked well for getting a very clean signal on tape -- as well as full, pure sound -- leaving lots of flexibility for EQ in the mix. Tones were manipulated by microphone choice and placement during recording.

This is not the only reason that stand-alone mic pre-amps are becoming more popular. The age of sampling has made it necessary for many musicians to record samples or sound effects in remote locations. Such people do not necessarily need multichannel, full-feature consoles to do their sampling, and many musicians are learning that clean, quiet, non-equalized samples offer the greatest latitude for manipulation. At most only two channels are needed for this type of operation; stereo pre-amps fit the bill perfectly.

There also seems to be a growing group of engineers doing stereo-pair recording, be it with spaced omnis or MS pairs, etc. These are typically purist recordings and a straight-through, high-quality signal path is desirable. Ambience is also playing a larger part in modern music, and I know of several engineers who have moved stylistically from close-mic to the MS technique, coincident pairs, or spaced omnis for recording sections. For these people a pair of high-quality, low-noise pre-amps is ideal.

Evaluation sessions This article came about because we thought it would be interesting to audition to some of these pre-amps for our own purposes. We discovered that there were, surprisingly, eight stand-alone mic pre-amps currently in production, and a couple more in the planning stages. After finding out how many pre-amps were actually on the market, we felt that it would be a nice idea to have a greater number of engineers benefit from our efforts to amass the necessary equipment. We therefore contacted 17 San Francisco Bay-area engineers, who expressed a genuine desire to participate in a listening test. The response was encouraging and so we proceeded with the program.

The units tested were the Benchmark 4X4, Innovative Audio UTMP-2, Jensen Twin Servo 990 (manufactured under license for Jensen Transformers by Boulder Amplifiers), GML Model 8600, Sontec MB-1A, Studio Technologies Mic Pre Eminence, Summit Audio TPA2, and a prototype built by a manufacturer utilizing a card produced by Harvey A. Rubens Technology (HART) which may possibly be released. Apart from the latter unit, all pre-amps were production models supplied directly by the manufacturer. (The exception was the GML Model 8600, which was supplied by a local owner. Apparently there was no stock on the GML pre-amp at the time of our test.
We asked several other manufacturers to supply units but either they had no stock or the units were not yet in production. Firms that reportedly will have pre-amps available in the near future include Cello, API, Hardy Company, Sound Research and Audio Engineering Associates.

The pre-amps fit into three basic groups. The Summit Audio TPA2 and Innovative Audio UTMP-2 are based tube technology. The Jensen Twin Servo 990 utilizes input and output transformers, while the balance of pre-amps are transformerless. Let's take a brief look at each of the pre-amps.

* The Benchmark 4X4 is a 1U (single space) rackmount unit containing four mic pre-amp channels. Each channel features a continuously variable gain control knob, as well as a 20dB pad and phantom power switch. There are no peak LEDs provided on the channels. The unit has detachable rackmount ears and utilizes an independent power supply. Access is via XLR-type connectors on the back panel. Outputs are balanced.
Benchmark offers a couple of options that were incorporated into our test model. One modification allows channel 1 and 3 and 2 and 4 to be summed to two XLR outputs on the back panel, turning the 4X4 into a miniature mixer. A front-panel headphone jack with gain control pot may also be added to monitor these summed channels. Incidentally, Benchmark plans to add a 2-channel pre-amp to the line; the pre-amp card is also available to OEMs.

* The Innovative Audio UTMP-2 is a 2U tube pre-amp that will take mic- or instrument-level inputs. Each channel features a 3-position stepped input gain control (30/40/50dB) and a variable gain output pot. There is a switch that allows the user to utilize either the front-panel phone jack for instrument-level inputs or the rear-panel XLR-type mic inputs.
Each channel features an LED peak indicator but there is no provision for polarity reversal. Phantom power with LED indicator is independently switchable on each channel. The unit contains a Jensen input transformer with an output transformer being offered as an option. Outputs are unbalanced.

* The Jensen Twin Servo 990 is a 2U unit which, with an expansion chassis, can be configured as a half rack or full rack model with a minimum of two channels and a maximum of 12 channels. It is also available as a table-top unit. I must admit that the Twin Servo 990 was by far the most impressive piece of construction in the review batch. It was elegant, solid and all mechanical functions had a very good feel.
Each channel includes a peak-reading LED, variable gain pot, phantom power switch with LED, polarity reverse with LED, and a microphone source impedance switch with LED that selects either 20 or 150=. Rear panel inputs and balanced outputs are via XLR-type connectors.

* The GML Model 8600 occupies 1U and is available as a 2- or 4-channel model; two units may be driven off the same independent power supply allowing up to eight channels of mic pre-amplification.. The front panel contains only the stepped gain pot and peak LED for each channel. The back panel holds a phantom power and polarity reversal switch plus XLR-type inputs and unbalanced output connectors.

* The Sontec MB-1A is a 2-channel table-top unit with a plastic casing. Each channel utilizes stepped gain pots and features signal and peak LEDs. A polarity reversal switch affects only the right channel, while an EQ switch activates an optional equalization circuit. Phantom power is selectable only by taking the unit apart and moving an internal connector. Rear-panel connectors are XLR-type and the power supply is independent. Outputs are unbalanced.

* Studio Technologies Mic Pre Eminence occupies 1U of standard rack space and contains two channels. Each channel has an infinitely variable gain pot, signal and peak LEDs, polarity reversal switch and phantom power switch. Rear-panel inputs and outputs are via XLR-type connectors. A rear-panel switch selection of balanced or unbalanced operation.

* The Summit Audio TPA2 is 2U high and features dual-stage variable-gain control pots for each of its two channels. It will accept instrument-level inputs accessed by the front-panel phone jacks and mic-level by rear-panel XLRs. Each channel features a switchable pad, and phantom power is switched on simultaneously to both channels. There are no function or peak indicators on this unit. Outputs are balanced and a Jensen transformer is offered as an option.

Measurement results I've assembled two charts -- Figures 1 and 2 -- that I hope will make comparisons fast and easy. This approach seems simpler than giving lengthy descriptions. All measurements were performed with a Audio Precision System One test system connected to an IBM PC.

The pre-amps were all subject to the same test criteria. We conducted all measurements with 30dB of gain in the pre-amps; this value was selected as an average amount of gain one might use in a true recording situation. It may or may not be the most flattering for making measurements, but is certainly indicative of real life.
I chose to put frequency response information into the charts -- as opposed to printing graphs -- because most of the units exhibited a flat response to at least 20kHz. The exceptions to this were the tube pre-amps. This rolloff is not due to the use of tubes but rather to the dual-gain controls.

As you will note from Figure 3 and 4, the response became less linear as input gain increased and output gain decreased.(a situation that can be likened to guitar amp gain staging for distortion purposes). Best response is obtained by keeping output gain at maximum and adjusting only the input gain.

The linear phase measurement were taken, which represents pure delay; any curve or deviation from a straight line would indicate unequal delay over the frequency range -- an undesirable feature This measurement is most important to look for regardless of the total number of degrees of shift. Because of time factors, however, group delay and deviation were not calculated and these factors may be even more revealing. The graphs themselves are not that revealing since everyone was fairly linear except at the lower frequencies. For this reason and space restrictions they have not been included here; instead we will provide a verbal interpretation.

The Jensen Twin Servo 990 revealed the most linear phase response, exhibiting no nonlinearity in the low-end. The next best group of pre-amps that exhibited less than 4 degrees of low-end shift included the GML Model 8600, Sontec MB-1A, and Studio Technologies Mic Pre Eminence. The Summit Audio TPA2 and HART prototype units displayed about 7 degrees of low-end shift, the Innovative Audio UTMP-2 about 10 degrees and the Benchmark 4X4 about 13 degrees.

Turning to the tube pre-amps, phase measurements indicated similar results to the frequency response. in that the dual-gain stage relationship significantly worsened phase (Figure 5).

What was most interesting -- or should I say surprising -- were the Common Mode Rejection measurements. The CMR shows how well the pre-amps reject signals that appear at both inputs (pins 2 and 3) in-phase at the same amplitude. (This unwanted signal could be RF or inductance into the mic line from some outside source.)
Only three of the units, the Jensen Twin Servo 990, GML Model 8600, and Summit Audio TPA2, showed any signs of matching components in assembly, a fact that is demonstrated by the similarity in the curves between two channels. The rest of the manufacturers should take a lesson from these guys and pay attention to matching. If we buy a stereo pre-amp, it certainly should be a priority to have each channel sound the same, especially when doing stereo pair recordings. I wonder if these manufacturers are measuring each unit that goes out the door; I would really be surprised if there are any listening tests being done.

Which brings me to the reason we put this article together in the first place: Listening Tests! The big question will be whether the measurements correspond with listening to the pre-amps. Next month we will look at the results of such tests conducted on these units by Bay Area engineers. Be there!