Gear‎ > ‎Microphones‎ > ‎

u87Ai

Neumann u87, u87A, u87Ai

The Neumann u87 large condenser microphone's reputation for low noise, flat response, warmth, sub-sonic rejection in a multi-pattern design is well known and today's price remains high because of this performance; and name recognition. The original u87 design (1968-1974), which was a FET version of the u67, was improved in the u87Ai model (1967-1986) with more gain (a capacitance change) and 3 dB lower noise thanks to a capsule HV bias change to 60V resulting in an increase in sensitivity from 8mV/Pa (-42dBv) to 28mV/Pa (-31dBv). The 'i' indicated 'international' which resulted from the connector change from a DIN to 3-pin XLR connector. The u87Ai, the most recent version, hosts changes only to the electronics. The u87A contained its own DC-DC converter to power a K67 capsule in the figure-8 pattern as the two capsule sides were electrically isolated in the u87 design. The acoustics of the u87, u87i and u87Ai are all identical. The lower sensitivity of the u87 "classic" afforded a +5dB in max SPL (122 dBSPL, 117dBSPL for the u87Ai).

"Classic u87" vs "Modern u87Ai" Designs

The primarily subjective arguments about the "sound" of a Neumann u87 vs u87Ai are easily found on the web. What is interesting is diversity of opposing perspectives in what users hear; or believe they hear. This is especially remarkable given the fact that Neumann states that the capsules are acoustically identical and the capsule accounts for the majority of the mic's characteristics (assuming the mic preamp does not have some significant issue or its specs are not exceeded.

Neumann's original u87 design became the u87i (International) when they replaced the DIN connector with an XLR. Certainly this had no effect on performance. The u87Ai was only an electronics upgrade which actually improved the sensitivity by 10 dB and lowered the self-noise by -6dB. A wide variety of claims have been made, mostly that the u87 "classic" design is "better" even though technically the u87Ai actually has improved performance. [ The one technical exception is that in achieving 10 dB greater sensitivity, the MAX SPL is 10 dB lower for the u87Ai but unless you really need to use a u87Ai above +117 dBSPL, that difference will never be missed. Generally, higher sensitivity with lower self-noise is much preferable. ] The u87Ai gained ~2.1 dB from the +47V to +60V bias increase and the remainder by preamp design changes which reduce the capacitive loading on the capsule. Technically, there is no basis for the claim that the u87 is "better" than the u87Ai but there are substantial reasons to use 10 dB more sensitivity with 6 dB less noise.

Newport NA87

Neumann pioneered a great set of LCM features into the u87 and then improved and simplified the design with the u87Ai. Unfortunately, the purchase cost of such a mic is well beyond the range of most very small studios. However, by researching the technology behind their design, it is possible to recreate a similar microphone with the goal of stereo recording using two matching 'copies' in a Blumlein configuration. To achieve this, it is necessary to break down the components of that design to create the Newport Acoustics "NA87" version of the Naumann u87Ai large condenser mic.

Critical Components

Capsule Design: The capsule is the most critical element in determining mic performance and its complexity and precise construction and materials all play a part. This includes membrane material and thickness, hole patterns on both sides, insulating materials, membrane tension, plating quality, etc. While cheap Chinese 'knock-offs' might look very similar to a Neumann or other professional capsule, their performance can be drastically different. Peluso Labs sell an independent designed capsule, the P-k87i, as a "replacement" for the u87 capsule with an extremely similar response curve as shown below. Note that since the Peluso capsule, by itself, has a remarkably similar curve to the Neumann u87 microphone, the capsule is obviously the primary source of the acoustic characteristics of this design.



FET Preamp: The choice of FET, selection of FET for specific properties, circuit DC balance, part selection for noise and frequency response, not to mention the actual circuit design, all contribute to higher SNRs. Fortunately, the 2N3819 JFET device, which is used in the u87 is still manufactured by Fairchild and a few other companies. As with the u87, these JFETs must be hand-selected for performance and for matched-pairs but that is just a technical step. With a little care, the JFET preamp will not detract from the capsule's acoustic response which so closely mirrors the u87 design.

Transformer: Transformers are not all alike as the core material permeability, coil winding ratios, wire type, winding type, etc. all affect the frequency response and transient response. Fortunately, Peluso Labs make a 9.5:1 transformer (P-T87) with the same specs as the one in the u87. There are no winding options with the P-T87 as with the u87, but for the normal configuration with 200 ohm output impedance, they are not needed.

Basket: Many opinions have been promoted on the web about the woven-mesh "basket" which covers the microphone capsule and, unfortunately, all are completely subjective "listening tests" where the comparative recording setups were not, and could not, actually be identical. In addition, most were not even double-blind comparisons. The "general consensus" promoted that having a "single grill/mesh" is superior or that copying the shape of the u87 is superior or that the dimension either promote or destroy standing waves has little, if any, basis in physics and even the Neumann company attributes no special features or characteristics to their own basket design. The only supported engineering perspective, and the only one this author can believe, is the value in having the basket well grounded as an electrical shield as the impedance of the capsule is > 10^9 ohms.

Copies

If a U87 did exist in "kit" form would it really be a u87?
Yes, assuming careful construction. There is nothing magical about combining the components based on an accurate design and the use of exact parts -- but then, there's no reason for Neumann to do this.
Can a Neumann u87 be copied (some call it "cloned")?
Yes, nearly identical major component blocks can be assembled into a working microphone with very similar features.
Will such a clone be "just as good as" a Neuman u87?
Possibly, but measuring the copy's performance would be very difficult outside of a well-equipped acoustic test lab. However, a lower-cost u87-like performance might be "just the sound" you were looking for, so does it matter?
Is building a u87 clone worth it?
If you can afford, and need, and can justify owning a Neumann u87Ai, then buy one. However, for those not afraid of a small project, a copy can be a reasonable method to acquire a very good, low noise, multi-pattern large condenser mic. If a "matched" pair is built using the same parts and construction, the result will provide a figure-8 Blumlein Stereo array with exceptional linear spatial separation.
What is the relative cost?
Assuming your construction labor is free, about 20-25% of the retail price.
What would I need to buy?
P-K87i capsule, P-T87 transformer, 3-switch LCM mic body (eg. MXL-2010), PC board set plus electronics parts.
Why the MXL-2010?
This is a pattern mic with the same switch options, is mechanically nearly identical to the u87Ai and only costs slightly more than some Asian "knock-off / no-name" mic bodies which are less similar to a u87 design.
Where do I get all the info?
The internet; of course, it just takes a little digging. Or ...... keep reading for more details.

Example Build

Specific Components
  • Peluso: P-K87i capsule, P-T87 transformer
  • OnLine: MXL-2010 pattern mic for the body, switches and connector
  • Mouser Electronics: Various electronic components
  • Newport: N87Ai PC Board Set