Xaudia offer microphone re-ribboning and repair services.


MOTM - Cadenza Rocket ribbon mic

Sometimes I see patterns or trends in what arrives on the bench. A year ago it was RCA 74s, and in spring 2011 it was Electrovoice ribbon mics. And then the summer brought Melodium 42bs. Of course these are just statistical anomalies or 'blips' in the random noise of what my customers send me to repair, but it does at least suggest ideas for the regular 'Microphone of the Month' column.

Cadenza microphone set with box, documents and stand

This month these little Cadenza ribbon microphones are in vogue - these were made by Simon SoundService Ltd in London, and were designed by Eric Tomson, Stanley Kelly Peter Bell. The mics have a very 1950s styling, and are often called 'rocket' mics.  However, having spoken to a couple of customers, the consensus was that many of these microphones no longer sound as good as they should - or at least as good as they look. The complaints are that they are noisy, with low output. 

The Cadenzas have dual impedance outputs, and can be wired for either 30 ohm or high impedance  (80 KΩ) output by changing the wiring at the connector. Neither option is really ideal for modern studio use, where something between 200 and 600 ohms is much more common for a mic output.

The ribbon itself is slightly unusual in that it has a fixed or nodal point half way along, where the ribbon is glued to an insulated support. This may well have been designed to reduce the likelihood of the ribbon being stretched, and one can imagine this either as two ribbons in series, or like a guitar string where the octave harmonic has been struck.

Cadenza mic transformer under the knife

From a technician's viewpoint (i.e., my opinion :p), there are a some weak points in the design which all relate to the transformer. Firstly, the ribbon clamps are connected to the transformer simply by winding the wire under a nut and tightening it - really not a reliable long term approach. Secondly, the transformer wires are extremely thin, and half a century later the insulation becomes brittle and tends to break, with disastrous results. It is more common practice to use thick gauge wire for the fly-leads to the primary, to keep resistance and noise to a minimum. And finally, as these thin wires become old and oxidised, the mics become noisy.

This all points to a transformer upgrade or rewind. However, it seems that Simon Sound Services Ltd. did a better job of encapsulating the transformer than they did of connecting it to the ribbon. The tranny is glued inside a mu-metal can with a hard, clear resin that resists removal. It was necessary cut the metal shell in half and then to soak the resin in dichloromethane for 48 hours to remove the resin. This allows removal of the laminations and a better inspection of the transformer windings.

Cadenza windings - primary (left),
and 30 ohm output (right)

The outer high impedance winding is a very fine gauge (approx. 0.07 mm), below which is a single layer of 0.4 mm wire which makes up the primary. The innermost winding is the 30 ohm output, which is approximately 0.2 mm in diameter. Unfortunately the solvent also damaged the bobbin, and and a replacement was found to make a new transformer using the vintage laminations. 

The new transformer has a single 600 ohm output, with thicker gauge wires to give lower resistance.  It took some effort but the result is a higher output with a lower noise floor, and the mic is much more usable. I'll post back with a 'before' and 'after' frequency plot for comparison.


Reslo transformers vs the WEE monster

This little dinosaur sculpture was made out of waste winding coils, stripped mostly from old Reslo and other microphone transformers. :)

So why are we stripping transformer coils?

Some older ribbon mics were originally wound for 30 or 50 ohm output impedance, and tend to give a low output level when connected to modern recording equipment.

Many of these low impedance mics, including Reslo and Tannoy ribbons,  can be rewound for a modern 200 or 600 ohm input, raising the output to a more useable level, and avoiding noise from having to crank up the preamps.

Bobbin from Reslo transformer with secondary winding removed.

The old Relso 30/50 ohm transformers have an inner (primary) winding consisting of just 12 or 13 turns of thick (0.8 mm) enamelled wire, and a secondary winding of 152 turns of 0.4 mm wire. The thick wire of the inner winding ensures that the primary resistance is low, which keeps noise to a minimum.

Reslo transformer rewound for 600 ohm output

Re-winding the transformer involves removing the outer winding from the original transformer and replacing it with sufficient turns of a thinner gauge to reach the desired turns ratio and output impedance. Usually the original primary winding can be kept in place. The transformer is then reassembled and dipped in wax to fix the windings and lams in place.

This makes the microphone much more usable in a modern studio - transformers can be would for 250Ω, 600Ω or any other desired output impedance.


Syncron AU7a revisited - Phantom power

A while back I wrote about the Syncron AU7a FET condenser microphones, which are sometimes badged as the Fairchild F22.

In that post I had sketched out the schematic. I have since converted one for a customer to run on phantom power, and spotted a glaring error in the schematic. Here is the revised version...

The transistor is of course a P-channel JFET, and the battery polarity is reversed, giving a positive ground. The batteries are switched off when the plug is disconnected, and the routing through the plug makes tracing a little tricky - that was my excuse anyway.

All of this means that some small modifications are needed for phantom power use, because negative ground is by far easier to implement. Using an N-channel JFET makes things much more straightforward - something like this...

The 'adjust' resistor is tweaked for best response to a sine wave applied across the head amplifier, and in this case the result was around 1kΩ. JFETs can vary quite a lot, and it is sensible to adjust this individually for each mic.

I built a small breakout board to supply the required voltages from the phantom power. The board fits neatly in the battery compartment.

The "110K" is again adjusted on the bench to ensure that the voltage is correct under load.

There is one more thing to note - now we have switched to negative-ground and an n-channel device, the output cap needs to be flipped round.

Here's a measured frequency response plot for the modified mic (the dips at around 150 Hz and 600 Hz are likely to be room modes)...

The microphone works perfectly, and it is nice to hear one brought back to life after all these years!