Microphone Input Transformer Characteristics

Right now I know very little about transformers, especially the mic input ones. I have used them before and I am familiar with a few characteristics like impedance so I can more or less match a transformer with an amplifier. It is my intention to understand more than that. This is what I have gathered about mic input transformers so far. They have:
1. Low primary DC resistance for little insertion loss
2. High permeability alloy lamination core material
3. Static shield between windings to reduce capacitive leaking
4. Relatively small core size, usually less than a square inch
5. Mu shield encasing in many cases to reduce external magnetic fields .

I know these are very rudimental traits of these transformers, but I have to start somewhere. :grin:

Here is a question I am seeking your help with:
- What sort of winding techiques are used in these transformers? (reverse winding, alternate winding, random winding, etc.)

Thanks,
Tamas

1. Low primary DC resistance for little insertion loss
2. High permeability alloy lamination core material
3. Static shield between windings to reduce capacitive leaking
4. Relatively small core size, usually less than a square inch
5. Mu shield encasing in many cases to reduce external magnetic fields.

Since a mike amplifier often has a very high input impedance, #1 isn't really about "insertion loss". In a big power transformer we want DCR around 2% of nominal impedance to reduce power loss, but in a typical mike transformer application even DCR as high as 50% of nominal impedance won't reduce output much.

50% DCR will however raise noise resistance to about 2dB Noise Figure. So we can be a little relaxed, but not careless.

#1 and #2 go together. For a given size, permeability and bass response, you need a certain number of turns. For a higher permeability, you can do it with less turns of fatter wire, reducing DCR automatically. Actually the key spec for high impedance windings is leakage inductance, which increases with number of turns; a hi-perm core extends the range between bass fall-off and trouble in the treble. With Permalloy you may get low DCR without trying. However Permalloy is expensive, so you tend to reduce core size and cost (hence #3), and DCR comes back up.

It is not an easy optimization.

#5: The mike transformer is generally the lowest level in the whole system. And systems have volts of signal and hundreds-volts of power laying around. So crap gets in the mike transformer. Distance is the cheapest shielding, and usually enough isolation for line transformers. Mike transformers very often need magnetic shielding.

#3: In an ideal transformer, we turn electric energy into magnetic energy, and turn that back into electric energy. The two sides of the transformer may be at very different voltage potentials, but that won't couple across the magnetic path. But to get good magnetic coupling, the two windings have to be very close together. That adds little capacitors from one winding to the other. Now any voltage potential on one winding will couple to the other, and usually in unexpected ways. If you foolishly wind a simple primary over secondary winding, and foolishly connect the outside of the secondary to the grid of the amp, at high frequencies all the common-mode voltage on the primary couples right into the grid. In this case, reversing the connections on the secondary couples the capacitive crap into the grounded side of the secondary, while the grid sees only the crap on the core which can be grounded and quiet. It gets more complicated when, to reduce leakage inductance, you interwind P-S-P-S-P so more of the primary inductance drives the secondary and grid (low leakage inductance); now you have capacitors all over and there is no way to phase them to avoid capacitive leakage into the grid. You can put electrostatic shields between windings, but that may raise total capacitance and lower the upper resonant frequency.

And mostly, low-impedance transformers like 150:1,000 don't have huge problems. They only need around a thousand turns, so insulation doesn't take up a lot of space like when you have many-thousand turns (you can make copper smaller, but there is a minimum thickness of varnish to avoid shorts). You can easily give them a hi-Z load so the leakage inductance is not too bad for treble. The capacitance is proportional to overall size, not number of turns, and may be 300pFd; this is "small" compared to 1,000 ohms but large next to 50K ohms. Yes Deane and crew worked hard on a "simple" 200:800 design, wringing every last bit of performance. But very simple techniques can give very-good performance at low impedance. Things get much harder over 10K nominal impedance, so much that compromises must be picked.

> What sort of winding techiques are used in these transformers? (reverse winding, alternate winding, random winding, etc.)

Depends mostly on the impedance. If you don't need extreme bandwidth, at low Z you can throw them on any old way. At high impedance, you may have to tediously compute several alternate winding techniques, and then compromise with what your winding machine can really do well.

PRR,

Thank You for taking the time to explain these. I really appreciate your help. Based on your advice I decided that I am going to start looking into parameters for a 1:2 step up transformer as the first affair.

Cheers,
Tamas

> start looking into parameters for a 1:2 step up transformer

All theory is hogwash. Experiment, especially if you can do it cheap. Get a 120VAC:48VAC 10VA power transformer: it is close enough to 2:1, isn't big bucks, and you can always use it as a power transformer afterward.

Simply because the power transformer is so competitive, performance/dollar will be very high. You'll get perfectly decent sound over most of the audio band, at 1/10th the price of even a modest "good" mike transformer.

When used outside their original design, power transformer parameters vary widely. I have two functionally equivalent power trannies on my bench, but different construction and different audio response.

In a mike-amp, you can load the secondary with a very high impedance input, so the leakage inductance isn't an issue. If you stay small like 10VA, the capacitance is small. There is a self-resonance in the winding, but even on the "high impedance" (120V) side it may be above the audio band and heavily damped by core loss. Run the 120V side into like a 5534's input, no secondary resistor, and measure. Try again with a few-K loading resistor; it is pretty easy to get it as good as it gets.

Bass response depends on source resistance. You might think it would be good to 50/60Hz, but remember that the wall output is an infinitely-low source resistance. Also bass response depends a lot on level: power iron is optimized for rated line voltage and high flux (and high THD), in audio we have to work at much lower level and the inductance will be lower. A 48V winding driven from a 200 ohm source will probably be flat to around 20Hz-30Hz, but it may depend how much they have optimized the 48V 10KG inductance at the expense of 0.5V 0.1KG inductance.

THD depends on level and driving impedance and frequency. Since our level is very low (about 1/100th of saturation), it should be tolerable, though this is one place where a more high-strung iron alloy would give much better performance (at the expense of performance at very high level).

Use it with a typical mike and source, but lay it near a large power transformer, hum will be objectionable. Out in the middle of the floor, hum may not be bad. In practice, even with careful location, we have to wrap some iron around mike transformers. Try a soup-can, an electrical junction box with cover, or some iron pipe. Mumetal will give similar shielding in less space but higher cost. But keeping the magnetic shield small helps just because it catches less hum.

Wire the secondary unbalanced, set amp gain for nice levels from a mike. Then disconnect the mike, tie both primary wires together, and feed a 1V square wave or CD player output from that point to amplifier ground. You will get a lot of "tizzy" leakage from capacitive coupling. Just like connecting a 100pFd cap from source to amp input. Grounding the core may help some. Reversing windings may help. Using a split-bobbin transformer makes capacitive coupling very small, but makes leakage inductance very large, and for most audio the leakage inductance is the harder problem to solve. A compromise is thick insulation between windings to reduce capacitance. We can reduce capacitive coupling very small by putting in an electrostatic screen. A wrap of copper foil between windings, grounded, but with insulation at the wrap so it does not become a shorted-turn. An interesting variation is to wind a one-layer dummy winding between the real windings, and just bring out one end for grounding.

The 48V 10VA winding will have 20-30 ohms of DC resistance. This is "small" compared to 200 ohm mikes, so noise performance (thermal noise, not hum and buzz) may be very good. This goes with the marginal bass response: we might like more turns for better bass, though DCR will start to affect noise.

So playing with a very "crude" transformer will show you how you can get decent performance without any special design, and which things need to be changed to make it better for our purposes.

i've recently been using the 15va toroids as inexpensive line output transformers in an api preamp. they work well as a step down (get the highest secondary voltage unit you can for least step down, 22+22 or 25+25). also using one of the dual primaries as a secondary for a 1:1 output works almost as well. good bass response, in some cases better than the profile 2503 repro, and the high end roll off isn't bad.

the toroids don't sound the same to me as the iron profile 2503, or the cinemag 50% nickel cores, but they are worth using, especially for the cost. and if you don't like them, like PRR says, use them as power transformers!

ed

Great thread! Thank you all!

great,now substitute for mic input xfmr 

Audio electronics spambots? I thought this day would never happen.

good stuff!