Mating Microphones - ever wondered why your low impedance mic won't work well with your recorder? Then read on...
"Why can't I use my low impedance Shure direct into this Portastudio? This AKG mic is supposed to be 200 ohms, but the manual on the console says the microphone inputs are 1200 ohms. Will it be ok? What impedance, if any, should I choose when I buy my next mic?"
To answer these perplexing questions, and more besides, Ben Duncan homes in on microphone impedance relationships.
With a few exceptions, the division between serious (ie. professional) and not-so-serious audio is manifest in the impedance of the microphones you're using. The first mic you bought was most likely of the high impedance variety, priced at £5 or whatever. Impedance has no direct relationship to the cost of the mic though - it's just that low cost mics tend to go with budget gear, and budget gear prefers to work with high impedance mics for reasons that will shortly become apparent. Also, it's more than likely that your first mic was of US or Japanese origin, because high impedances aren't taken very seriously by European manufacturers.
Tying in with last month's episode on the perils of cable capacitance, let's refresh our memories by noting that high-Z (Z is shorthand for impedance) mics are bad news with long cables. Bad at least if you prefer to have any sort of top-end (treble) response.
For example, let's assume our high-Z mic is tied to 4 metres of cable, with a capacitance of 280pF/metre. That's 1120pF overall. The source impedance of our mic will be: at least 12kohms, so the signal will drop by 3dB or more at just above 12kHz. Basically, it spells goodbye vocal and guitar harmonics, and goodbye to cymbals...
So what are the advantages? The big plus for high-Z mics is that their output level is sufficiently high for relatively noiseless processing by simple amplifier circuitry, using low-cost parts. In other words, when the designer's budget is tight, economy can be found in the microphone input circuitry, by making it suit high-Z microphones.
Getting down to specifics, high impedance mics typically produce 14 times the output (+23dB) over their low impedance cousins. You'd be right to imagine - somewhat cynically - that this extra output voltage comes hand-in-hand with a loss somewhere else in the equations. It does: the source impedance is raised by the product of the voltage gain squared, viz. (142) = 196 times greater. So, thinking of mics where the impedance can be switched from low to high, the 150 ohm source impedance in the low-Z position rises about two hundred fold to 20,000 ohms when you switch over to high-Z. Now you can see that the high-Z mic's restrictions on cable length are as a direct consequence of attaining a higher output voltage.
High-Z mics need to see a high load impedance at the input socket of the mixing desk, Portastudio, or whatever. This statement may appear to be self-evident, but the reverse isn't necessarily true for low impedance mics - as we'll shortly discover.
With analogue audio equipment we always go for voltage matching, between equipment and transducers. For our purposes, this means we aim to transfer the output voltage of the microphone to the equipment, with negligible loss, rather than bothering about current or power transfer. Amazingly, these latter particular two parameters don't matter at all, even when good voltage matching pushes them down to infinitesimal levels, like nanoAmps.
Anyhow, as a rule of thumb, voltage matching means the load impedance (at the equipment input) should be at least 5 times - and preferably ten times - the mic's own, internal, or source impedance. Let's use a motor analogy: shifting into 4th gear at 25mph in a 950cc Fiesta will overload the engine, and ignoring the kangaroo leaps, the revs - analogous to the microphone's output voltage - will fall. With a bigger, 1600 XR2 engine (representing a lower source impedance), we can load the engine to a much greater extent without losing too many revs, but we still can't start off in 4th gear; the XR2 model merely offers us a pro-rata increase in the allowable loading on the engine.
Having established the concept of loading ratio, it's time to put it to use. The source impedance of most real, high-Z mics is in the 10 to 20k region - which ties in neatly with the earlier calculation concerning the impedance differences between high and low-Z models. Applying our 5:1 loading ratio, the minimum load impedance is therefore between (10k x 5) and (20k x 5), namely between 50k and 100k (100,000 ohms). Higher load impedances, as found for instance, on some valve gear, will do no harm, although they may be noisier than is necessary. The head impedances being discussed here are also known as rated impedances by manufacturers.
From time to time, you may come across medium impedance inputs, of 20k or 30k. Assuming these aren't really aimed at line-level signals, they're strictly a relic of oddball design from an earlier age. Alternatively, you may also find odd mics with rated impedances somewhat below 50k. This is compensated for if the capsule (source) impedance and/or output voltage are proportionately lower, but you're still left with excess load impedance if yours are of the conventional 50k or 100k type. The answer is to solder a resistor in parallel across the mic's output: with care, you can easily fit this inside the Cannon or jack plug. The exact value to use can best be discovered by experimentation.
The reason for being fussy about the load impedance, in this instance, arises because a mic with a non-standard impedance rating suggests the model is sensitive in this way. Back to the motor-load analogy: Before extra loading on the engine causes the engine revs to drop significantly, the characteristic resonance shifts. Put simply, the sound of the engine changes in line with slight changes in loading.
With moving-coil mics, fairly large changes in load impedance (-30% to +100%) can usually be accommodated without serious detriment to their sound. But, if a particular model is sensitive in this regard, the manufacturer may set a specific (if arbitrary) load impedance, against which the mic will be tuned to produce the intended response curve. In this instance, the mic will be more readily upset by 'strange' impedances.
Even so, with the smoothness of typical high-Z mics not being anything to write home about, smaller variations, say +/- 30%, shouldn't make much difference in practice, and we'll confine ourselves to noting the gross effects. Firstly, if load impedance presented by the equipment is too low (say less than 60% of that recommended by the microphone manufacturer), there may be a loss in top-end response. Secondly, if taken to extremes, with a load impedance equal to the mic's own source impedance - or lower - the mic's overall output will drop-off drastically.
If in doubt, you can test for this by temporarily soldering a resistor around 47k to 100k in series with a high-Z mic's centre (electrically positive) terminal. The output may drop, and the mic will also be noisier (more hiss), but if the top-end response 'picks up', you can be sure the load impedance presented by the equipment is too low for your microphone, so seek other arrangements.
For any serious work, high impedance microphones are best avoided altogether, even if the inputs on your present equipment suggests otherwise. Microphone impedances of either persuasion can be converted, but transforming the output of low-Z microphones to a high impedance - at the equipment end only please - makes most sense in the majority of circumstances.
If you must use high-Z microphones, you can make the best of them by using low capacitance cable - below 100pF/metre would be ideal, and by keeping cables as short as possible. But how short? Well, working from the table in last month's Interconnect, you can discover the maximum length if you know the cable's capacitance. Failing this, 3 metres is a good rule of thumb. If this proves too short - or just plain antisocial (!) - on occasions, the high impedance input on most DI (Direct Inject) boxes will provide a suitable, local impedance step-down. The subsequent cabling can then be run as long as you like, but you'll then need to find a low-Z input to feed into. But, as we'll see in a moment, the whole idea of using low-Z microphones is to transcend petty limitations on cable lengths, and sundry related nuisances.
As implied earlier, the output of most low-Z mics is significantly below that of high-Z types. In return for this inconvenience, we can use any practical cable length - up to a hundred feet even - without worrying about top-end losses. Back to the actual output: for any microphone, it's as long as a piece of string. It all depends on (a) how loud and (b), how close the sound source is.
Typical output levels are around 2 to 10mV for normal conversational speech, 100mV for loud vocals, and 500mV (½ volt) or more when our low-Z mic is positioned inside a bass drum. Returning to ordinary speech for the moment, noise generated by the mic amplifier (in the mixing console) needs to be at least 65dB below 2mV (-52dBu) if hiss isn't to intrude on the lowest level signal we're likely to be dealing with. This implies a noise level of around -117dBu, that's 65dB below 2mV (-52dBu).
Budget equipment can't as a rule offer low-noise performance of this order, that's why its circuitry is made to suit high impedance microphones; the extra 20 to 25dB of output makes the best of a poor S/N (Signal to Noise) ratio, so saving us from unnecessary hiss.
Returning to low impedance microphones, their nominal source impedance is typically 200 to 250 ohms, particularly when they come from European makers like AKG, Beyer, and Sennheiser, whereas in the US, lower source impedances lying between 50 and 150 ohms are more common (EV, Shure). Returning also to our loading ratio, the minimum load impedance for 200-250 ohm mics is 5 times greater, at 1000 to 1200 ohms. For this reason, you'll find that most UK and European mixing desks designed for low-Z mics feature a nominal 1k (or 1k2) input (or load) impedance as standard. As with high-Z mics, the exact impedance isn't too fussy, and the loading ratio of 10:1, say, achieved when the desk is mated with a US microphone with a 100 ohm source impedance is perfectly satisfactory. Indeed, loading problems will only arise if the desk's input impedance is much below 1000 ohms.
For certain US mics, particularly older models, power matching may come into the picture. This prehistoric concept is still embedded within the psychology of US audio engineering. To power match, we make the load impedance equal to the source impedance. This reduces the output voltage, and hence the S/N (Signal to Noise), ratio by 6dB (50%), which is all rather pointless, not to mention unnecessary - at least from the perspective of European mics. But unless you have specific information to the contrary, it's fair to assume that any US microphone will operate satisfactorily into the standard UK 1k load impedance. In other words, forget power matching! The only likely by-product would be a slight peaking-up in the treble response, particularly when long cables are in use. But this anomaly can be made to disappear by simply wiring a resistor around 620 ohms - 1k2 across the microphone's output terminals. The mic will then see the ideal load impedance, albeit with a 6dB drop in output.
Earlier, I implied that there was nothing intrinsically wrong with feeding a low-Z mic into a high-impedance input. Putting any mild HF response anomalies to one side, this is broadly true. It's just that (1) high-Z inputs tend to be designed for the higher levels commensurate with Hi-Z mics, and (2), the physics behind noise mechanisms are apt to make high-Z inputs noisier than their 1000 ohm counterparts.
Taking these facts into account, you may nevertheless find that a low-Z mic works satisfactorily into a high-Z input. Providing the signal level is high and/or a little extra hiss isn't too much of a problem. In a tight situation, this means you may be able to press a high-Z input into service for a low-Z mic, especially on loud instruments like bass drums and brass instruments.
This category includes electret models. The big difference with powered-microphones (such as capacitors) is that we're interfacing active electronics with the mixing desk input, rather than (passive) coils of wire. The internal circuitry will generally exhibit a very low source impedance - typically 50 ohms or less, but there may be limitations on the output-current capabilities of the circuitry, or failing this, there's generally a need for protection against short circuits and unacceptably low load impedances.
For these reasons, a resistor, typically 47 ohms to 100 ohms, will often lie in series with the output, and this will largely define the mic's source impedance, as viewed from the output terminals.
As a general guide then, capacitor mics will exhibit source impedances of 100 to 200 ohms, and will mate happily with a load impedance 5 times bigger - or anything greater. For example, Calrec 600 series mics have 100 ohm source impedances, and can therefore be plugged into US desks (with 500 or 600 ohm load impedances), but are also equally at home with the 1000 ohm impedance seen at the inputs of most UK consoles.
So from the impedance angle then, capacitor mics are very similar to low-Z moving-coil models, but there's one difference - output level is much higher, and often not far removed from the levels achieved by high-Z mics. For example, the output of Calrec 1000 series models is about 15dB higher than a typical low-Z moving-coil mic, whereas the 600 series models are around 20dB more sensitive, and thus on a par with high-Z models.
The extra output means that the S/N ratio needn't suffer. In a word, hissless! Indeed, in most instances the only problem - as such - will be one of the potential overloading, especially if the padding and gain input control facilities on the high-Z input are inadequate. But, assuming you can identify such occurrences, and take care of them with plug-in pads (attenuators), capacitor mics will, by and large, mate with any input, be it high or low.
Microphone matching is a confusing topic, principally because manufacturers of mics and consoles have failed to agree on definitive standards amongst themselves. Even in Europe, where the 200/1000 ohm mic/equipment standard holds good, not all low-Z microphones exhibit a 200 ohm impedance. But the situation with Japanese and American equipment is much worse - even the terminology is hopelessly confused! For example, one US company has used the term 'rated impedance' to indicate the microphone's source impedance. This is totally out - 'rated impedance' always refers to the input impedance of the equipment!
If you're using a UK or European mixing console, it's a fair bet that any low-Z microphone can be mated without difficulties. If in doubt, try a parallel resistor, if only to confirm that the response isn't significantly modified/improved. This evaluation will only normally apply to US mics.
On US consoles, a lower input impedance, often 500 or 600 ohms may be encountered. This will affect some European mics, but not others. Sadly, short of changing the desk's input transformers, or disposing of the mics that don't match properly, there's not a lot you can do to remedy the situation. Essentially, if your mic collection - like most people's - incorporates AKGs, Sennheisers, Beyers and the like, it's best to steer clear of US consoles unless you're certain that the input impedance is around 1Kohm - or is readily modified.
Sadly, it's not possible to be specific about the effects of wrong impedances. This is because cables, cable lengths - and the resulting reactances - not to mention transformers, all exert a major influence above 1kHz. The best remedy to this uncertainty is to make your mic cables no longer than they need be.
Shure's low-Z models have diverse source impedances - 38 and 150 ohms for the SM57/8, 85 ohms on the SM81, and 160 ohms on the SM59 model, for example. But Shure now recommend voltage matching for all their models, so 1000 ohm console inputs will suit in every instance.
Nearly all Beyer's microphones have a 200 ohm source impedance, and the minimum recommended load impedance is 500 ohms on some models, but 1000 ohms on most. With AKG, 500 ohms is the recommended minimum load impedance for most capacitor mics, but the source impedance of their dynamic models is mostly around 300 ohms, so a minimum of 1k input impedance is recommended with these.
For Sennheiser, the common dynamic models (eg. MD421, 441) are all at 200 ohms source impedance, so a standard 1k input impedance will suit nicely. Last, some of the older Electro-Voice models follow a similar pattern to Shure, with a range of source impedances (often adjustable), but as a general trend, the more recent models are all 150 ohms. This impedance will mate satisfactorily with most consoles.