Microphones + HiMD = confusion

[niftydog]

Recently, whilst considering a 2nd hand Rode NT4 I came here in search of some answers to questions regarding the microphone input on my HiMD (MZ-NH900). What I found was some useful information, some well-meaning misinformation and a whole lot of confusion! Here's a few things that I learned along the way that will hopefully help someone in future, and in the meantime it should clear up some confusion.

First cab off the rank, microphone power. MD units with microphone inputs supply some DC (1.5V or thereabouts) power at the socket to power small microphones. The only units I have tested that use this feature are the usual suspects from Sony. This DC voltage will not hinder the operation of "conventional" condenser microphones, however, nor will it suffice to power them properly either. The most ideal condenser microphones for use with MD units are self-powered by a 9V battery or similar.

Most microphones utilise balanced signals. Balanced signals are less susceptible to EMI and are therefore ideal for low-power signals such as microphones. The problem with balanced signals is, they need a balanced input. Portable MD devices do NOT have balanced inputs. Therefore you need to convert from balanced to unbalanced before inputting the signal to your MD device. There are two common methods;

1. make a cable that shorts one phase to ground and connects the other phase to the signal input.

2. make a balanced to unbalanced converter using a transformer or some electronics.

The problem with solution 1 is that you end up with a signal that is half the amplitude of the full balanced signal. The MD preamp has to be run with a lot of gain to get a decent signal, but more gain equals more noise. (Folks, this is THE problem with the factory Rode 5 pin XLR to minijack cable!) I found that it was far too much noise to be acceptable, so I had to go with solution 2.

Solution 2 preserves the full amplitude of the signal thus reducing the gain required, thus reducing the noise recorded! You can buy balanced to unbalanced converters of the passive (transformer) and active (electronics) variety. Passive converters are preffered in this case as you don't need a power source for them. You can make passive converters quite easily - if you can get a hold of some old condenser microphones. I salvaged two transformers from broken microphones and made my own stereo passive converter housed inside one of the salvaged microphone cases.

Regarding the Rode NT4:

I am awaiting information direct from Rode, but it is my understanding that the factory 5 pin XLR to minijack cable does NOT convert from balanced to unbalanced. For this reason, many people have been disappointed with the results.

The resistors in the minijack housing seem unlikely to be attenuators although I cannot confirm or deny that at this stage. They are more likely some kind of load resistors that are an attempt to provide good conditions for the microphone output stage. This microphone has a solid state output stage (not a transformer) and it's not ideal to just short one of the phases to ground. Thus, the unused phase might be tied to ground via some load resistance. As I said, awaiting information from Rode on this one.

If you are going to make your own version of this cable, don't waste your time mucking around with two cables into one connector! Use a standard two core mic cable and put the load resistors (or shorting links) in the XLR housing. The minijack only requires three connections, left, right and ground - save your sanity by only putting three wires inside it! Wiring the "factory" XLR splitter is somewhat more tricky, but if you can get a hold of quad core cable that makes it much easier.

There was a question about the "49 mV" spec in the manual. This is the minimum voltage for line inputs. Given even a high-output microphone would be lucky to generate even half of this voltage you shouldn't be trying to put microphones into line inputs. Always put microphones into the microphone input and run with as little gain as possible to get a decent signal.

Anyone else got some tips?

[dex Otaku]

Good stuff, niftydog.

Anyone else got some tips?

Not really. Or, well, maybe.

You can find transformers suitable for making simple balanced->unbalanced interfaces at many surplus electronics stores. In Canada there a line of stores called Princess Auto, many of which will occasionally have 1:1, 1:2 and 1:5 audio transformers in stock [though I will not vouch for their quality. In the world of analogue, you definitely get what you pay for.].

Balanced vs. unbalanced: If you have to run a balanced mic any kind of [cable] distance to an unbalanced recorder, *use the balanced cabling* for the long part of the run, and a balanced->unbalanced interface of some kind right beside your recorder. Balanced connections = less signal loss and more noise rejection.

That said, the type of cable included with the NT4 is an example of a single-cable, no-power-required solution to the problem of getting the signal from the mic to the recorder, assuming the recorder is very close to the mic. Any solution of better quality will require outboard equipment, and most will require their own power.

Thoughts on the Rode NT4: the only one I've had the chance to play with [a rental] came with one of Rode's 5-pin to 3.5mm TRS cables [as well as 5-pin to dual 3-pin XLR]. The signal output from this mic [using a fresh 9V battery] was so hot that it easily overloaded the mic preamp in my RH10 with any sound louder than someone shouting within a couple of meters. Jacked into the line-in and used as drum overheads, levels would reach the -12dBfs mark, but YMMV. The same mic also overloaded the input preamps on a M-Audio Delta 1010LT regardless of what their input gain was set for [eventual solution: don't use the preamps on the 1010LT at all, they appear to be useless with any mic more sensitive than an SM57].

Related stuff:

http://en.wikipedia.org/wiki/Balanced_audio

(PDF document) Mackie's "Balanced Lines, Phantom Powering, Grounding, and Other Arcane Mysteries" which is worth reading despite many people's opinion about Mackie.

Jensen Transformers Application Schematics.

In particular, for the real DIYers, (PDF document) JT-MB-C "Real" Mic Inputs & Phantom Power for Sony DAT which is exactly the interface needed for stereo [dual XLR] balanced mics -> stereo unbalanced [3.5mm TRS] mic audio. Would be nice if they also had a PCB design to download.

Side-note: whether the 5-pin to 3.5mm cable used with the NT4 is suitable for your use really depends on what you're recording. For loud sources, it should be just fine. For quiet sources [i.e. nature recordings], a real balanced interface is more likely to be necessary.

[ozpeter]

When making up XLR cables involving stereo 3.5 plugs, buy a premade cable such as 3.5 to 2 x phono, chop off the phonos, solder on the XLRs. Saves struggling with soldering those pesky little 3.5 plugs.

[Malcolm Stewart]

First cab off the rank, microphone power. MD units with microphone inputs supply some DC (1.5V or thereabouts) power at the socket to power small microphones. The only units I have tested that use this feature are the usual suspects from Sony. This DC voltage will not hinder the operation of "conventional" condenser microphones, however, nor will it suffice to power them properly either. The most ideal condenser microphones for use with MD units are self-powered by a 9V battery or similar.

Anyone else got some tips?

Thanks for an interesting posting. My experience with XLR terminated mikes is zero, and I use small electret types, Plug-in Power, most of the time. For camcorder video I've tried low cost directional mikes but they're not good enough for audio recording.

I've measured the "Plug-in Power" voltage for my two Sony Hi-MD units, it's 2.38V, and also my older Panasonic MD recorders, where it's slightly lower at 2V. I normally use Yoga tie-pin microphones (Dummy Head Stereo) and they seem to work quite well. I've yet to measure their noise level. Still wondering how to do, it as I'll obviously need a good level of acoustic insulation for the test to be meaningful.

The article http://www.wildlife-sound.org/equipment/hi...evels_meas.html indicates that feeding directly into the Mic-in socket incurs a dynamic range limitation which apparently doesn't occur when using the Line-in socket. I've tried it using a 9V battery box but have yet to find reliable and repeatable loud sounds which would demonstrate any differences.

[niftydog]

Thoughts on the Rode NT4... The signal output from this mic [using a fresh 9V battery] was so hot that it easily overloaded the mic preamp in my RH10 with any sound louder than someone shouting within a couple of meters.

This brings up another tip:

Condenser mics are notoriously sensitive. I wouldn't place the Rode NT4 or similar mic within a meter of an acoustic guitar let alone a drum or a good vocalist. Distortion can be generated within the mic itself. So, if you get distorted results, don't automatically blame too much gain, it could be that the sound source is too close to the mic(s).

Naturally, it's all relative to your equipment (YMMV!) My MZ-NH900 is set to low sensitivity and runs quite low mic gain (around 7-9) with good results.

[dex Otaku]

This brings up another tip:

Condenser mics are notoriously sensitive. I wouldn't place the Rode NT4 or similar mic within a meter of an acoustic guitar let alone a drum or a good vocalist. Distortion can be generated within the mic itself. So, if you get distorted results, don't automatically blame too much gain, it could be that the sound source is too close to the mic(s).

I've been close-mic'ing [25-50cm] instruments and even amp cabinets for about 15 years with various condensors without distortion. Most stage and/or studio mics have pretty high max SPL ratings [-WAYYYYYY- louder than almost any acoustic instrument is capable of being played], especially if properly biased [i.e. 48V phantom] and jacked into even a half-decent preamp with lots of headroom. Even mediochre portable studio or FOH mixing gear have worked fine in literally every situation I've ever encountered. I didn't start getting preamp or mic distortion until I started using MD, but then - it runs off one battery, you have to expect there to be some sacrifices.

[62v8]

Gents,

Well I took my Rode NT4 out for it's 1st full blown rock show on Sunday (New Christs).

Venue has a very solid small /pubclub PA (stacks both side of stage) that probably sat at around 120dB all night and lets not forget the amps onstage. Stood about 6m from FOS.

I'm "happy as Larry" with the results.

Took the earlier advice and got the side cutters out and redid the 5 pin connector to remove the pad. Ran off the mic internal 9V and straight into line in.

I'm lazy, it was dark and I was drinking from 5pm so by show time all I was bothered with was "just press record" and let the machine do it's thing.

In the quiet parts there is no hint of mic noise and the performance of the mic itself was astounding. Best drunken show I've recorded.

I did also run a set of MM cardioids on recorder 2 (another HiMD). They sound absolutely gutless in comparison but that isn't really fair ... a sub $100 set up vs a $500 mic.

Definitely a revelation to my recording arsenal.

Regards

GT

[niftydog]

I didn't start getting preamp or mic distortion until I started using MD, but then - it runs off one battery, you have to expect there to be some sacrifices.

That's something to add to the tips, remember that a mic running from batterys is never going to have as much headroom as one running off full 48V phantom power.

I guess I overload mics more because I'm a bass player and lots of moving air really upsets condensers! That is to say it doesn't have to be loud to distort. For instance, take a condenser and gently blow on it to see the effect.

[AllanH]

Sorry to be contentious about this, but as the 'discoverer' of them, I've found

removing the pad resistors in the NT4 XLR to 3.5SJ cable does not result in more noise. It gives a substantial improvement in signal to noise ratio (more signal, same noise).

The Rode cable converts to unbalanced as you suggest in your method 1. The pad resistors substantially reduce the signal at 3.5SJ plug in comparison with the input from the mic (about 12dB loss of signal). The 4 pad resistors are surface mount types on a tiny pcb in the XLR5 plug.

Suggest you have a look at:

http://www.uwm.edu/~type/audio-reports/Rod...leMod-index.htm

It's been discussed with Rode, as the above page notes, and is now a recognised 'mod' for the NT4 in field recording circles.

I have certainly not been disappointed by unbalancing the signal in this way - it has always been an improvement. But I haven't used long unbalanced cables. For long cable recording I use balanced cables from a pair of NT1As to an ARTII twin phantom supply, with a short unbalanced cable from the phantom box to the recorder (as Dex suggests). This works very well - should also work with NT4. Its worth noting that there was some discussion about choice of phantom supply last year on these pages, with some types conflicting with the PIP supply from the recorder.

I've not suffered from excessive signal overloading the mic pre-amp - but then I've always been making field recordings, not bands.

There's lots more on microphone noise, pre-amp noise, outboard phantom supplies etc on the uwm.edu pages mentioned above.

Hope this helps.

[niftydog]

Sorry to be contentious about this, but as the 'discoverer' of them, I've found

removing the pad resistors in the NT4 XLR to 3.5SJ cable does not result in more noise.

That's not what I was saying - the noise was due to the extra gain required because of the unbalancing technique used. In fact, I had the same trouble even without the pad resistors.

I can't seem to get a response out of Rode on this. I asked them directly to provide me with a schematic so I could see for myself what the resistors do but they have not come to the party. (I do not own the factory supplied cable so I can't inspect the PCB.) I must say I'm still surprised that they are indeed attenuators, but you can't argue with the facts! Obviously the bipolar output stage is happy to have one phase shorted to ground... seems odd to me though that Rode would subject their design to such treatment, but without seeing a schematic of the mic I can't understand their reasoning on that either.

Thanks for the info, very interesting stuff.

Still experimenting with my new toy though, getting some odd results too, I will update the thread if I find some answers.

[AllanH]

I guess I must be misunderstanding you - I find I need substantially less gain to achieve a given recording level with the pads removed. In comparison with them present and this also has the effect of pushing the noise floor down. Maybe I'd need even less gain if I used a fully balanced signal chain with no pads.

With regards to unbalancing the signal - Zafiro give the following adsvice for using unbalanced microphones with the AD-20 Inbox ADC:

"in order to connect unbalanced mics to the Inbox you should use a cable that will properly balance the signal - usually this involves a small resistor/pad between ground and the 2nd leg of the balanced input that isn't being used."

Zafiro advised me that the value wasn't critical "1kohm to 10kohm should work."

I've made up an XLR5 to 3.5SJ lead with 10kohm resistors as described above and used it with the Rode NT4 and NT1As. I've noticed no difference in the quality of the recording made or in the output level of the mics.

On a different topic - NT4 capsules can easily be contaminated and go noisy. One of mine did so and Rode were very good in providing a replacement (they simply screw on/off). I believe the same capsules are used for NT4, NT5 and NT6.

Regards etc Allan H

[niftydog]

I guess I must be misunderstanding you - I find I need substantially less gain to achieve a given recording level with the pads removed. In comparison with them present and this also has the effect of pushing the noise floor down. Maybe I'd need even less gain if I used a fully balanced signal chain with no pads.

I totally agree with everything you're saying, I just got the impression you thought that I had said that removing the pad resistors somehow adds noise.

If you do go down the path that I followed (making a stereo transformer unbalancing box) you'll find that you'll probably have TOO MUCH level! This is the problem I'm encountering. Using my MZ-NH900 set to a high mic sensitivity to record my bands rehearsals, even a record level of 1/30 results in distortion. It also results in a very low level signal leading me to the conclusion that it's actually the microphone preamp stage that is being over-driven - prior to the record level control stage! So, it seems that the record level adjustment doesn't set the microphone preamp gain at all, but in fact it sets the level being fed to the analogue to digital converter.

Conveniently my band is on a short break and I haven't had a chance to try the low mic sensitivity setting as yet!

"in order to connect unbalanced mics to the Inbox you should use a cable that will properly balance the signal - usually this involves a small resistor/pad between ground and the 2nd leg of the balanced input that isn't being used."

Well, that's not balancing the signal, but it is a better way than just shorting the unused leg directly to ground. This is what I initially thought the Rode cables mystery resistors did, but as you showed me they are in fact attenuators. Baffling!

BTW, still no word from Rode which is pretty disappointing.

[dex Otaku]

Using my MZ-NH900 set to a high mic sensitivity to record my bands rehearsals, even a record level of 1/30 results in distortion. It also results in a very low level signal leading me to the conclusion that it's actually the microphone preamp stage that is being over-driven - prior to the record level control stage! So, it seems that the record level adjustment doesn't set the microphone preamp gain at all, but in fact it sets the level being fed to the analogue to digital converter.

#1 - Why on earth are you using high sens mode to record anything louder than, say, a lecturer all the way across a hall, or birdsong?

#2 - The way the gain on the mic input works appears to be like this:

input -> preamp w/gain setting [low or high, ~15 and 35dB I believe] -> manual level control [this is NOT variable gain for the preamp, it's a level control AFTER the gain is applied, hence the preamp clipping regardless of what levels are set at - though in your case, turn off high sens mode, that should at least help a bit]

[AllanH]

Getting a bit off topic, but you might like to see these pages for info on mic pre-amps nad mic sensitivity/gain in the HiMD recorders:

http://www.wildlife-sound.org/equipment/hi...evels_meas.html

http://www.wildlife-sound.org/equipment/hi...myths_meas.html

[niftydog]

Why on earth are you using high sens mode to record anything louder than, say, a lecturer all the way across a hall, or birdsong?

Hey dex, While you probably didn't intend for that to come across kind of aggressive, it did.

If you must know it was an error on my part. Normally I would use the low setting but as I discovered to my horror it was not set properly first time around. As I said I will be trying the low setting at the next possible opportunity. The important part about what I was saying is that my stupid mistake has lead me to have a better understanding of how my HiMD works and I wanted to share that with the forum.

The way the gain on the mic input works appears to be like this:

input -> preamp w/gain setting [low or high, ~15 and 35dB I believe] -> manual level control

Um, thanks for that, that is the same conclusion I had already come to - but I guess it's reassuring that others think the same thing as me.

AllanH, more great stuff, thanks again!