The View Up Here

Random scribblings about kites, photography, machining, and anything else

Powering the EM172 Capsule

Posted by Tom Benedict on 16/01/2016

Chris Hass wrote a very nice article on building microphones around the EM172 capsules on her site, Wild Mountain Echoes. In it she mentions the issue of power. The datasheet for the EM172 specifies a supply voltage of 5-10v, but most handheld recorders supply something considerably lower than that. Chris and I compared notes, and her Sony PCM-M10 and both of my Tascams supply something closer to 2.3-2.7v. My question to her was how this affects performance, and what my options are for doing something about it.

Chris suggested bypassing the recorder’s own built-in power and using an external battery box to supply a higher voltage to the mic. She pointed me toward the boxes made by Church Audio. I followed her advice and bought a Bat 2B from them. It should be here in a couple of weeks.

Since my 70D has XLR inputs I decided to pursue another possibility as well. Most recorders can supply 24V or 48V phantom power on their XLR inputs. The only trick left is to drop that down to the 5-10V the microphones want. I ran across a thread on the Yahoo! micbuilder forum that referenced a circuit by David McGriffy called Simple P48 WM61 (referring to a simple circuit to power the Panasonic WM61 microphone from 48v phantom power). Richard Lee uploaded a document to the forum describing McGriffy’s circuit, along with modifications for using it with (you guessed it) the EM172 capsule. I still have a bunch of Mogami cable left over from building my earlier mics, so I ordered the remaining parts for McGriffy’s circuit from Mouser Electronics. he parts should be here in a couple of weeks as well.

In the meantime I figured it would be a good mental exercise to try to predict what each of these approaches would buy me in terms of performance. All of this ties back to a set of graphs on the micbuilder forum. It’s in Files/EM172/Primo EM172 Sens Noise vs RL VL.pdf. The graphs show the performance of the EM172 capsule as a function of supply voltage and input impedance. Using a battery box or McGriffy’s XLR circuit will let me change the supply voltage, but the input impedance is a function of the recorder. Here are some cases:


Tascam DR-05 and DR-70D 1/8″ Inputs:

Both the Tascams supply just under 3v for plug-in-power. The input impedance on the DR-05 is 25k ohms, and the DR-70D is 10k ohms. The graphs only go up to 10k ohms, so I’m using that number for both cases. Bumping the supply voltage from 3v to 9v should have the following effect:

Sensitivity: -38.6dB -> -36.7dB (smaller negative numbers are better)
Noise Floor: -112.7dB -> -116.1dB (bigger negative numbers are better)
S/N: 74.1 -> 79.4dB (bigger numbers are better)

In reality the DR-05 should get a bigger bump since its baseline performance will be lower than at 10k ohms, judging by the trend in the graphs. But the preamps on the DR-05 are noisier than those on the DR-70D, so I may not be able to hear the improvement.


Tascam DR-70D XLR Input:

The DR-70D’s XLR inputs have an input impedance of 2k ohms. Since I’m starting at 5V it should have the following performance:

Sensitivity: -38.3dB
Noise Floor: -116.8dB
S/N: 78.5dB

The noise floor is better than on the 1/8″ input, but the sensitivity won’t be quite as high. If I re-sized the resistor in the McGriffy circuit to provide something closer to 10v I’d get the following results:

Sensitivity: -37.7dB
Noise Floor: -116.8dB
S/N: 79.1dB

No change in the noise floor, but the sensitivity would improve by another 0.6dB. I’m not sure I can hear that, so it’s probably not worth dinking with.


Sony PCM-M10:

I also ran the numbers for Chris’s recorder. The Sony has an impedance of 3.9k ohms. Bumping from 3V to 9V should have the following effect:

Sensitivity: -37.8dB -> -37.2dB
Noise Floor: -115.1dB -> -116.5dB
S/N: 77.3dB -> 79.3dB

Almost 1.5dB improvement in noise floor, and 2dB overall improvement in signal to noise.


Sony PCM-D100:

The input impedance of the higher-end companion to the M10, the PCM-D100, is 22k ohms. It should see a similar performance bump to the Tascam DR-05, but since the preamps on the D100 are so much better than the DR-05, this will likely make for an audible improvement in the performance of the mic.


From the standpoint of mic performance, both approaches provide a clear gain. Whether my ear is sensitive enough to tell the difference remains to be seen (or heard!) From the standpoint of convenience, additional gear complexity, etc. each one has its pluses and minuses.

On the up side, the Church Audio battery box supplies 9V and will work with any recorder with a 1/8″ input, so I can use it on both of my recorders. Another up side for me, personally, is that so far I’ve built all my EM172 mics with 1/8″ plugs, so it requires the least re-work in order to test. On the down side it means I have to add a 9v battery, battery box, and cable to my setup. Velcro will go a long way toward making this a non-issue (mostly), but I wish this kind of thing could be designed in from the get-go. (Recorder manufacturers take heed! Being able to dial in a particular plug-in-power voltage would be nifty!)

The up side with the XLR approach is that from the standpoint of gear it amounts to changing the plug at the end of the cable. All of the circuitry fits inside the XLR plug. As an added bonus I’ll be able to plug EM172 mics into all four XLR inputs on my 70D, which is pretty darned cool. (The 70D only has one 1/8″ plug, which is tied to channels 1 and 2 only. Up until now I’ve only been able to do two channel recording on my four channel recorder.) The down side is that the 48v phantom supply on the 70D is a battery hog. So even if it works it means I’ll have to pack extra batteries or an external battery pack.

Good news is neither approach was all that expensive, and even with the Bat 2B or the external battery to compensate for the extra load from the 48v phantom power, neither adds too much bulk to my bag. For the moment I’m looking at it as having more options rather than having to choose between one approach or the other. In the extreme case it would give me the ability to plug two mics into my DR-05 with the Bat 2B, and another four into my DR-70D using XLR plugs. Six channels at once!

Now all I need is a subject that actually needs six channel audio. But that’s for another day.



4 Responses to “Powering the EM172 Capsule”

  1. […] « Powering the EM172 Capsule […]

  2. […] is the last in a four part series about powering the Primo EM172 microphone capsule. Part 1 outlined the problem of how to provide 5-10v to the capsule and predicted some results. Part 2 […]

  3. Tim said

    Hi Tom,

    I love these 4 blog posts. So, for those of us who are challenged by electronics but OK with math, would you mind walking thru how you calculated the sensitivity change when adapting the em 172 to ~10V with the xlr inputs of the dr70? Its a great teaching moment. And, secondly, the dr70 has 2kohm impedance on its inputs, like many xlr inputs of digital recorders. I assume that this is fine with the 2.5kohm impedence of the em 172 in that “it works” and that the change in calculated sensitivity arises from this parameter for the xlr inputs. Thanks for your help and can’t wait to hack some of these together later this year.


    • Tom Benedict said

      Unfortunately it’s a lot less mathematical than that. Most of it boils down to reading a set of graphs that’s posted on the Micbuilder forum:

      The file in question is Primo EM172 Sens Noise vs RL VL.pdf

      But re-reading it now I’m pretty sure I messed up. In my calculations I read “resistance” and thought it meant “input impedance”, but what the graph is actually talking about is the load resistance used in the circuit in the EM172 datasheet: In the example in the datasheet they’re using a 3.9k resistor with a 5V supply voltage.

      In the case of the 3.5mm input powered through plug-in-power, the load resistance will be dictated by the internals of the recorder. Tascam doesn’t have that listed anywhere on their site, so I can’t even begin to hazard a guess what the load resistance would be.

      In the case of the XLR inputs there’s a bit more info to work from. Henry on Audio Improv has a neat write-up on a number of mic circuits: In that article he cleared up a long-time confusion of mine: how the output stage of the Schoeps circuit actually works. In it he points out:

      “The output stage looks a bit weird, but that’s because only half of it is in the microphone. The rest of it is inside the mixer or mic preamp. The circuit is a PNP emitter follower which like the source follower has no voltage gain, but has considerable current gain. Because of the 100% local negative feedback, distortion and output impedance is very low. The emitter load resistors are the 6.8K resistors inside your 48V phantom power supply, normally part of your mic preamp. The 100K resistors furnish bias current to the PNPs.”

      Which should mean RL for most XLR inputs is 6.8k. Which, going by the graphs, should mean that a bias voltage between 5V-8V is about right for the EM-172 when plugged into an XLR input, and should give a sensitivity of around -26dB, noise around -106db to -107db, and an SNR of around 80dB.

      Dang. I really need to re-write that whole string of articles now. Thanks for bringing this up.

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s

%d bloggers like this: