No announcement yet.

Condenser Mic preamp for PC

  • Filter
  • Time
  • Show
Clear All
new posts

  • Condenser Mic preamp for PC

    I was going to try posting this in the Amplog as kind of a build project but have been unable to create a blog. I already had to register separately from the main forums, I hope that was intentional, and I can now see the other blogs. (correction: I could only see them occasionally, probably while that session was timing out, now it asks me to log in again and does not accept my password any more). But trying to create a blog I cannot get past the step where it asks you to designate a *type of blog* since there is nothing in that menu to choose from, or something else is wrong.

    Maybe this could be moved to a blog later, isn't that supposed to be an easier way to handle pictures & schematics?
    Regardless, other builders and comments would be helpful so here goes, all I have yet is links so this first post is kind of a homework session for anyone interested who is not already aware of these initial design considerations:

    Well, how about a preamp which is application-specific for interfacing a decent condenser microphone directly to the soundcard input of a PC, especially a laptop for portable recording?

    The objective is to capture CD-quality tracks using as few analog components as possible between the mic and the analog-to-digital conversion process, where the PC's stereo Line Inputs are employed since they are usually direct connections to the Analog-to-Digital conversion chip.

    I guess that means two preamps would be good since using the Line Inputs we have two channels. We'll build only one circuit to begin with for testing, leaving room in the enclosure for the second channel later once its finalized.

    Then as long as the noise floor of the combined analog chain (mic plus cable plus preamp, that's all there is with this setup) is below the level which can be reproduced by a traditional audio CD (44.1KHz 16-bit), you have no audio compromises when recording a track at those sampling conditions then burning a CD from that track compared to a professional digital studio.
    Well, provided the sound itself you are recording is good enough without any further processing, and the preamp itself actually sounds good, or transparent, or just doesn't sound bad in any significant way, it must pass the audition after all.

    You also may possibly have an unfair audio advantage even compared to studios where tracks are sampled at higher rates than 44.1, but heavily remixed, and processed by numerous series and parallel digital processing modules and the trivial analog components involved (which depending on our preamp design will outnumber our analog components), then resampled down to 44.1 for producing the final CD. Especially if originally sampled at rates not evenly divisible by 44.1.

    Now I've been doing recording like this since 2003 using a low-noise stereo opamp preamp I had previouly built for recording vinyl LP's direct to the PC employing RIAA compensation. That's when I got the laptop, and that preamp was already built & proven, and there was enough gain even for weak dynamic mics when the RIAA was bypassed. Plus it ran good as a portable on a pair of 9v batteries for the +/- opamp supplies. But the whole thing is unbalanced and was not originally designed for the purpose of *taping* or recording raw live mic tracks for burning or digital remixing.

    This new preamp is going to be application-specific for live taping.

    The main functions are simple, impedance compatibilty between the condenser mic & PC Line Input, the right amount of gain to allow complete control from the PC (since the preamp will have no knobs), and supplying the phantom power to the condenser mic.

    Here's the background material (in order):
    (I suggest creating a folder for this project and save each web page to your hard drive)

    from this following ESP project we will be concentrating mostly on fig. 2:

    briefly consider a nice 48V power supply which can produce much greater milliamperage current available than needed for my small preamp project with only two microphones:

    At this point I would like to give extreme thanks to Rod Elliot for his expertise and posted documentation. This project is largely inspired by his material and is being simply adapted for my application. If a builder does not yet have the experience & confidence to follow my efforts for this preamp once its developed, I can see no better way than to gain skills by assembling some of the interesting projects designed by Rod at ESP using the prefabricated blank circuit boards he has available. check out lots more he has to offer at

    However, we'll be skipping the power supply for now, powering the first single-microphone unit by completely hum-free batteries instead as a reference instrument and for portability also, and concentrating on the final assembly having low current drain for long battery life.

    I'll be posting a concept drawing shortly.


  • #2
    You might like this project if you haven't seen it before:



    • #3
      I think that PAIA MS unit would be worth experimenting with, although I have had good luck with simple direct stereo for live concert taping. I get the idea that MS technique was developed for orchestral recording. Thanks for the link, that is the kind of input I want, in addition to messages from any other builders that may be interested as I post more material, like the final schematic and pictures.

      Well, I hope nobody thinks my preamp is much of a grandiose project, it is supposed to be simpler than the ESP and PAIA assemblies.
      Plus it runs on batteries now so I'm not even addressing a phantom power supply at this time.

      I am starting with the cap-coupled balanced mic input rather than transformer-coupled, not much differently than Rod Elliot, since a very low cost preamp can be assembled this way, and later a small audio transformer specifically for this purpose can be used to replace the caps if desired. I want to be able to build at least one preamp from scrap, surplus, or parts I already have which are not costly, and this original design is expected to be useful for Ampagers to record their own amps using a relatively excellent phantom-power condenser mic (which can be borrowed from other musicians if you can't afford one yourself), and using a homemade preamp which should be within every hobbyist's reach, while getting CD quality or close to it. To some extent it is a recycling project, it has been designed to work on stacked 9V batteries but they do not need to be brand new, depending on how completely your friends wear out their batteries before replacing them, you can likely use their discarded batteries if they have at least 7V remaining, so under ideal conditions battery replacement is at no cost and will occur very seldom. I hope. We'll see how low the current draw turns out to be.

      There are going to be two channels in one battery-powered box, each having their own separate coupling caps or input transformers, depending on which you solder in. So you could make one channel cap and the other xfmr. It wouldn't be too difficult this way to A/B the exact same preamp design when the only difference was caps vs. xfmr and there are very few other parts anyway. Although you can get a limited amount of virtually noiseless gain from an input transformer if you use the step-up type, I'm just not using them at this time.
      Take a look at the Jensen website for some good xfmr ideas for simple balanced preamp input, here's a basic example:

      seems like some of their transformers will sometimes require load resistors and/or caps for the primary & secondary, to give specified performance, so there may be a few extra complementary components in addition to the mic input xfmr when you go to replace capacitor inputs with xfmr.

      Now here's the transformer way they have to do the MS micing:

      wiring these type transformers into the box in this way would make it a dedicated MS preamp, so you would need a cardiod and a bidirectional mic set up appropriately, and the two channels of preamp output would then be Sum & Diff rather than Left & Right or 1 & 2. You would however require 3 channels of your software mixer to mix these back to stereo afterward, since you will need a duplicate of the *Side* channel copied to an empty track at 180degrees out-of-phase with the original so you can have one copy of it panned to the left and the other copy panned to the right while mixing with the *Mid* channel. But it could be pretty good, especially for monophonic-compatibility.
      Check this out if you didn't read it already:

      So as expected a single laptop and two mics can capture live stereo when needed, OTOH you can also record one or two separate (but syncronized) studio tracks at a time, of individual instruments for regular studio type remixing with previously captured rhythm or live tracks for instance. None of that USB latency at least. Even with just two channels of live input at a time, the number of total tracks you can accumulate before final mixdown should only be limited by your software and PC hardware. Also, with a desktop PC you could quite easily use a professional soundcard offering much faster sampling rates than the 44.1KHz CD standard, and deeper bit depth as well, and you might like this preamp under those conditions just fine but it is not expected to have the abysmal noise floor which would be preferred when you require the raw tracks to be much better than CD quality.

      In live taping, with two laptops each having a 2-channel battery preamp and MS mic setup, you should be able to mix down to surround sound 5.1, this could give you some ideas:

      Now here's Elliot's original drawing:

      and here's how I modify this input wiring before going to the preamp:

      I have removed R3 and the power switch from where they were originally, since I will be switching the power ahead of where it said +48. Also, C1 is shown as a local filter for the phantom power coming in, but for the prototype the batteries will be ripple-free enough by themselves, and a 220uF will be used as a filter for the preamp power rather than the phantom power. The prototype preamp is application-specific for an AKG Perception 200 Mic interfacing to a Dell Inspiron 8200 laptop. Probably end up working for just about any PC, and many other makes & models of mic. From what I can tell by reading the mic specifications from AKG, the MIC itself needs 48v +/- 5VDC (or +/- 4VDC depending on which spec sheet you look at). There are other AKG condenser mics that are specified to require between 9 and 48V, but this is not one of them. In operation at 48V, I measure the mic consumption at 3milliamps, the matched 6.8K phantom resistors (which drop the voltage from the phantom power source) operate in parallel effectively as a single 3.4K, and drop about 10volts across them. Therefore to provide the mic itself with an actual 48V, I will need about 58V coming in at the B+ point, whether from batteries or an AC supply. Removed the 20db pad since the mics have a 10db pad switch on the shell. Changed the protective zener diodes so that they protect primarily against signal-input or cap-differential spikes only as shown, rather than protecting each input from 48V phantom power spikes. (This is because I will be ultimately powering the preamp from the same batteries as the phantom power, and will have the preamp be unbalanced this time, so the -Input of the preamp will be connected to B- in the end. Therefore any spikes will still be sent to *ground* in my assembly. This doesn't have to be the case, we'll take a little theoretical look at the possibilities for a balanced preamp and one powered by separate batteries than the mic (which are separate from the PC batteries too), so at this point I think its best to leave the concept drawings floating. Keep in mind that B-, Vzero, Ground, Earth, and Chassis do not always need to be in contact with each other, and with full battery power the possible variations are numerous.) I'll be using different value zeners too.

      Realistically, on a complete battery-powered system (mic+preamp+laptop) without an Earth reference at all, the simple *unbalanced* stereo Line Inputs on the PC behave no differently than balanced inputs, so it would probably be better in that case to have a twisted pair for the line-level signal leading from the preamp to the PC as well, so that these two leads can reject some commom mode themselves, otherwise the single-inner-conductor shielded wire which is normally used (actually one inner conductor for each channel if not two parallel single shielded lines where they connect to the 1/8 stereo plug) will have the Signal- line exposed to interference since it is the cable shield, while the Signal+ is protected by the shield. If you had twisted-pair-plus-shield instead, you could connect the shield only at the preamp end (and maybe a suitable Earth at the same time), leaving the PC completely floating.
      The laptop I'm working with right now does have a 3-prong plug on its AC adapter, and upon checking, the regular shield of the 1/8 Line Input jack is Earthed when the AC adapter is connected. So if balanced wiring was employed to connect the preamp to the PC, it would default to regular unbalanced when you plugged the laptop in to wall power, provided you wired it correctly and carefully. Then the shield on the regular PC line input cord will be effective at keeping out noise since it is Earthed. You would need to be even more vigilant if you were powering the mic, the preamp, and the PC all using rectified AC power having a common ground, since you must consider how having a Grounded Signal- output on some balanced preamps is allowed while on other preamp designs it could potentially throw sparks or release smoke.
      So where it says *Cable Shield* in the above drawing is indicated being connected to B-, it means only the cable to the mic in the concept drawings, not necessarily the cable from the preamp to the PC.

      So here's the next concept drawing:

      Here all I have done is given the basic outline, but it can be seen how the amplification module is still balanced and floating at this point. It could very well be an opamp preamp powered by its own dedicated pair of 9V batteries, +9 & -9 for V+ & V-, with a floating Vo connected where the batteries join in the middle, often encountered. In that case Vo would be in contact with the shield of the PC line input if it was plugged into the PC directly, and the pair of 9V batteries actually connected at this point to the PC. However, if this was unwanted a pair of coupling caps or a transformer could be inserted between the preamp and the PC to make a fully floating system, and the presently-shown coupling cap from the preamp module to the PC would be unnecessary with many preamp types.

      Well, that's enough of a message for today, but it is already no longer just an idea, the prototype seems to be pretty good so far, here's what it looks like:

      and here's Take 1 from the young musician who came to pick up his Strat that day after I had intonated it:

      He has a Duncan Vintage Rail PU at neck position, Hot Rail mid, and what looks like a stock PU bridge, slightly rewired. Amp is my '60's EL84 SE Class A Gibson Skylark, equivalent to the Kalamazoo Model 1, but highly modified. Kendrick Blackframe speaker, 10inch, open back. As you all know, this amp only puts out 3watts clean, it was not cranked, neither was the guitar.
      Amp was sitting on the floor, 6feet from the mic, with mic 30inches above the floor.
      He was standing a foot or two in front of the amp, facing the mic. The vocals came out pretty good, even though there was not a separate vocal mic, its just a raw unprocessed snapshot.
      There is a lot of background noise in the lab, but I think the tone is good.

      Then I tried the classical guitar about 6inches from the mic, a little boomy, but its the exact same software settings as the above track, there are no level controls on the preamp:

      More pictures later as I document a bit of the assembly process, its in a 3x4x5inch aluminum BUD box or equivalent.