I build and measure loudspeakers at home as a hobby. But I have never had a decent mic preamp for my unmodified Panasonic ECM capsules. I use a USB sound card to accept the mic audio and analyze it with ARTA software.

After thinking about it for awhile I decided to go to the mountain and use the Linkwitz circuit for my measurement mic preamp, but with a couple changes. See here at Microphone http://www.linkwitzlab.com/sys_test.htm

The Linkwitz circuit is designed for capsules that are cut. That means the capsules are referenced from ground to V-. So I thought why not design the circuit so that it can handle uncut as well as cut capsules. My capsules are all uncut but I may try out cut ones in the future.

So, I figured a switch could be used to switch the capsule reference from "ground to V-" to "ground to V+". That change is on the schema below. Thoughts?

I believe my Panasonic capsules put out about 6 ma at 90 dB SPL which is where I do most of my testing. The Linkwitz circuit has a gain of about 10 dB. But the cut capsules are more efficient than uncut. I am uncertain what gain I need to properly feed my sound card, so I figured why not make the circuit variable gain to provide for the uncut capsules. What I figured was to use a rotary switch to change the resistance of R3 in the schema below. A four position switch with the resistances shown should be able to provide a gain variation of 10, 20, 30 and 40 dB. But at the higher gains the effective R3 starts to get pretty small. Am I in trouble there? Would it be smarter to only have two levels of gain, like 10 dB and 20 dB? How High should I go to get a proper level to my sound card? Would it be smarter to vary R4? Suggestions?

The output of the circuit will go to line-in of a sound card. Cables will be very short, a few inches. So do I even need R5 and C2?

As for the pot, it may not even be necessary given the switched gain choices in the circuit. Thoughts?

I can use any value for the Pot if it makes sense to have it at all. Suggestions?

The resistances in the circuit were chosen by Linkwitz to keep noise down. Is there any good reason to change them? Especially considering the uncut capsule interface with the circuit? I understand the uncut capsules like an input impedance of about 20k or higher. Thoughts?

The LM833N OPA is a dual chip. Is there anything smart that can be done with the second amp?

And last, how long a cable can an uncut capsule push signal down before the high frequencies begin to be cut? I have seen home theatre mics with cable lengths of 20 feet or more. How far can I go with mic cable? Linkwitz suggests less than 30 ft with a cut capsule. What about a stock capsule?

Here is my schematic:


  • \$\begingroup\$ @Andy Yes I did mean 6mV. No tone controls here. Its a measurement system. I am concerned with insufficient signal to line in or overloading line in rather than clipping the OPA. Hence my concern for the gain range I should shoot for. Cables are "low Z mic cable" no specs. MIC capsules spec'ed "less than 2.2K impedance". I should be safe with 20 ft mic cables. Perhaps a tiny MIC buffer that fits inside an XLR connector? Can you point me to a circuit that will fit in an XLR connector? And ya the 833 is an oldie but a goodie.How about the rail switching switch? Any problems there? Thanks \$\endgroup\$
    – oneoldude
    Commented Jul 16, 2013 at 9:16

2 Answers 2


Using a pot as a voltage divider as you do is not a good volume/gain control. The problem is that it makes your output impedance dependent on the gain. This has the effect of making your frequency response vary with gain as that output impedance interacts with the capacitance/impedance of your cable and whatever you are feeding this output into. You want the output impedance to be as low as possible. Do this by deleting R5 and R6.

To do the gain control without impacting the output impedance use a pot to replace either R3 or R4 (or both).

The nice thing about this type of gain control is that it actually controls the gain. With your configuration if the opamp was clipping and you turned the gain down you would still be clipping the opamp. But with the modified gain control, if you turn down the gain the opamp will stop clipping.

Do not remove C2, since it blocks the DC signals to your sound card. You might even consider increasing the size of C2 to as high as 68 uF. The larger C2 is, the better your low frequency response will be. 10 uF is the lowest I would go, but 22, 33, 47, or 68 uF are better choices.


The rotary switch turning R3 down to 100 ohms isn't going to be a problem.

I'd make the variable pot element a 10K in series with a 1k and replace R4 with it for the reasons outlined by David Kessner i.e. putting it in the output affects the output impedance of the circuit.

The LM833 is a better choice than the OPA2134 in my opinion and leave C2 in circuit.

You say 6ma output for your capsule - do you mean 6mV?

2nd amp - why not add a tone control or maybe a LED limit circuit that tells you are close to clipping?

The cable issue needs your input as to the type of cable and the output impedance of the microphone.


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.