Dynamic Microphone LRC Filter - Anyone Familiar With This?

[rodabod]

Whilst dismantling a well known and well respected dynamic microphone, I came across something I had never seen before.

It seems to be a series LRC filter wired in parallel with the capsule.

Here is a schematic (drawn in PaintCAD.... err..ok... Paint):

So, does anyone know what exactly this is for? I'm guessing it is maybe there to improve the response in some way?

I calculated the low impedance resonance frequency of the RLC section as 1/Rt.(L*C), which gives Wo = 26860 which is 4275Hz.

But having said that, I still don't really know what is going on. Any ideas?

 

[clintrubber]

All I could think of is that's there to dampen a certain peak in the response. The environment may alter it, but as it is it's obviously creating a dip on that frequency. Q seems around 3.

(Hmm, I used f = 1/(2*pi*sqrt(L*C)) and you a different one but they both give the same freq. I'm too lazy now to 'rewrite' both formulas but given fixed values of L & C the freq is now determined (ignoring the environment, just like certain countries...)) - but R may change to adjust Q.
So one wouldn't expect it in the freq-determining formula like in the one you used, but I think I remember having seen before that both formulas are indeed the same - time for some pencil & paer, but not now.

But back to your question, I haven't seen such EQ (electronic peak/dip compensation) in mics before. Curious which model btw...

Bye,

Peter

 

[rodabod]

[quote author="clintrubber"]All I could think of is that's there to dampen a certain peak in the response.[/quote]

That is the first thing I thought when I found it. Then I thought it might be there to extend the high end, but I was not sure at first.

(Hmm, I used f = 1/(2*pi*sqrt(L*C)) and you a different one but they both give the same freq.

Yeah, I calculated Wo (radians/sec) and you calculated fo (Hz). W=2*Pi*f

It's all the same.

...but R may change to adjust Q.
So one wouldn't expect it in the freq-determining formula like in the one you used, but I think I remember having seen before that both formulas are indeed the same - time for some pencil & paer, but not now.

Well, I think Q=Wo*L/R for a series LRC, but I'm not sure if the value of R would be 560 Ohms or 760.

(ignoring the environment, just like certain countries...)

Oh dear... The jokes get worse and worse. Stop polluting the board!

 

[Samuel Groner]

Dynamic mics are basically resonances stringed together--unlike omni condenser mics which are inherently flat up to a certain frequency, a dynamic mic is never truely flat.

There are many ways to introduce poles and zeros in a response--capsule resonance, transformer response etc. This LCR-filter is just another possibility.

Samuel

 

[TedF]

That rings a bell :?
But are'nt dynamic mics just the same as loudspeakers the other way around? And where's the 'string of resonances' hung together there?
:?
But I know what Samuel means really....

This little network looks to me like a little bit of frequency shaping to damp the main diaphragm resonance; it's quite a low 'Q' (the resistor controls that). AKG D202?

 

[rodabod]

[quote author="TedF"]That rings a bell [/quote]

Ted, I hope that wasn't intended as a filter resonance joke...

But are'nt dynamic mics just the same as loudspeakers the other way around? And where's the 'string of resonances' hung together there?

They are the same, but the loudspeaker doesn't incorporate venting and helmholtz resonators which some capsules use to improve response. I think I'd maybe compare a sophisticated moving-coil capsule to a loudspeaker mounted in a cabinet - eg. using bass vents to extend low end by adding a lower resonance, etc.

This little network looks to me like a little bit of frequency shaping to damp the main diaphragm resonance; it's quite a low 'Q' (the resistor controls that).

Right, so do you think it is simply damping a main resonance? I wasn't quite sure what effect the LCR would have in series.

Also, does anyone know if I were to calculate Q-factor and fo correctly, would I need to incorporate the values of the resistance and inductance of the capsule too?

AKG D202?

Good guess! I see where you are coming from. I'm working on a couple of D202s at the moment.

I thought I maybe shouldn't say what mic it was, but ok, it is a B*eyer M*201.

 

[clintrubber]

[quote author="TedF"]That rings a bell :?
But are'nt dynamic mics just the same as loudspeakers the other way around? And where's the 'string of resonances' hung together there?
:?
But I know what Samuel means really.... [/quote]
You're obviously not in need of it, but I thought this might be a nice opportunity to mention 'The 'BIG' Mic Thread!':
http://www.freewebs.com/tekker/harveygerstmicthreads.htm

Can't remember having it seen here yet.
There's some on resonances - and a whole lot more.

Regards,

Peter

 

[pstamler]

[quote author="TedF"]That rings a bell :?
But are'nt dynamic mics just the same as loudspeakers the other way around? And where's the 'string of resonances' hung together there?
:? [/quote]

Well, you got your woofer fundamental resonance, your box resonance, your cone breakup, your dust cap resonance, your frame resonance, your tweeter fundamental resonance, your dome breakup...and that's only a two-way system.

Peace,
Paul

 

[PRR]

"Q" in low-Q circuits is meaningless until you define Q. Considered just electrically, there is 560+200= 760 ohms in the tank. But what we really want is the response, which is 560/(560+200)=0.73 over most of the band, tending to unity at the resonant frequency. If we define band-bump Q as the range between -3dB points, we may be frustrated because the response hardly bumps 3dB.

The real way to look at it: they had a flat mike. They wanted a Vocal Mike. Shure SM58 out-sells all others 100:1. Rockers need the Shure Peak. This RLC loads the capsule except at ~4K, giving an upper midrange peak. {EDIT: Wrong!} The amount of peaking is surely set by ear, not by calculation. Cut and try.

 

[clintrubber]

[quote author="PRR"]"Q" in low-Q circuits is meaningless until you define Q. Considered just electrically, there is 560+200= 760 ohms in the tank. But what we really want is the response, which is 560/(560+200)=0.73 over most of the band, tending to unity at the resonant frequency. If we define band-bump Q as the range between -3dB points, we may be frustrated because the response hardly bumps 3dB.[/quote]

I see what you mean, the Q=3 above is just meaningless, it's ignoring the surrounding stuff.

Rockers need the Shure Peak. This RLC loads the capsule except at ~4K, giving an upper midrange peak. The amount of peaking is surely set by ear, not by calculation. Cut and try.

I should be sleeping i.s.o. typing, but isn't the RLC-section loading the capsule the most at that ~4kHz ?

Regards,

Peter

 

[dayvel]

This is a lot like the inline bass drum mic EQ I made a while back (works great, thank you) so I was able to plug the values into my SPICE sims for that. Making the outrageous assumption of a purely resistive mic impedance going into a 4K4 load, this circuit gives a 2dB dip at about 4300kHz.

 

[rodabod]

[quote author="clintrubber"]

Rockers need the Shure Peak. This RLC loads the capsule except at ~4K, giving an upper midrange peak. The amount of peaking is surely set by ear, not by calculation. Cut and try.

I should be sleeping i.s.o. typing, but isn't the RLC-section loading the capsule the most at that ~4kHz ?

Regards,

Peter[/quote]

I agree. Resonance occurs when the impedance is minimised in the LCR network.

So, at resonance, if the LRC is at minimum impedance then is it not loading down the capsule?

 

[clintrubber]

[quote author="rodabod"]
I agree. Resonance occurs when the impedance is minimised in the LCR network.

So, at resonance, if the LRC is at minimum impedance then is it not loading down the capsule?[/quote]
There's series & parallel resonance (ok, and som stuff in between, but let's stick to filters here, not oscillators).

(Do you have the Tremaine-book/file ? Good pics on page 68.)

Ignoring R for now, for series the impedance of L+C drops to little, parallel: L//C becomes high. So when the LRC-series branch is at minimum impedance (and it is for that ~4kHz) then loading is the most, resulting in a dip in the freq response, like in Dayvels graph.

Bye,

Peter

 

[PRR]

> resulting in a dip in the freq response

Yes, I was wrong.

So they had a peak at 4KHz and didn't want it, LC notching was somehow better than fixing the capsule.

 

[clintrubber]

[quote author="PRR"]So they had a peak at 4KHz and didn't want it, LC notching was somehow better than fixing the capsule.[/quote]
In general one could wonder how mechanical & electrical stuff that has to counteract would match & would keep track under the conditions it'll be subjected to.
But as you pointed out the Q isn't high so I guess it'll work OK enough then.


Regards,

Peter

 

[rodabod]

[quote author="PRR"]So they had a peak at 4KHz and didn't want it, LC notching was somehow better than fixing the capsule.[/quote]

I have a frequency response chart of the microphone without the LRC installed - is anyone interested? I could post it.



[quote author="clintrubber"] In general one could wonder how mechanical & electrical stuff that has to counteract would match & would keep track under the conditions it'll be subjected to.
But as you pointed out the Q isn't high so I guess it'll work OK enough then.
[/quote]

I'm not sure I understand what you mean by "matching" and "tracking" under different conditions. Can you explain what you mean?

I see what you mean, the Q=3 above is just meaningless, it's ignoring the surrounding stuff.

I don't quite undertsand this either. What "surrounding stuff"?

The filter does have a Q of ~3 right? Do you just mean its effect on the response isn't that noticeable?

Roddy

 

[clintrubber]

Hi,


[quote author="rodabod"][quote author="PRR"]So they had a peak at 4KHz and didn't want it, LC notching was somehow better than fixing the capsule.[/quote]

I have a frequency response chart of the microphone without the LRC installed - is anyone interested? I could post it.
[/quote]
Sure !

[quote author="clintrubber"] In general one could wonder how mechanical & electrical stuff that has to counteract would match & would keep track under the conditions it'll be subjected to.
But as you pointed out the Q isn't high so I guess it'll work OK enough then.

I'm not sure I understand what you mean by "matching" and "tracking" under different conditions. Can you explain what you mean?

[/quote]Sorry, let me explain. What I had in mind (but what's of course less clear for the world outside that mind of mine ) was this: 'mechanical' side of the story is the capsule, mic-body, grille etc. It's made with a certain precision/tolerance. The 'electrical' side of it is that RLC you showed. Has a tolerance as well. So that's the starting point: they're intended to counteract/cancel each other. And unless both are made in a Swiss watchmakers-workshop the counteracting won't be perfect. So I used 'matching'.

Then in addition to that the actual life of the mic begins, likely resulting in changes in the mechanical & electrical values of the parameters involved. I called that 'tracking'.

The mismatch at the start could very well be the most influencing factor already.

I see what you mean, the Q=3 above is just meaningless, it's ignoring the surrounding stuff.

I don't quite undertsand this either. What "surrounding stuff"?

The filter does have a Q of ~3 right? Do you just mean its effect on the response isn't that noticeable?

Roddy

Right, the surrounding stuff: everything else than the series-LCR.
As PRR pointed out, the resulting (effective) Q is much less, so indeed less noticable.

BTW, sorry I haven't responded to your PM yet.

Bye,

Peter

 

[rodabod]

Hi Peter,

I understand you and PRR now. Thanks for the good explanation!

I will post the frequency response chart once I get my digital camera back from my friend.

Can you guess where there is a peak in the response without the LCR? (Rhetorical question).

 

[Rossi]

Sennheiser's new guitar cab mic, the e906 (the current successor to the ancient MD409) has a frequency switch. From what I heard it must use a similar kind of passive filter for the upper midrange (peak, flat, dip).

BTW. I have an older MD409U. When I took it apart recently, I discovered that it is actually an omni capsule. But the diaphragm is so large that it becomes a cardioid at mid to high frequencies.

The Beyer M201 is an awsome mic. I'm curious to see your frequency plot without the filter!

 

[rodabod]

Ok, here we go:

Like I said before, spot the peak at 4KHz.

You can compare this to the frequency response for a B*eyer M201 - notice the lack of the peak there.

It's a nice response though. Those Germans really know how to make great dynamic mics.