Balanced High Pass Filter

[bruce0]

These are the little inductors.
The metal case is 55mH, 55 Ohms DCR
The plastic case is 118mH, 143 Ohms DCR

 

[abbey road d enfer]

These are the little inductors.
The metal case is 55mH, 55 Ohms DCR 

That one would be barely usable (about 10dB attenuation only)

The plastic case is 118mH, 143 Ohms DCR

This one not. Only 6dB attenuation in the best case.

 

[earthsled]

I got several lovely little shielded inductors from Nebraska Surplus

Your inductors look nice bruce0, thanks for the info. 

Does Surplus Sales have any more of the metal case 55mH models, or did you get the last of them?

Some have too high DCR; in your application, you're looking at DCR >40 ohms.

I measured the original MD421 inductors at L1 = 31.8R and L2 = 45.2R

Mouser's best 82mH choke is listed as having a DCR of 71R, so probably too far out of spec., I would imagine.

Thanks, all!

 

[bruce0]

actually... it wasn't nebraska surplus it was here... they appear to have them.

http://www.electronicsurplus.com/Item/4978/TRW%20-%20Inductor_%20coil_%2054mH_%20Shielded_%20-%20TFR20YY%20%20CLF-690/

 

[abbey road d enfer]

Some have too high DCR; in your application, you're looking at DCR >40 ohms.

I measured the original MD421 inductors at L1 = 31.8R and L2 = 45.2R

Mouser's best 82mH choke is listed as having a DCR of 71R, so probably too far out of spec., I would imagine.

Sorry, typo; I meant DCR <40ohms.
But you're right, 71 ohms is too high, unless you want very mild attenuation.
What construction are the inductors in the Sennheiser?

 

[bruce0]

The MD421 specs say 200 ohm output impedance at 1000 Hz, so rolloff at 863 ish using the calculations matches the graph well. Newest Sennheiser product sheet shows attenuation all the way up to 10K, maybe because of the load this puts on the mic?

Original post looked for TLM103 filter, that has output impedance of 50 ohms transformerless, and min load impedance of 1000 ohms. 

What happens when a big LF frequency wind flutter goes out of the transformer-less buffer on the mic and into what looks like 100ohms load presented by this filter, is there a concern that that overloading the mic might cause distortion in the high frequency signal at that moment?  Maybe a little HF distortion is ok?

I think this is a fantastic thread, thanks for keeping it going (all these years)

Maybe post is no longer talking about the TLM103?

 

[abbey road d enfer]

The MD421 specs say 200 ohm output impedance at 1000 Hz, so rolloff at 863 ish using the calculations matches the graph well. Newest Sennheiser product sheet shows attenuation all the way up to 10K, maybe because of the load this puts on the mic?

That's very likely. Inductors have losses that are equivalent to an additional load.

What happens when a big LF frequency wind flutter goes out of the transformer-less buffer on the mic and into what looks like 100ohms load presented by this filter, is there a concern that that overloading the mic might cause distortion in the high frequency signal at that moment?

You are absolutely right, this is a concern. For a condenser mic (or any mic that relies on built-in amplification), a 2nd-order filter (C-L-C) may be preferrable because its input impedance is higher. The difficulty is passing phantom power, so the filter becomes something like 3 caps (one big), 3 resistors, one inductor. Obviously, it's much easier to do that in the head amp.

 

[tv]

With transformerless pres there will usually be additional capacitance (input caps), adding an additional reactance into the passive circuit. The preamps' input impedance may be seen as a series resistance (~~) to caps' reactance and may be sufficiently high to be seen as unimportant - but what happens if the micpre has "impedance tweak" pot going sufficiently low (and not necessarily in front of the caps, making this essentially a L-C-R tank, and the pre of course "listening" what happens at the "R" node). This is likely going to skew the response in lower mids (from top of my head). Perhaps even more than pres with transformer input?

Enough offtopic..

 

[abbey road d enfer]

With transformerless pres there will usually be additional capacitance (input caps), adding an additional reactance into the passive circuit. The preamps' input impedance may be seen as a series resistance (~~) to caps' reactance and may be sufficiently high to be seen as unimportant - but what happens if the micpre has "impedance tweak" pot going sufficiently low (and not necessarily in front of the caps, making this essentially a L-C-R tank, and the pre of course "listening" what happens at the "R" node). This is likely going to skew the response in lower mids (from top of my head). Perhaps even more than pres with transformer input?

If the designer has done his homework, the reactance of the caps will act at a very low frequency. Impedance switches or pots are usually before the caps, so it won't interact with them.
It will however interact with a passive HPF. In the case of a simple C-R filter, it will shift the frequency upwards. In the case of an CLR filter, it will reduce the Q, making the filter less steep but over a larger BW.

 

[earthsled]

Sorry for the delay...

What construction are the inductors in the Sennheiser?

I've attached an image. Upon closer inspection, the inductors look more like a transformer - sharing the same core. Unfortunately, it doesn't look like an off-the-shelf type of component.

For a condenser mic (or any mic that relies on built-in amplification), a 2nd-order filter (C-L-C) may be preferrable because its input impedance is higher. The difficulty is passing phantom power, so the filter becomes something like 3 caps (one big), 3 resistors, one inductor. Obviously, it's much easier to do that in the head amp.

This sounds like a design I'd like to try - a filter that could accommodate both passive and active mics, located in the preamp rather than the mic itself.

It will however interact with a passive HPF. In the case of a simple C-R filter, it will shift the frequency upwards. In the case of an CLR filter, it will reduce the Q, making the filter less steep but over a larger BW.

If the filter is kept passive, could it be designed to accommodate a wide variety of mic outputs and preamp inputs while still doing an adequate job of reducing LF rumble?

Thanks everyone for your comments!

 

[abbey road d enfer]

Upon closer inspection, the inductors look more like a transformer - sharing the same core. Unfortunately, it doesn't look like an off-the-shelf type of component.

As I surmised, they had to use a core of about the same size as they would for a mic input xfmr.

For a condenser mic (or any mic that relies on built-in amplification), a 2nd-order filter (C-L-C) may be preferrable because its input impedance is higher. The difficulty is passing phantom power, so the filter becomes something like 3 caps (one big), 3 resistors, one inductor. Obviously, it's much easier to do that in the head amp.

This sounds like a design I'd like to try - a filter that could accommodate both passive and active mics, located in the preamp rather than the mic itself...        ...If the filter is kept passive, could it be designed to accommodate a wide variety of mic outputs and preamp inputs while still doing an adequate job of reducing LF rumble?

It can be done, however it's gonna be a compromise. As microphones' actual impedance vary from 100 to 500 ohms, you have to shoot in the middle, so let's use the nminal 200 ohms value.
Then you have to target a corner frequency. Let's say you want 80Hz.
You end up with the basic circuit of 2x 22uF caps and one 300mH inductor. You can't use the standard x-over calc formulae because the source impedance is not zero.
A simulator is a great help there.
The response varies significantly in regard to the actual mic impedance, not so much in regard to load impedance. I'll put the graphs in the next post.
As I think you want to make it switchable, I have also drawn two possibilities, whether the input is with a xfmr or not.

 

[abbey road d enfer]

Red is nominal 200r source
Cyan is with 100r source
Green with 400r source

 

[tv]

...
The response varies significantly in regard to the actual mic impedance, not so much in regard to load impedance
...

I also played with the sim wrt to my previous remark and your statement is true when you deal with a "well-behaved" micpre.

When you lower the "Z-in" enough, like when a preamp has a "Z-in" knob, the frequency response starts looking funnier. So my hunch vas valid, but my simulation was a bit too naive to fully trust it.

Btw. do you have any "good enough for most cases" simulation of a "mic source"?
I used an ordinary R at 200 Ohms as a mic "sim", but that's quite, well, naive.

 

[abbey road d enfer]

Btw. do you have any "good enough for most cases" simulation of a "mic source"?
I used an ordinary R at 200 Ohms as a mic "sim", but that's quite, well, naive.

I've spent some time working on it when I found I had a ribbon mic acting funny.
Dynamic and ribbon mics can be modelled with a resistor of about nominal value, an LCR circuit that simulates the impedance peak created by mechanical resonance and an inductor that shows the inductive nature of the transducer. However no two mics are equal in this respect; you can spend some time and create a rather accurate model of an SM57 or a Royer 121, but that would be useless because the behaviour of mic preamps is predictible enough in the presence of the simplest model.
Condenser mics are easier to model; transformerless are an almost perfect resistance, those with transformers are accurately modelled by a resistor and an inductor in series.
You will probably learn more by using different values of resistor than by using a complicated model.

 

[tv]

Mehhhhhhh ... condensers are only easier to model, imho, if they are the "open emitters", schoeps type output. Otherwise there will be all sorts of weird hacks wrt the coupling and love/hate relationships with the micpre inputs.

I was hoping to find something simple to implement wrt the dynamic mic "sim", like an ac current source driving the (mic coil) inductance followed by mic's .... 200 Ohm resistor, haha. Iow, something that "just works", similar to some hacks that represent the guitar/bass pickup.

Offtopic, ignore... SPICE GIRLS ALERT

 

[earthsled]

Hi all,

I just wanted to know if this Tamura inductor (attached PDF) would fit the spec for Abbey's HPF circuit. It seems to have an acceptable DC resistance of 40R. Is 4mA okay for the DC current?

Thanks!

 

[abbey road d enfer]

I was hoping to find something simple to implement wrt the dynamic mic "sim", like an ac current source driving the (mic coil) inductance followed by mic's .... 200 Ohm resistor, haha. Iow, something that "just works", similar to some hacks that represent the guitar/bass pickup.

Isn't it exactly what I wrote: "those with transformers are accurately modelled by a resistor and an inductor in series."

"love/hate relationships with the micpre inputs" is an anthropomorphic concept that has nothing to do with electronics.

 

[abbey road d enfer]

Hi all,

I just wanted to know if this Tamura inductor (attached PDF) would fit the spec for Abbey's HPF circuit. It seems to have an acceptable DC resistance of 40R. Is 4mA okay for the DC current?

Thanks!

Due to the rather large DC resistance, the slope will be a little less steep - about 10dB/oct instead of 12.
In any case, there is no DC current in the inductor.

 

[earthsled]

I'm fairly new to circuit simulation, but I've recently been trying out the online version of CircuitLab.

Here's my first attempt at simulating Abbey's filter:
https://www.circuitlab.com/circuit/q7q4qd/2nd-order-hpf-v1/

The simulation results are attached.

I wasn't quite sure what parameters to set for XFMR1. The software allows for turns ratio, DC resistance for primary and secondary, and inductance for the primary. Currently, I have these parameters at the default values.

Please let me know if I'm on the right track for the simulation.

Thanks!

 

[abbey road d enfer]

I wasn't quite sure what parameters to set for XFMR1. The software allows for turns ratio, DC resistance for primary and secondary, and inductance for the primary. Currently, I have these parameters at the default values.

Turns ratio is irrelevant to this calculation; only the primary matters. A good start is L at 5H and DCr at 20 ohms.