Neumann W75k passive bandpass inductor values?

GroupDIY Audio Forum

Help Support GroupDIY Audio Forum:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.

MrShhh

Well-known member
Joined
Oct 26, 2004
Messages
98
Location
Northamptonshire, UK
Hi, I'm trying to make sense of this schematic:

http://i985.photobucket.com/albums/ae333/MrShhh_2009/W75kschematic.jpg

Sorry about the quality, it's all I could get hold of  :-\ You might have to zoom right in to read some of it.

1) Does anyone know what kind of passive bandpass this is? If I know the type, perhaps I can research how to correctly work out the inductor tapping values for Dr4 and Dr5.

Is it as simple as working out the crossing of the function for the capacitive reactance Xc = 1/(2 Pi * f * C) and the function for the inductive reactance XL = 2 Pi * f * L?

2) Any ideas what Dr1 is? 2 inductors back to back? Some kind of ferrite bead? A transformer? If it's an inductor, there seems to be no way of knowing how many mH it is.

Thanks  :)




 
MrShhh said:
1) Does anyone know what kind of passive bandpass this is? If I know the type, perhaps I can research how to correctly work out the inductor tapping values for Dr4 and Dr5.
The highpass is a 2nd order job essentially governed by Dr2 & 3 in conjunction with R5/C15. The lowpass is a 3rd order balanced, so the equations are not as easy as the ones you have posted, which are correct only for 1st order filters. I suggest you enter the schematic in LTSpice and tweak the inductors till you get a more or less Butterworth response. Shouldn't be very difficult since both inductors are identical.
2) Any ideas what Dr1 is? 2 inductors back to back? Some kind of ferrite bead? A transformer? If it's an inductor, there seems to be no way of knowing how many mH it is. [/quote] Dr1 is a common-mode filter that rejects RF interference but does not affect the signal. Essentially a bifilar inductor
 
Fantastic. Thanks again abbey road for helping out ;D

Dr2 & Dr3 are variable inductors.

Presumably they're variable so you can set the amount of high frequency roll-off: 52mH gives XL=6535ohm @ 20kHz.

Or maybe they're for signal balance adjustment? Probably both at once, since one must affect the other.

Could I get away with fixed value L here? (Depends I suppose, on how well matched the components in each signal path are, although components have come a long way since 1959).

I'm trying to find a decent way of running LTSpice on OS X. When I've done the simulation (which should be interesting, I've never done one before) I'll post the results.

Cheers

 
It's always difficult to find out what the designers had in mind, but I think, since these units were for broadcast and the stringence of regulations in Germany, that the inductors were adjustable to make sure the attenuation fell perfectly in the tolerances. Also it is possible that they were fine-tuned for perfect HF common-mode rejection.
What is your interest in this unit? What precision do you expect?
Fixed values should be acceptable unless you really want to adhere strictly to standards...it is only pertinent to FM broadcast, nothing else
Have you been to the LTSpice forum?
http://tech.groups.yahoo.com/group/LTspice/
They may help you with OSX.
 
abbey road d enfer said:
What is your interest in this unit?

I'd like to recreate the filter to use between a U47 mic and a V72 pre as they're very hard to find in good order.

abbey road d enfer said:
What precision do you expect?
I doubt I need the precision required for FM broadcast! I'll use fixed values to begin. I don't want to introduce more noise than necessary but the passive design should be quiet enough hopefully.

abbey road d enfer said:
Have you been to the LTSpice forum?
Not yet, thanks  ;D
 
MrShhh said:
abbey road d enfer said:
What is your interest in this unit?
I'd like to recreate the filter to use between a U47 mic and a V72 pre as they're very hard to find in good order.
OK, I understand for the highpass, but do you really want the lowpass? Since you don't need the RFI filtering nor the 15kHz cutoff, that would make for a much simpler build.


 
I'm after the "sound" of that particular mic/bandpass/pre combo, and the 15kHz cutoff is part of it, but I will try without the lowpass first. As you say, it will be much easier. If it's no good, I can change it later.
 
Ok, it took a while but now I have LTSpice running in Virtual PC.

Schematic:
http://i985.photobucket.com/albums/ae333/MrShhh_2009/W75kLTSsch.jpg

AC Analysis:
http://i985.photobucket.com/albums/ae333/MrShhh_2009/W75k16HzNoseriesR.jpg

This doesn't look quite right yet. Have I missed anything? The only thing I can think of is loading and the series resistance of the inductors. I have not implemented these at the moment.

LTSpice kicks ass  ;D thanks abbey road!

PS Is there any way to show analysis curves for multiple component values overlaid on a single plot? It's really awkward having to change component values and rerun the analysis each time, losing the plot you had before.
 
MrShhh said:
Is there any way to show analysis curves for multiple component values overlaid on a single plot? It's really awkward having to change component values and rerun the analysis each time, losing the plot you had before.

In most simulators you can save the curves and later import them into your current plot. I haven't used LTSpice, so I don't know how to do it in that specific software.

You can also make multiple copies of your circuit in the same drawing which allows you to simulate them at the same time. (But there can be some limitations of maximum number of components in one drawing...)
 
audiox said:
In most simulators you can save the curves and later import them into your current plot.

Cheers audiox, I hadn't thought of that, I'll give it a go  ;D

audiox said:
You can also make multiple copies of your circuit in the same drawing which allows you to simulate them at the same time.

Tried this, but it opens them all in separate analysis windows  :-\
 
MrShhh said:
audiox said:
You can also make multiple copies of your circuit in the same drawing which allows you to simulate them at the same time.
Tried this, but it opens them all in separate analysis windows  :-\

Draw the same circuit two (or more) times on the same page and connect inputs of all these sections to the same generator. Now the software treats it as a one circuit and you are able to plot the frequency response for all the sections simultaneosly. Just place a plotting probe to the output of each section.
 
Thanks spreemusik, I'm looking into that, it sounds very useful. The help file I'm finding difficult to understand just yet, but I'll stick at it.

Anyone know how to model how the source and load affect the circuit?

e.g. suppose I want to look at how a 50ohm mic source & 600ohm pre load affect the performance of the circuit?

I'm happy to read tutorials if you know of any good ones.

 
Parameters are actually very easy, once you get the hang of it. Instead of your value for L, for example, you put {Lx} , then a spice directive .params Lx=100m Now the inductor has 100mH. With .step param Lx 200m 1000m 100m you can step the value from 200mH to 1000mH in 100mH steps.

For the source, you can specify the series resistance in the voltage source, works for me. I guess it can be a lot more elaborate. For a load, I simply use a resistor ;-)
 
spreemusik said:
For the source, you can specify the series resistance in the voltage source, works for me. I guess it can be a lot more elaborate.
I found the parasitic series resistance earlier & thought "nah, that's far too easy"  ::)

spreemusik said:
For a load, I simply use a resistor ;-)
Same again, seems too easy! I'm not complaining though.  ;D

Having said that, there's probably a dozen different ways to make it super-difficult....

Cheers spreemusik!
 
I know this is an old thread, but thought it would be better to ask here than open a new one...

I'm also interested in building this circuit and wanted to understand some components.

Especially the two 100 Ohms resistors at the input and the 900 Ohm at the output.
The unit is made to use between the mic and the preamp, right?
Are the two 100 Ohms resistor not to low for the common 200 ohm outputs of the microphones? Would that be an impedance matching scenario which results in level loss?

And what about the output resistor?
The input impedance of most preamps is between 1k-2k ohm. Wouldn't the input impedance of the preamp be in parallel with the 900 Ohms and thus lowering it?

I tried to simulate the circuit in spice and get bad results when I strap a 1k5 resistor at the output. Seriously cutoff at both top and bottom... Without the load it is flat from 40Hz to 15Khz. Like the original.

I feel like I'm missing something  ::)

Edit: Ok, I read somewhere that this unit was also made to use before the V72 preamps which have an input impedance of about 2k Ohms. With that load it seems ok... lowering it results in worse response. So replacing the 900 Ohm resistor with a 25k pot would give more flexibility with other preamp inputs, right?
 
Murdock said:
Edit: Ok, I read somewhere that this unit was also made to use before the V72 preamps which have an input impedance of about 2k Ohms.

The input impedance of a V72 was upwards of 4K if I recall  - The input transformer was a 1:10 and the secondary had a 425K termination. 
 
Winston O'Boogie said:
The input impedance of a V72 was upwards of 4K if I recall  - The input transformer was a 1:10 and the secondary had a 425K termination.

Yeah, there seems to be different revisions. One had a 250k termination resistor and another one has 400k.

Another little question.
The termination resistor on the secondary. Is it in parallel with the primary reactance?
So if I had, for example, an input transformer with a primary reactance of 1000 Ohm at 40Hz. How would the secondary termination resistor affect the input impedance at that frequency?
 
Murdock said:
Another little question.
The termination resistor on the secondary. Is it in parallel with the primary reactance?

Yes in parallel.  With the correct value of termination, there should be negligible effect on your filter.
I haven't looked at your schematic ( perhaps I should 😛) but these sorts of filters are generally defined by the source impedance - which was usually assumed to be 200 ohms in Europe - and that the load would be sufficiently bridging to that impedance.

Edit:. I looked at the schem but couldn't see clearly enough I'm afraid. 
 

Latest posts

Back
Top