mic pre HPF

[mitsos]

OK, so I was toying with the idea of throwing a HPF into a (312) pre I have here.  I thought I could do it simply with a series C and shunt R (a pot with R in series, to vary Freq).  Is this wise? I was reading another thread saying the frequency would change with the mic attached. Well, I did some playing in Excel and I found that with a 0.1uF series C with a 50K shunt pot in series with a 7K5R (so R would vary from 7K5 to 57K5) would yield a HPF with -3dB between 28Hz and 213Hz.

Can this work? I assume from the other thread that this should not be put on the input, so where would a good place be for something like this? After the input trafo maybe? Or should it be done another way altogether?

 

[mikefatom]

Mitsos,

I plan on doing something similar to my 312 pres. I think that a switchable RC network after the 2520 would make more sense. The output is low impedance and will not have as great of an effect on the cut off freq. A simple voltage follower buffer circuit would probably work as well to isolate the 2520 from the rc network.

I will try both and post the results if I get around to it.


Good luck

mike

 

[abbey road d enfer]

Putting the HPF before the xformer is definitely a bad idea (although it's been done a number of times), because the xformer sees a high impedance, which generates distortion.

 

[mikefatom]

Interesting... I shall try both before and after the 2520 to see which sounds best. 

 

[mitsos]

Thanks for the replies/suggestions guys.

My first thought was to put it at the input. As I said above I found a thread (rochey's I believe) that said it was a bad idea...that along with Abbey's comment tells me it's a no-no. My next thought was to do it after the the trafo, which would be easier; as Chae pointed out only one RC nework needed.  But I'm curious, since we have relatively high impedances, is it best to do it after the opamp, as mike suggests?  In either case, I want to avoid any buffers, and just have an RC network.

Does the RC network I described above make sense? I guess it wouldn't even need to be switchable if I could get the frequency low enough at one end of the pot travel, but I know it'd be best to get the thing out of the circuit when not wanted with a SPDT switch.

Mike, are you thinking of doing a fixed frequency (HPF on-off) type thing of some type of variable F?

 

[mikefatom]

I was thinking more of an on-off/fixed 120Hz HPF.

I believe that using a pot for continuous freq selection, means planning out the min and max cut-off freq for you HPF while staying away from affecting the hi-freq response of the circuit. An active circuit would probably be best in this case.

 

[mikefatom]

I put this quickly together. Probably not perfect but just to give the gist of it.
What do you think?

It's to be inserted before or after the 2520.... depending on sonics.

opinions?

Thanks

Mike

 

[CJ]

there is nothing wrong with doing that, just make sure that the reactance and resistance are high

compared to the value of your source and you should be alright.

if they are low, like close to the mic's Z, then the magnetizing current, which follows the B-H curve of the

particular input core you are using, will flow into the RC circuit. since this mag current follows B-H and not a

sine wave, it will cause distortion on your output also, since what happens on the pri, comes out the

ass end..

now there is already mag current flowing threw the mic transformer on a 57, unless you removed it,

if you can find the right cap/resistor combo, you might even be able to improve the distortion specs by

off setting the weird mag waveform generated by the mag current thru the source,

 

[mikefatom]

Thanks CJ for the input! I knew the input transformer impedance would be something to watch out. Your knowledge of transformers is way beyond mine. I vaguely remember field theory  :-[. Much less than I would like...

R2 in the schemo is there to insure a high enough input impedance in case this would be inserted right after the input TX.
Make up gain could then be applied to counter act the insertion loss of the passive filter.

 

[CJ]

have you tried messin with a cap across the feedback resistor? that is easy pickens for tone contour.

build an api and hard wire in a 57, tweak the cap.

build another api and hardwire in an electrovoice, tweak the cap, by the time you get 8 of theses custom suckers, you have some game.

 

[Harpo]

A non-inverting integrator between DOA out and its inverting input by a low-offset/low-drift opamp might do it, though not being passive-only. Basicly a servo circuit, but -3dB cutoff set some octaves higher than usual. A 100nF/10K for upper -3dB @ 160Hz, 1M rheostat in series to this 10K varying this down to 1.6Hz. The big lytic in the shunt arm of the mic pre circuit could probably be replaced by a wire link.

@CJ "have you tried messin with a cap across the feedback resistor?"
for varying LPF.

 

[peterc]

The big lytic in the shunt arm of the mic pre circuit could probably be replaced by a wire link.

This has an effect on the low end. What if this was decreased to cut off freqs up into the  20 - 200 Hz area?

Peter

 

[Harpo]

The big lytic in the shunt arm of the mic pre circuit could probably be replaced by a wire link.

This has an effect on the low end. What if this was decreased to cut off freqs up into the  20 - 200 Hz area?

Peter

Varying the gain setting resistor (quite normal for a preamp) will vary the HPF frequency.
-3dB cutoff is set by 1/(2*PI()*R*C). Varying R (between 200R and 22K2 for the usual 312 mic pre with 20K feedback resistor) and fixed C at maybe 220uF, this is between 0.036Hz @ 20K2 and 3.6Hz @ 200R for highest gain.
Exchanging this 220uF with a 4u7 cap will vary cutoff frequency in range of 1.7Hz to 170Hz, but depending on dialed in gain setting resistor, so probably not desired.

 

[mitsos]

Harpo... curious about the DC-servo type thing you mentioned. Would you switch in the HPF at the servo output?

You also said the shunt lytic can be jumpered. Is this lytic there to block DC from the pot? So if you have a DC servo (and hence no DC offset) you don't need it, right? 

 

[Harpo]

mitsos, something like this, just for the idea, whatever its worth...
(input stage and  transformers obviously missing, no 2520/1731/990/.. DOA in my lib, so picked a AD797 instead. Servo could be opa604,ad706,lm11,.., depending on supply voltages. Added a bias trim, similar to J.Hardy's MPC-1, whatever offset polarity your DOA  may throw out).
Caps in shunt arm and at DOA output could be shorted out (at the cost of the servo circuit and might buy you other artefacts).

 

[mitsos]

ok, that's a bit to ponder... actually you kinda lost me.  ;D  Is it necessary to have both switches? and then do you lose the servo, as you say?

 

[Harpo]

Is it necessary to have both switches?

If you like the hpf at defined positions, yes, else substitute both switches (could be a 2x6 lorlin) and switched resistors with the 1M pot+10K as drawn below left side, though the usable range of the pot seems silly small. It might be better rearranged as 1M+10K+50K pot and shorting out the 1M for 'hpf ON' as drawn below right side. Maybe reload updated schematic.

and then do you lose the servo, as you say?

obviously my crappy english, this hpf is the servo.

 

[simonsez]

Typically the best place to do it is at the input, to avoid amplifying the low-frequency energy that you don't want, thus reducing your total headroom.

Agree...I'm building green pre now, is it possible to add hpf at the input?

 

[abbey road d enfer]

Typically the best place to do it is at the input, to avoid amplifying the low-frequency energy that you don't want, thus reducing your total headroom.

Agree...I'm building green pre now, is it possible to add hpf at the input?

Please read carefully all posts. Putting the HPF at the input creates CMRR problems (transformerless pre) and non-controlled behaviour in regard to source impedance and gain. Worse if you use a transformer.

 

[simonsez]

Thanks Abbey, my logical says that it's better to filter the rumble before amplify it, but  as you say... it will create another problem...thanks for telling me...