Best place for an active filter?

Status
Not open for further replies.
S

StellarRat

Member
I'm building a spy ear type sound amp. I have an electret microphone connected to an op-amp pre-amp (OPA2134) that feeds an op-amp active filter (300hz - 3000hz bandpass (OPA2134) ) and that feeds a power amp (TPA4861) for the headphones. My question is should filter be the first stage after the mic or where I have it now (after the pre-amp)? I know it can't be last stage because it could never handle the power from the power amp. The circuit is currently fully operational, but I wondering if moving the filter would improve it.

Also, I've seen some circuits of a similar type that actually have a FET as the first pre-amp and that feeds an op-amp instead of just an op-amp as the pre-amp. Is there any particular advantage to that?
 
Last edited:
I would put it as far down the signal chain as possible so it rolls off all the noise from those circuits, thus just in from of the power amp as you have it is good.
 
crutschow said:
I would put it as far down the signal chain as possible so it rolls off all the noise from those circuits, thus just in from of the power amp as you have it is good.
Click to expand...
That's interesting. Isn't it better to get rid of the signals you don't want to amplify as early as possible in the chain? Can you explain your reasoning in more detail?
 
The amplifiers contribute noise of their own. If you are talking about eliminating the filter stage by incorporating the filter characteristics into the preamp feedback network, so that the preamp drives the power amp directly, that certainly will work. But if there is to be an intermediate gain stage between the preamp and the power amp anyway, put the filter there and it will reduce some of the noise created by the preamp.

ak
 
I was more worried about extraneous tones coming in from actual sound sources like fans, cars, etc... My goal is to hear human speech at a distance. My thinking is that the op-amp noise is probably minimal compared to all the garbage sounds generated by our machines and picked up by the microphone. I was thinking that eliminating garbage sound is more important than any noise the electronics generate. From my admittedly newbie electronics knowledge, my understanding is that op-amps perform better when the frequencies they have to amplify are narrower. Of course, I could be all wet and I welcome more knowledge.

BTW, I want to say thank-you to AudioGuru because I'm using his electret pre-amp with my circuit. I changed the op-amp from a TL071 to a OPA2134, but it still seems to work just fine (obviously I rewired the pin configuration.)
 
Last edited:
StellarRat said:
my understanding is that op-amps perform better when the frequencies they have to amplify are narrower.
Click to expand...
Don't know where you picked that up, but - no. That sounds like someone was using the wrong part for the job, like using a LM741 or LM358 to make 60 dB of gain at 20 kHz. Of course the device must have enough gain and bandwidth to do the task. Beyond that, there is a thing called intermodulation distortion that audio designers worry about, but it doesn't really apply to a relatively narrow bandwidth like yours. You are using high-end audio grade parts for a low-end telephone-grade project. You should have no "perform better" problems.

This sounds like a shotgun or parabolic microphone project - ?

ak
 
Yes, that was my original intention. I haven't got as far as a directional set-up yet. It's actually pretty amazing how well a sensitive microphone just hanging in space works with the circuit so far. I found some sensitive mics at Mouser (-25 DB) and so far I'm quite happy with the results with zero mic housing. What got me interested was my alarm system. They can hear people talking in the bathroom all the way across my house (that was the test.)
 
StellarRat said:
That's interesting. Isn't it better to get rid of the signals you don't want to amplify as early as possible in the chain? Can you explain your reasoning in more detail?
Click to expand...
The preamp will contribute some undesired noise, even if its small, so that's why I suggested putting the filter after it.

And the active filter circuit itself will have some in-band noise that would be amplified by the preamp if the filter was before the preamp.

Putting the filter after the preamp won't affect the filtering of the ambient audio noise outside the desired frequencies unless the noise is high enough to saturate the preamp, since a filter rolloff is not absolute but is relative to the passband signal.
 
Thanks crutshow! I obviously have a lot to learn. If I had any math skill I think I would have much preferred to be an EE (had I known.) I couldn't handle calculus. They told me I me I was really good with concepts, but not so good with execution. I ended up in computer software/hardware, but I've never really thought that was my place. A bit late now, I'm 56. LOL. Anyway, I'm doing this to stimulate my old brain.
 
Generally, the first stage of the amplifier chain should have the highest gain, to get the signal up and out of the noise. This amplifier needs to have the highest gain bandwidth and lowest noise (to enable high gain while adding minimal extra noise to the system)

The input impedance of an amplifier goes down as the loop gain goes down. The loop gain goes down as the closed loop gain goes up, which contradicts the design brief above (keeping the closed loop gain of the first stage high).

However, the input impedance of the OPA2134 is very high so you should be able to gain up the input signal without it adversely loading the output of the microphone.
 
So, since I should get as much gain as possible from the pre-amp is there a way to boost the gain from the pre-amp circuit I'm using? It's same one that sticky'ed on the home page of this section under "Basic opamp circuits : Audioguru's electret mike pre-amp". All I did was substitute an OPA2134 in place of the TL071. If the opamp I'm using is more capable maybe something can be changed? I tried the TL071 first and I really can't tell much difference in the sound quality or gain (would not expect a gain difference though without changing something.)
 
The benefit is basically due to the gain difference between the closed loop gain and open loop gain. The part you chose has an 8 MHz gain BW so it is a very good one. I don't think you need to change anything. You should set the gain of the amp stage to get the signal up to the level you need.

If you are concerned about noise, I recall the NE5532 is one of the best for that and also has a very wide bandwidth (like 10 MHz). It used to be the go to part for pre amps. It is not FET input so it may not match your microphone as well as a FET device.
 
Last edited:
A TL071 and OPA134 have almost the same very low distortion and fairly wide frequency response but the OPA134 has less hiss noise.
If you have too much gain then loud sounds will cause an amplifier to produce clipping. The gain of my electret mic preamp is simply 1+ (RV1/R4). So if RV1 is 470k then the gain is 215 and if RV1 is 2.2M then the gain is 1001.
 
I would think that a design for a mic pre amp would benefit from some kind of AGC (auto gain control) that would amplify enough to boost a wide range of input signals so the output would be in a specified signal range.
 
bountyhunter said:
I would think that a design for a mic pre amp would benefit from some kind of AGC (auto gain control) that would amplify enough to boost a wide range of input signals so the output would be in a specified signal range.
Click to expand...
Yeah, it would save my ears too when there's an expected loud noise.
 
A really good AGC circuit is much more tricky than most people think because they (AGC circuits, not most people) are stupid. They can't tell the difference between a loud clang and a loud vowel, and they can't interpret either and make a judgement about whether they represent something worth correcting because they are the first of many such noises, or something to just clip and let go because the next sounds will be quiet again. Your brain does this in microseconds. Chips can't.

Adafruit has an AGC mic preamp module based on a Maxim mic preamp chip, and it is the easiest way I've ever seen to evaluate a simple AGC to see if it does more good than harm in your application. Next is a more complex AGC based on the RMS value of the audio rather than the peak value, but that is a very different discussion.

ak
 
bountyhunter said:
I would think that a design for a mic pre amp would benefit from some kind of AGC (auto gain control) that would amplify enough to boost a wide range of input signals so the output would be in a specified signal range.
Click to expand...

Generally that's a VERY bad idea - far better to have a manual gain control, so YOU are in charge of what's going on.

For a simple tape recorder (for amateurs to use) it's a good idea, but that's about all.
 

I know that the electronic ear muffs for shooting have some limiting circuits that work pretty well.
 
Status
Not open for further replies.

Similar threads

  • Question
NEW INFO! Electrostatic Air Filter Using Neon Transformer?
Replies
4
Views
2K
rjenkinsgb
7
  • Question
im looking for an 8 bit push on, push off latch dip ic
2
Replies
22
Views
3K
danadak
Best low voltage pulldown resistor
Replies
10
Views
2K
rjenkinsgb
J
  • Solved
Suggestions for an Angle Sensor (one axis)
Replies
17
Views
3K
_John_
J
S
  • Question
Is this an encoder?
Replies
6
Views
1K
stuhagen
S
Menu
Log in
Register
Cookies are required to use this site. You must accept them to continue using the site. Learn more…