eHam.net - Amateur Radio (Ham Radio) Community

Call Search
     

New to Ham Radio?
My Profile

Community
Articles
Forums
News
Reviews
Friends Remembered
Strays
Survey Question

Operating
Contesting
DX Cluster Spots
Propagation

Resources
Calendar
Classifieds
Ham Exams
Ham Links
List Archives
News Articles
Product Reviews
QSL Managers

Site Info
eHam Help (FAQ)
Support the site
The eHam Team
Advertising Info
Vision Statement
About eHam.net



[Articles Home]  [Add Article]  

Passive Audio Filter

Alan Applegate (K0BG) on March 30, 2007
View comments about this article!

Passive Audio Filter

If you're an aficionado of HiFi SSB, wideband AM, and other enhanced modes, this article isn't your cup of tea. However, if you operate mobile operation where clear reception is essential, then read on.

Passive audio filters have been around for many years. I remember one made during the 70s that utilized a small 2 inch speaker mounted in a tuned cavity made from PVC. At the mouth was a little sleeve you could adjust to set the center of the bandpass. If your receiver didn't have a CW filter, this little jewel worked fairly well. It sold for about $15 at the time.

Passive filters for SSB have been produced over the years by just about every manufacturer of amateur equipment. When these units were first introduced, few radios had built in bandpass or IF shift features, to say nothing of the DSP systems we have nowadays.

Unlike the aforementioned CW filter, all of these designs used a combination of inductors and capacitors typically built into a station speaker. Some were lowpass, and some bandpass. Front panel switches set the mode you desired.

In the early 80s, I built a passive bandpass filter into an old Heathkit mobile speaker. The parts (coils and capacitors) I used were scavenged from a couple of HiFi speaker crossover networks. Besides the bandpass filter, there is a separate high-cut tailored to remove most frequencies over 2,400 Hz. This speaker has served me well for many years, and I still use it!

You might be asking why add a passive audio filter to a radio with a decent IF based DSP? If it were a radio primarily designed for base station use, that question might be difficult to answer. But for a mobile installation, no matter how good the DSP, some high frequency rolloff increases readability.

Designing one isn't difficult. Audio bandpass and lowpass filters aren't any different than their RF counterparts, except for the frequency in question. That, and they're less susceptible to stray capacitance. All of the formulas are available in the ARRL Handbook, and even on line.

\In a recent QST, a New Product announcement described a passive bandpass CW filter from the Xtal Set Society. What caught my eye was the fact the filter could be configured for SSB. A quick look on their web site confirmed this, and that it was designed by Phil Anderson, WXI, of Kantronics fame. It cost all of $19.95 plus shipping. A photo of the CW version of the filter is at left. It took all of two days to reach me via priority mail.

It comes with everything you see in the photo except for 10 feet of #24 hookup wire. About all you need otherwise, is a box to mount it in if there isn't room inside your mobile speaker, and perhaps a jack, and matching plug. For the casual home brewer, having most of the hard-to-find parts all in one envelope is a bonus.

The instruction manual (to be honest) is a little lame. It's geared for the CW version, but an addendum is included covering the SSB version. It also states the filter is in bypass when the switch is in, instead of out. Perhaps this is nit-picking, but to a neophyte it's important fodder. It took me about 30 minutes to put the filter together including the winding of the three toroids. With the help of a little super glue (a hot glue gun would suffice), the toroids were secured in place.

\The circuit board is double sided, but it is NOT plated through, so you'll need to solder on both sides of the circuit board to assure a good connection. I found this out the hard way! It also helps to look at the pictorial which comes with the filter.

There are a couple of caveats to keep in mind. First, the audio amplifiers in amateur radio equipment are almost always transformerless and integrated. This usually means that both leads to the speaker are above ground potential. In other words, you can't chassis ground them, and if you do you run the risk of destroying the IC. Therefore, care must be used when mounting the filter either inside your speaker or in its own box.

One drawback is a reduced audio output level. The amount of reduction depends on the design and cutoff frequencies of the filter in question. It may be anywhere from 2 dB to as much as 12 dB. In this case, it's about 4 dB, so you'll have to turn up the audio to compensate. However, too much audio (way too much actually) can cause the filter to ring, which will cause clearly heard distortion.

The filter works as advertised, although I'd like a little more high frequency cutoff (you might not). The above chart, which came with the filter, exemplifies my bench results. However, I should point out that there can be slight variations depending on the feed and terminating impedances if other than 8 ohms.

Building a passive bandpass filter is a good, hands-on learning tool. Speaking of which, the Hand-On Radio column, starting with the March 2007 issue of QST, starts a series of articles describing the basics of filter design.

Alan Applegate, KBG
www.k0bg.com

Member Comments:
This article has expired. No more comments may be added.
 
Passive Audio Filter  
by WA2JJH on March 30, 2007 Mail this to a friend!
TNX Allen fer excellent artical. I used to foll around with Telco toroids years ago. I got too much ringing.

This would also make for an excellent Newbie project as well.
 
Passive Audio Filter  
by W4QO on March 30, 2007 Mail this to a friend!
Alan, Great article. Some people seem to give more to ham radio than others and you are at the top of the list! Your website if extremely helpful as well.

W4QO
 
RE: Passive Audio Filter  
by K3AN on March 30, 2007 Mail this to a friend!
"some high frequency rolloff increases readability."

Alan, I am puzzled by that statement. As people get older and their hearing starts to go, it's the high frequencies that are lost first. This results in a loss of "readability" and most hearing aids try to compensate by BOOSTING the highs with respect to the lows.

With your filter providing a rolloff in the 1500 Hz range, listeners will lose the "sibilance" sounds so they'll be less able to distinguish between "f" and "s" for example. In my own case, I adjusted the carrier offset frequency in my 1000MP to shift the 1.8 kHz IF filter response to cover a higher range of frequencies (500 to 2300 instead of 300 to 2000). It doesn't sound as natural, but it is far more readable.
 
RE: Passive Audio Filter  
by AB0WR on March 30, 2007 Mail this to a friend!
K3AN:

You are exactly right about voice.

There was a *big* thread about this recently on a different forum.

There were several people who posted various bandwidth "voice prints" mixed with varying levels of white Gaussian noise (i.e. excerpts of Ted Koppel as well as amateur voices). It was obvious that an upper cutoff of 3500hz provides a significant increase in intelligibility in the presence of white Gaussian noise. An upper cutoff of 4500hz is better yet. Even better intelligibility is seen when a 3-to-6db per octave pre-emphasis is applied to the voice spectrum above 500hz.

It is this very concept that the MFJ speech enhancer product operates on. It can be confirmed by studying the Articulation Index graphs in the book SSB Circuits and Systems by Sabin and Schoenike which show the AI levels for various bandwidths of speech both without and with pre-emphasis.

Consider a 2500hz bandwidth audio signal running 100watts average output. In a 0db SNR situation you get an AI of 0.2. To raise the AI to 0.4 (minimum acceptable for a communications link) you have to increase your SNR to 10db. That's an average power increase of 100watts to 1000watts. Since the average voice has a 15db dynamic range you would either have to have a transmitter with a peak power output of 31kw or you would need to do about 12db of clipping/compression which would limit your peak power to about 2kw.

A 3db pre-emphasized 5khz audio bandwidth starts out with an AI of 0.4 at a 0db SNR. Because of the increased noise from the wider bandwidth (3db for twice the bandwidth) you would need a 200watt amp to maintain the 0db SNR. The 3db pre-emphasis cuts the dynamic range to about 12db. That means you need a peak power output of about 1600watts with no further clipping or compression.

The spectrum effiency and interference mitigation issues associated with a 1000watt/2.5khz signal vs a 200watt/5khz signal are many, varied, and open to discussion. The 200watt signal will allow higher propagation stacking levels but the narrower signal allows more signals per hz (if the 1000watt power level allows it due to receiver limitations).

Certainly the wider bandwidth signal should only be considered for use when conditions either require it or allow it. E.g. formal messge traffic handling could benefit greatly from increased intelligibility in noisy conditions as could emergency communicatons, however these don't happen regularly. When the band usage is very low then using wider bandwidths affects no one although it would contribute greatly to lessening listener fatigue when participating in large roundtables.

Certainly significant increases in the understanding of intelligibility have occured in the just the past decade. Dependence on the economic decision that the telephone companies made 75 years ago that 2.5khz is "good enough" is rapidly fading. The big issue is how we integrate this understanding with the associated issues of spectrum efficiency and interference mitigaton. It's a wide open area for experimentation and development.

tim ab0wr
 
RE: Passive Audio Filter  
by WB2WIK on March 30, 2007 Mail this to a friend!
>RE: Passive Audio Filter Reply
by K3AN on March 30, 2007 Mail this to a friend!
"some high frequency rolloff increases readability."
Alan, I am puzzled by that statement. As people get older and their hearing starts to go, it's the high frequencies that are lost first. This results in a loss of "readability" and most hearing aids try to compensate by BOOSTING the highs with respect to the lows.<

::Probably absolutely true for people suffering hearing loss. Not necessarily true for the rest of us. Using the DSP filter options in the ORION-II, I can adjust low and high cut frequencies all over the place and find the *majority* of weak signals under noisy band conditions -- which is the norm on the HF bands -- are much more readable when the high frequency cutoff is selected to be about 2.4 kHz. I can also pick 2.7, 3.3, 3.5 and lots of other frequencies at the turn of a knob; but selecting one of those does not improve readability, it detracts from it because the wideband noise that comes in along with the signal reduces S/N markedly. If I were hearing impaired, I might hear this differently. But since I'm not, there is no question at all that the narrower BW with lower frequency cutoff improves readability for voices on HF SSB circuits which are typically noise laden.

WB2WIK/6
 
RE: Passive Audio Filter  
by K6AER on March 30, 2007 Mail this to a friend!
Looking at the pass band of the audio filter the 20 dB pass band is from 250 Hz to 2800 Hz but the real 3 dB band width is from 500 to 1500 Hz. This I would describe as a terrible audio bandwidth for communication. Being most communication radios are designed for 300-3000K on FM units the intelligibility of frequencies above 1500 Hz are greatly needed. I set my DSP filter on my Pro-III for the bandwidth depicted in the filter sheet and the received audio sounded terrible. If you re into the hand over the hostage mouth audio I guess that is great.

The one other statement that was made,

There are a couple of caveats to keep in mind. First, the audio amplifiers in amateur radio equipment are almost always transformerless and integrated. This usually means that both leads to the speaker are above ground potential.

A quick check of typical equipment and I found the FT-920, FT-2000, TS-570, TS-2000, IC 756-Pro III, FT-100D, IC7800 all had grounded speaker jacks.

What amateur HF gear made in the last 15 years has a floating speaker output?
 
RE: Passive Audio Filter  
by WB2WIK on March 30, 2007 Mail this to a friend!
>RE: Passive Audio Filter Reply
by K6AER on March 30, 2007 Mail this to a friend!
What amateur HF gear made in the last 15 years has a floating speaker output?<

::Funny you should ask, Mike! Ten Tecs do. I don't know why. There's even a notation in the manual to NOT ground either side of the speaker...

73

Steve WB2WIK/6
 
RE: Passive Audio Filter  
by WA1RNE on March 30, 2007 Mail this to a friend!


"Looking at the pass band of the audio filter the 20 dB pass band is from 250 Hz to 2800 Hz but the real 3 dB band width is from 500 to 1500 Hz. This I would describe as a terrible audio bandwidth for communication."


>>> True, but as Alan indicated the filter is geared toward CW, but an SSB version is available. The response shown in the article is usable for CW.

Aside from that, an Active Filter Kit would be much more useful and would likely cost about the same using a 9 volt battery or vehicle as the power source. An optional wall wart would bump the cost up a few bucks....


...WA1RNE
 
RE: Passive Audio Filter  
by W6TH on March 30, 2007 Mail this to a friend!
.

Is it possible that the "IF" shift could resemble this filter? Turning the shift to the left will bring in the lows decreasing the highs and tuning to the right emphasize the highs and attenuate the lows? Also an advantage, will not decrease the audio output.

.:
 
RE: Passive Audio Filter  
by W3JJH on March 30, 2007 Mail this to a friend!
I wonder if the frequency response data shown is measured or calculated.

The input impedance of a loudspeaker in not purely resistive. For a simple driver or a driver in a closed-box enclosure, the input impedance is inductive at low frequencies below the electromechanical resonance and then capacitive above resonance to some middle frequency above which the voice coil inductance becomes a factor. It looks like a resistor at only three points--the voice coil resistance at dc, a higher value (depending on the system Q) at resonance, and a value somewhat greater than the voice coil resistance at that mid frequency above resonance.

In high quality loudspeaker systems it is not uncommon to find conjugate networks in parallel with the drive terminals to provide controlled loads for the passive filters in the crossover networks.

A typical communications loudspeaker would have its resonance near the low frequency corner of the response curve shown, and I'd expect that the interaction would significantly affect the filter performance. Also, the inductance of the voice coil could become an issue affecting the high frequency rolloff.
 
RE: Passive Audio Filter  
by W6TH on March 30, 2007 Mail this to a friend!
.

I imagine the response curve of this filter was done with an 8 ohm resistor or such at the output.

.:
 
Passive Audio Filter  
by KE4ZHN on March 30, 2007 Mail this to a friend!
Interesting article. Ive always found that narrowing the bandwidth of my 756 Pro receiver definitely increases readability on weak signals. Its nice to listen to good quality ssb audio (No, not over eq`ed crap!) with wide receiver bandwith if the signal is strong, but for pulling the weak ones out of the crud, you cant beat cutting the high end off to reduce the atmospheric noise that comes in along with the signal. It may not sound wonderful but you'll copy every word. It seems that many amateurs get confused between fidelity and actual readability. Communication is the name of the game, not trying to sound like an FM radio station on HF. That extra wide crap may sound good to your ego maniac buddies who brag with you every night, but in the distance your signal is very poorly readable. All that mushy bass makes for difficult copy in the noise. But...as the warden said in the movie Cool Hand Luke...some men ya just cant reach...
 
RE: Passive Audio Filter  
by K6AER on March 30, 2007 Mail this to a friend!
Steve,

Youre kidding me? The TenTecs have floating audio? Those boys in Tennessee dont get out much.
 
RE: Passive Audio Filter  
by AB0WR on March 30, 2007 Mail this to a friend!
KE4ZHN:
<<Interesting article. Ive always found that narrowing the bandwidth of my 756 Pro receiver definitely increases readability on weak signals. Its nice to listen to good quality ssb audio (No, not over eq`ed crap!) with wide receiver bandwith if the signal is strong, but for pulling the weak ones out of the crud, you cant beat cutting the high end off to reduce the atmospheric noise that comes in along with the signal. It may not sound wonderful but you'll copy every word. It seems that many amateurs get confused between fidelity and actual readability. Communication is the name of the game, not trying to sound like an FM radio station on HF. That extra wide crap may sound good to your ego maniac buddies who brag with you every night, but in the distance your signal is very poorly readable. All that mushy bass makes for difficult copy in the noise. But...as the warden said in the movie Cool Hand Luke...some men ya just cant reach... >>

If cutting highs down below 2500hz gives the audio more intelligibility then you need to publish a scientific paper refuting all the studies of intelligibility that have been done over the past decade.

Take a look at this site for and explanation of how intelligibility can be scientifically measured in a repeatable fashion.

http://www.isvr.co.uk/spe_hear/comint.htm

If you want to cut something cut the frequencies below 500hz but leave all the highs you can. You get a lot of power below 500hz but most of the intelligibility is above 500hz.

That's why the Articulation Index for 300-5000hz speech is higher at 0db Signal to Noise ratio than 300-2500hz speech at 10db Signal to Noise ratio.

You can quote all the anecdotal stuff you want. I could list a page full of scientific studies over the past decade that show this with repeatable results.

If you really want to snag that rare DX, cut your lows below 500hz and raise your highs as far as you can.

tim ab0wr
 
RE: Passive Audio Filter  
by K3AN on March 30, 2007 Mail this to a friend!
">>> True, but as Alan indicated the filter is geared toward CW, but an SSB version is available. The response shown in the article is usable for CW."

OK, if the filter is intended for CW operation it will provide some benefit. I guess I thought since the article talked about its use in a mobile environment, I assumed phone operation. My bad. However, in re-reading the article I was unable to determine for sure whether the filter is intended for CW or for phone communication.


 
RE: Passive Audio Filter  
by AB0WR on March 31, 2007 Mail this to a friend!
">>> True, but as Alan indicated the filter is geared toward CW, but an SSB version is available. The response shown in the article is usable for CW."

k3an:
<<OK, if the filter is intended for CW operation it will provide some benefit. I guess I thought since the article talked about its use in a mobile environment, I assumed phone operation. My bad. However, in re-reading the article I was unable to determine for sure whether the filter is intended for CW or for phone communication. >>

The reason cutting the highs in mobile work usually helps in the mobile environment is because any highs that are coming out of the speaker usually represent only masking noise, at least for ham radio. The audio chain and IF filters in most ham transmitters today only pass high frequencies up to the range of 2.2khz to 2.5khz. You can't hear something that isn't sent.

I guess the question could be asked: If you have good 2.2khz IF filters then where is the noise in the speaker above 2.2khz coming from? Perhaps the design of the product detector and following audio amps needs to be looked at?

tim ab0wr
 
RE: Passive Audio Filter  
by KE4ZHN on March 31, 2007 Mail this to a friend!
Tim, I'm no scientist or engineer, but I know what I hear. All the studies you point out are fine and I'm not going to argue the fine points. In fact in a lab with audio testing I`m sure what you say is correct. But its a different world from a quiet sound lab to listening to a ton of static and white noise and trying to pull a weak radio signal from it. All I can tell you is it works for me. Any time I hear a real weak signal buried in the noise, Ive narrowed the RX bandwidth to even as narrow as 1.8 khz and managed to pull the signal out of the noise. Yes, the fidelity of the received audio inst wonderful with all the high end chopped off, but most of the snap crackle pop static goes away also. Listening wider makes it more difficult to hear the other station because of the increased noise floor.
 
RE: Passive Audio Filter  
by WA1RNE on March 31, 2007 Mail this to a friend!

Keep in mind, this filter is offered by a web site dedicated to crystal radio affectionato's.

It might be a fun project, but no passive audio filter with a 6dB response at 1 khz off each side of center frequency is going to make up for a marginal receiver design or a station receiver set up in a noisy environment.

If the receiver has decent overload and IMD performance and has a good noise blanker that's all you need - in addition to eliminating the actual noise source when possible and common mode noise suppression.


...WA1RNE
 
RE: Passive Audio Filter  
by AA4PB on March 31, 2007 Mail this to a friend!
On the one hand the highs can improve intelligibility. On the other hand narrowing the bandwidth (which includes cutting the highs) improves the signal to noise ratio. It is quite probable that improving the signal to noise ratio has more effect on intelligibility than the highs do (at least thats been my experience). I'm sure all the studies were done in a low signal to noise environment.
 
RE: Passive Audio Filter  
by AA4PB on March 31, 2007 Mail this to a friend!
Correction: All the studies were done in a high signal to noise environment (i.e. low noise, high signal).
 
RE: Passive Audio Filter  
by WO7T on March 31, 2007 Mail this to a friend!
This is a novel little filter, to commit to the innards of a station or mobile speaker I would think. If you need want a little more, go to: http://newenglandqrp.org/nescaf and spend $29 for the NESCAF kit.... Knobs, pots, and switches included.

72,
Mark
 
RE: Passive Audio Filter  
by AB0WR on April 1, 2007 Mail this to a friend!
ke4zhn:
<<<<Tim, I'm no scientist or engineer, but I know what I hear. All the studies you point out are fine and I'm not going to argue the fine points. In fact in a lab with audio testing I`m sure what you say is correct. But its a different world from a quiet sound lab to listening to a ton of static and white noise and trying to pull a weak radio signal from it. All I can tell you is it works for me. Any time I hear a real weak signal buried in the noise, Ive narrowed the RX bandwidth to even as narrow as 1.8 khz and managed to pull the signal out of the noise. Yes, the fidelity of the received audio inst wonderful with all the high end chopped off, but most of the snap crackle pop static goes away also. Listening wider makes it more difficult to hear the other station because of the increased noise floor. >>>

If that's the case there were no highs there for you to hear. That could be caused by many things. Perhaps the masking noise in your receiver at the time was not a white Gaussian noise that is considered to be so typical of the HF bands that it has been given its own name AWGN (Additive White Gaussian Noise) and is used in many, many HF communication channel studies. What is the noise figure of your receiver? I'm not even sure the ARRL even measures that in their evaluations. But on a weak signal even a 2-3db noise figure can add enough masking noise to obliterate the weaker, high frequency parts of a signal. Remember that on an unprocessed audio signal, the higher frequencies can be as much as 26db below the low frequencies.

I'll reiterate. A 200watt signal with a 3db pre-emphasis on the speech power above 500hz on a 250-5000hz bandwidth has the same intelligibility as a 1000watt signal with no processing and a bandwidth of 250-2500hz. This can be directly calculated using the integral of the voice spectrum and of a Gaussian noise source. Now, I'll add the caveat that this is based on a voice that follows the mathematical model that Sabin and Schoenike used - but that model has been widely used in the literature on hearing studies and, to my knowledge, no peer review of any study using that model has ever questioned its validity.

The mere fact that people like you legitimately point out that narrower bandwidths *help* intelligibility for you tell me that there is lots of room for experimentation and study of why our ham receivers work in such a manner when the scientific world says it just shouldn't be that way.

If you want confirmation of that conclusion take a look at the sales of the MFJ Speech Enhancer which does nothing more than add pre-emphasis to the higher audio frequencies *after* they have been received. It would be ever so much more efficient to do that ahead of time. Their sales do confirm, however, that the upper frequencies do add significantly to the intelligibility of speech.

tim ab0wr
 
RE: Passive Audio Filter  
by AB0WR on April 1, 2007 Mail this to a friend!
AA4PB:
<<<On the one hand the highs can improve intelligibility. On the other hand narrowing the bandwidth (which includes cutting the highs) improves the signal to noise ratio. It is quite probable that improving the signal to noise ratio has more effect on intelligibility than the highs do (at least thats been my experience). I'm sure all the studies were done in a low signal to noise environment. >>

AA4PB:
<<Correction: All the studies were done in a high signal to noise environment (i.e. low noise, high signal). >>

Nope. If you read closely I said that Sabin and Schoenike showed that a 250-5000hz audio signal with 3db preemphasis above 500hz and a SNR of --- 0db --- has the same intelligibility as an unprocessed 250-2500hz audio signal with a --- 10db --- SNR.

A 0db SNR is not a low noise environment. Even if you consider it to be such, consider what happens to the narrow band audio when you drop it to 0db SNR.

tim ab0wr

 
RE: Passive Audio Filter  
by N3JI on April 2, 2007 Mail this to a friend!
I was a pretty heavy contributor to the other site's thread Tim-AB0WR mentioned earlier. I'll keep it short here. I have been experimenting with different types of Amateur Radio audio for several years now, and the following transmit audio specs seem to work very well for me when trying to bust a pileup or grab that DX contact: ~400-3400 (3 kHz total bandpass) with some fairly heavy audio compression, and a lot of boost in the 2200-3400 range. This was done essentially by trial and error before I read too much about the subject. The cool part is that I basically reproduced the results of the studies done that Tim mentioned with one exception. Since most Amateurs don't bother to listen wider the 3k (many listen at 2.5 or even 1.8 as someone else mentioned), going out past 3400 probably doesn't gain you much. Another thing to keep in mind is when the going gets tough, a lot of folks (me included) shift that IF up. I believe this is where the extra few hundred Hz helps. If Amatuer "A" is running a 2.4k TX filter, he might get audio to 2800 and that's it. I'm running out to 3.4k, so when Amatuer "B" who hears both of us shifts his IF up, I have a lot of energy up there where his passband is now going and I'll sound much clearer than Amatuer "A" even though we're both cutting the bottom end around the same place.

That about sums up in the "Cliff's Notes" version of what I have found in the last few years. Having now read a lot of the studies out there, and considering how the average Amateur listens (especially during pileups, contests, etc.), it's easy to see how that slightly extra bit of frequency response can be used to your advantage. Of course, the trick is finding a rig that has that extended transmit response capability. Unfortunately, the latest Icoms cut us at 2900 (and it's brick-walled in the DSP), but several later models of Kenwood have it available (I run the TS-950SDX with about 4k max TX bandwidth), and the newer Ten-Tecs (Jupiter, Orions, Omni VII) can go out to about 3800. The new Yaesus (FT9000, FT2000) can go a bit wider as well, but I'm not sure how far.

I said I'd keep this short, so that's enough for me. Other than that, everyone needs to take another look at what the ARRL is trying to get into the FCC again. This time, they're trying to take 10m & up away from us in regards to this type of SSB experimentation. Of course the big things are the robots & digital formats they're trying to push on us, so pay close attention and get those comments in!!

73,
Joe, N3JI
 
RE: Passive Audio Filter  
by N1XBP on April 2, 2007 Mail this to a friend!
This would be a fun kit to couple with a small amplifier stage to make it zero gain when switched in and out.. I'd imagine it would make a fun SSB accompaniment to my NESCAF for CW.

The only thing missing then would be a variable analog notch, but I haven't figured out how to make a notch sharp enough to be useful without using DSP or introducing a lot of ripple.

I don't know about everyone else, but I'll take analog ringing over the odd digital sounding distortion DSP creates sometimes any day. I often find that more distracting and fatiguing than the noise that is being filtered.
 
Passive Audio Filter  
by KG6OMK on April 2, 2007 Mail this to a friend!

I'd like to know what really makes a voice more intelligible.

Doesn't the Heil HC5 mic cartridge give a _boost_ to the higher voice frequencies so as to make you more intelligible. Sounds un-natural but they say it is easy to copy.

What filter is best may be different depending on where you place it in the chain. In the mic, in the xmitter's AF or in the receiver before or after the audio amp. Could be the answer is different depending on where you filter.

At first when I saw this article, I figured a low pass would be the worst thing you could do to improve voice intelligibility. but then I thought about it more. I'm thinking now that the SSB transmitter has already cut off everything above about 2.8Khz so what is the low pass inside the speaker doing? Likely removing "garbage" that got put into the signal path someplace between the transmitter's 2.8Khz cutoff filter and the receiver's speaker. Basically it is removing "stuff" over then voice. Seems now that the best filter to put on a speaker would be a bandpass that matches the audio bandpass of the transmitter. So I can see why this would work
 
RE: Passive Audio Filter  
by K0BG on April 3, 2007 Mail this to a friend!
A few people missed an important point. The filter comes in two different configurations. One for CW, the other for SSB. I used the pics off of their web site, which is of the CW version. The response curve is the one for the SSB filter.

It turns out that all of the values for the three combined networks are all the same. It is a clever design as you don't have any "special" values.

I agree that the dynamics of the speaker effect the response curve. How significantly depends on too many factors to discuss here. However, the speaker impedance doesn't appear to be too critical; most mobile speakers are about 8 ohms anyway.

Steve, WB2WIK, hit the nail on the head. Cutting the highs in a noisy band environment is the key here. Extended bandwidth intelligibility notwithstanding, those who think this "trick" doesn't work, have never operated mobile.

Alan, K0BG
www.k0bg.com
 
RE: Passive Audio Filter  
by WB2WIK on April 3, 2007 Mail this to a friend!
>RE: Passive Audio Filter Reply
by K0BG on April 3, 2007 Mail this to a friend!
Steve, WB2WIK, hit the nail on the head. Cutting the highs in a noisy band environment is the key here. Extended bandwidth intelligibility notwithstanding, those who think this "trick" doesn't work, have never operated mobile.<

::Agreed, it does. Other tricks that help is bringing the sound much closer to your ears so extraneous noises coming from other sources don't overwhelm the signal you're trying to hear. A high-quality "communications" speaker -- one with limited range to pass mostly 300-3000 Hz and roll off below and above those points -- at head level aimed right at the operator helps enormously, for me!

-WB2WIK/6
 
Email Subscription
You are not subscribed to discussions on this article.

Subscribe!
My Subscriptions
Subscriptions Help

Related News & Articles
JT65-HF -- an 'Odd' but Fun Digital Mode
The 'Wadetenna' for HF Pedestrian Mobile


Other How To Articles
Possible Cure for Relay Contact Cancer?