Speaker filters.. Reduce the hiss

[ke7trp]

I replaced the cracked Jensen 10 in my R46B with a new Jensen MOD speaker. They cost about $35.  The sound was great but the high end of the new speaker was way up there.. After a few days the high end Hiss really bothered me.

We played with different caps across the speaker terminals.  A 10 to 12 UF cap takes the hiss right out. A 22 makes it sound nice and mellow. almost no anoying his.. I dont hear much if any loss of fidelity while listening to BC stations or ham stations. Its now warm and smooth.

C

[WBear2GCR]

Not sure what a Jensen MOD speaker is... but...

Assuming it is a single driver:

I'd do two things before I put a cap across the speaker terminals.
A cap across the speaker terminals looks like a dead short at some high frequency, maybe within the audio band.
Not good for the amp that feeds it.

An inductor in series will only raise the impedance... not much of a problem.
A resistor in series with a shunting cap will limit the rolloff and shelve it (usually good enough) and still prevent
the impedance from getting too low.

An inductor in series and a cap shunting will give you a 12dB/oct rolloff - but again you can put a resistor in series with the cap and across the inductor to moderate that rolloff if you wanted to.

There are various "crossover simulators" and calculators online now that you can plug values into and see what the electrical part of the crossover will do - it won't show the combined acoustic + electrical response though. You'll have to use ur ears, or else measure that.

But, the easiest and safest way to go is the venerable acoustic roll off method.
Sounds good when you say it that way.
That means that you put a bit of material of the proper thickness and type (cotton, felt, wool, etc) between the speaker and you. Usually behind the grille cloth! Thicker & denser material will roll off lower in frequency. Quite nicely. A bit of experimenting will get it just right... try draping a bath towel over it for starters... 

In the recording biz, the technique often involves scotch tape and some tissues. Thus the "one, two or three tissue tweeter".
Of course with the tissues the idea is to get a rising response above ~10khz to be relatively flat...

IF it is a two or three way (two or three driver) speaker then the methods to use are similar but a bit different - you can simply pad down the tweeter and/or alter some values in the crossover to suit... or at least to start.

                      _-_-bear

[ke7trp]

Hmm. Never thought of the cap loading down the audio amp. This is how the filters in most ham speakers are made. Just a cap across the terminals inside.  The Hallicrafters used caps in some of the big speakers. I studied some schematics.  It sounds just wonderfull.  All the very high end hiss is gone. But none of the Freqs in a persons voice is reduced at 22 UF. 

The Jensen MOD is a replacement speaker.  visit Jensentone.com  They sell different speakers marketed towards guitarists. The original was a jensen Alnico.  They make that same speaker but its $90.  I decided the $30 MOD would be fine and it is.. It really sounds fantastic.  Lots of low end. My factory speaker had a tear in it from coil to mount.

C

[WD8BIL]

Antique Electronics Supply has the Mod 10 for $35.00.

[flintstone mop]

During my TV servicing days that was common to see that done to alter the response of audio for whatever characteristics the manufacturer was trying to achieve.

Fred

[The Slab Bacon]

Many receivers in the past put a cap across the primary side of the outpoot transformer to accomplish the same effect. Usually somewhere between .001 and .01. Also sometimes seriesing a pot with the cap for a somewhat rudimentary "tone" control. This was very common in "antique" radios.

so putting a larger cap scross the speaker would give the same effect. the larger cap is necessary due to the much lower impedance of the outpoot side of the transformer.

                                                    the Slab Bacon

[KM1H]

The Hallicrafters R-42 speaker has a front Hi-Fi/Communications switch with just a capacitor and it does an excellent job here knocking down hiss and static. The later R-48 & R48A speakers also have the switch. For a steeper rolloff a simple LPF can be built using one of the on line calculators.

Carl
KM1H

[K3ZS]

Many receivers in the past put a cap across the primary side of the outpoot transformer to accomplish the same effect. Usually somewhere between .001 and .01.

                                                    the Slab Bacon

My Multi-Elmac PMR-7 receiver had a cap across the primary.    I clipped it and AM sounds much better on the speaker I am using.

[ke7trp]

Yeah.. The cap wont hurt or load anything.  We just tore down some Famous model huge klipsh speakers that had Caps right on the Crossover board.  The JBLs I worked on also had big paper caps in the network.

I spent a few hours listening.  Don and some other guys where comming in BIG late last night.  On Don,the 22 cap has little effect on his tone. The HiSS is GONE.  The tone is mellow.  One of the guys he was talking to was hifi and had a wide range of audio.  The 22 did knock down his high end quite a bit.  A 10, 12, or 14 UF is what you are looking for to knock JUST the hiss and noise out but still have full hifi audio.

Without the cap, I get listening fatigue and just want to turn the damn thing off after an hour.. ITs playing some freq that just bothers me. With the cap, I can listen for a long time without being bothered.


C

[WBear2GCR]

yeah, caps on the xover board - but in what position??

In a SERIES xover the cap appears across the woofer, but in series with the tweeter, and the inductor appears
across the tweeter but in series with the tweeter.

In a PARALLEL xover there would be a cap in series with the tweeter and an inductor in series with the woofer.

That would be for a first order filter.

If you take the PARALLEL xover configuration and add a cap across the woofer, and after the inductor, not before(!)[/i] then the woofer now has a 2nd order or 12dB/octave slope.

The reason that the cap is not before the inductor is not because that changes the filter's frequency response (much), but because it dramatically changes the impedance the amplifier looks at.

The fact that it is a quick and dirty solution and has been used by various manufacturers to roll of woofers, doesn't make it good practice, or a good solution.  I suppose with flea power amps its not much of a big problem, but some solid state amps will become rather unstable when faced with a mostly capacitive load...

                            _-_-

[ke7trp]

Klipsch and JBL disagree.  Both of them have caps right across the midrange.   So does the Icom and yaesu speakers with filters built in. There are probably 10 ways to do this

C

[k4kyv]

Sounds like either there is an acoustical resonance in the speaker at the nuisance frequency, or a spurious peak in the response of the audio amp. If you have a smooth, wide frequency response, you should not have this problem.  From my experience, putting caps across something to limit high frequency response also kills the sibilance response and results in loss of voice articulation frequencies and thus loss of intelligibility. My outboard receiver audio amp (a 50's vintage 10w monaural "hi-fi" amplifier) has bass and treble controls.  I keep the bass control at flat, but adjust the treble from time to time for the best compromise. The Sherwood synch detector has a 3-position treble boost/flat/cut control.

[Steve - WB3HUZ]

For sure Don. If the IF bandwidth of the RX is correct, any sort of lowpass filtering in the audio chain is useless or will only further reduce the high frequency response. Most ham radios use such techniques because either the last IF is very noisy or the audio amplifier is very noisy.

[ke7trp]

Yes, All my receivers are broken and or of poor quality.

R390
R390a
SP600 JX14
Sp600 JX21
SX62
RME 6100
SX100
SX101
SX110
RME42
75A1
75A2
74A4


I better sell them off or just throw them out and get something else.. They all hear band noise and hiss!

Clark

[Steve - WB3HUZ]

Only because the IF bandwidth is too wide. It's receiver agnostic. Drama not needed.

[ke7trp]

I can adjust the bandwidth.. Come on..   The 10UF cap kills the hiss.  Does not effect anything in the human voice as recieved.

C

[The Slab Bacon]

Having somewhere around 15 receivers in my stable (give or take a few) I guess it's time for me to chime in again.

I have to agree with Clark on this one. Sometimes the high frequency band "hiss" and ambient QRN / QRM can be a bit fatiguing on the ears. On many occasions I run the '390A in the 4kc bandwidth to eliminate some of the background noise.

But this brings one particular receiver that I have to mind. It is an old WW2 military RBS receiver. It has very interesting audio filtering / shaping that eliminates much if not all of the high frequency component from the audio outpoot. This makes it very quiet between tuned signals as you hear no "front end noise" at all, or at least very little. So little that you sometimes wonder whether or not it crapped out. Yet it has little effect on the audio in the normal voice range and does not sound at all "muddy". If anything it actually improves the percieved signal to noise ratio. It is a very pleasant radio to listen to and "easy on the ears", especially on a noisy night. It also creates a very "quieting" effect even from the weaker signals that wouldnt normally quiet out the receiver. Everyone that has heard it comments on how nice it is to listen to.

However, the audio filter / shaping network is a lot more than just a cap across the outpoot. It is quite involved with a series of caps and chokes before the outpoot tubes.
I had thought about bypassing them out for more HiFi audio (it has P-P 6V6s),
but it would have detracted from the personality and charm of THAT receiver, so I left it stock. I have enough HiFi receivers. It seems to have just the right audio passband for good voice reproduction with minimal / no background noise. It may be a little lacking in syballance (for some) but no one that has heard it has had anything bad to say about it.

I guess they figured that tightening up the audio bandwidth was easier (or cheaper) than tightening up the IF bandwidth when they designed it.

                                                    The Slab Bacon

[The Slab Bacon]

Oh Yeah, and, err, furthermore....................

The caps that they put across the primaries of the outpoot transformers of some of the "antique" radios did wonders to improve what came out of the speakers.

[ke7trp]

FB bacon. 

I would like to make a filter box.  With speaker terminals in and out.  This will give me different levels of sound shaping.  It really makes the operating experience much more pleasurable to my ears!  Hearing that hiss and static all night really wears you down.

C

[KM1H]

Back in the 60's there was an article about improving voice communications intelligibility and reducing operator fatigue by filtering out high frequency hiss. Now, if I remember this was from NASA, USAF, or a similar organization so I have a tendency to believe them.

I do know that it worked for me on a 75A4 and has been installed in its headphone circuit ever since.

For the TS-940's and TS-830's used with transverters I use headphones with a sharp cutoff.

The article was reprinted in either CQ or HRM (Im somewhat sure   )if someone has time to do the searching.

Carl
KM1H

[Steve - WB3HUZ]

What's the difference if the hiss is removed at the IF or AF level? If I have an 8 kHz IF bandwidth, I should not need a 4 kHz lowpass filter in the audio chain. If you do, then the IF or AF amps in your receiver are very noisy.

Good receiver design dictates narrowing the bandwidth as early in the receiver chain as possible. Using audio filters is a band-aid approach for a weak, noisy or too wide IF. This is not something I just made up. The concept has been around for many decades.

So, think what you will, but you aren't arguing with me. You are trying to refute engineering concepts that have been widely known and accepted before we were born.

Oh Yeah, and, err, furthermore....................

The caps that they put across the primaries of the outpoot transformers of some of the "antique" radios did wonders to improve what came out of the speakers.

[ke7trp]

I am not arguing.  You dont understand what I am talking about.  This cap is not limiting bandwidth at all.  I think you cant get past that. I just played 20 to 20K through the reciever. I can hear clearly 80 to 10K.  Without the cap, 80 to 10K. The cap is removing very high Freq noise and hiss that has NOTHING to do with what is comming out of your transmitter.  There is no bandwidth limiting going on. Just no damn anoying HISS and band noise.

Do you understand now  

C

[Steve - WB3HUZ]

No, I don't understand. Please explain how the cap knows the difference between hiss at some frequency and voice at the same frequency.

[K5UJ]

Yes, All my receivers are broken and or of poor quality.

...
75A1
75A2
74A4

I better sell them off or just throw them out and get something else.. They all hear band noise and hiss!

Much to my relief the 75A3 isn't in there.  It must be noiseless 

This is a timely topic because I have been searching for a nice big speaker in a box for my recently arrived A3, and having found nothing that appears to be what I had in mind, have begun to seriously contemplate HB.  A friend is a wood worker hobbyist and I'm thinking a dark stained mahogany looking cabinet around 20" high and wide and maybe 12 or 18" deep.  Dark gray grill with silver thread.   Get one of those Jensen speakers, 25 watt 4 ohm 15 inch diam. however I think that might be too much of a load for the little 2 1/2 watt AF output stage in the A3.  Maybe 10 watts is better.  Anyone here ever built their own speaker cabinet?

rob

[ke7trp]

The 12 or 15 will work great. Look at what QIX uses.. A huge speaker.

I would like an R12 hallicrafters Style speaker with a 12 inch jensen.  We are going to build a panel for the rack that holds a 12.  These speakers are realy effiecient.  Go to jensentone.com and read the charts and specs. Most are 95 to 96 DB for 1 watt input. 2 watts would be loud.

C