D-104 MODIFICATION

[k4kyv]

Does anyone know if this modification keeps the stock D-104 crystal element?
http://home.comcast.net/~esprepair/d104.htm

I find the 3 kHz "hump" in the stock D-104 frequency response helpful.  It adds crispiness and clarity, without creating a "tin can" sound.  I even add about a few more dB's of boost with pre-emphasis in my speech amp. I find that a flat response leaves a dull sound over the air, especially since so many AM receivers (ham and broadcast alike - even the newer ones) have audio response that begins to roll off below 2000~.  The upper midrange boost compensates for this poor receiver frequency response, and brings the articulation up out of the background noise.  A better solution would be to add some rolloff at the receiver end if you find the sound too harsh.

But the low frequency response in the sound file sounds better than anything I have ever been able to achieve with a stock D-104.

BTW, to me a "stock" D-104 means the original unamplified version.  The "power mike" amplifier that Astatic started putting in the base of the microphone to appeal to the CB crowd is a piece of crap, with limited frequency response and plenty of distortion.  The first thing I would do would be to rip it out and connect the shielded mic cord directly to the crystal element.  Of course, that would be a problem with a rig with low impedance mic input.  I use a  homebrew mic pre-amp, with audio input directly to the grid of the 12AX7, and a 5-meg grid leak resistor to load  the mic.

I did try using a tube-type Altec preamplifier, which has separate bass and treble controls.  You can get plenty of bass boost, and it sounds good in headphones, but the problem is that the bass boost reverses the phase of the audio at the lower frequencies, so that the waveform is asymmetrical in the opposite direction than it is at higher frequencies.  The result was that I had to sacrifice at least 10 dB of average audio power in the sidebands in order to avoid overmodulating in the negative direction.

Before investing in the custom modification, I would advise checking the waveform symmetry of a modified microphone, throughout the frequency range, to make sure this modification doesn't do the same thing.

The best I could ever do with the D-104, to add good bass response that maintains the proper phase of the asymmetry, was to mix in some bass from a separate microphone.

[W2XR]

Hi Don,

Your comment as to boosting the bass reverses the phase of the audio at the lower frequencies is interesting.

Can you clarify for me as to why this phenomenon occurs?

Thanks & 73,

Bruce

[Todd, KA1KAQ]

Does anyone know if this modification keeps the stock D-104 crystal element?
http://home.comcast.net/~esprepair/d104.htm

Good question. It sounds like the same mic with bass added.

BTW, to me a "stock" D-104 means the original unamplified version.  The "power mike" amplifier that Astatic started putting in the base of the microphone to appeal to the CB crowd is a piece of crap, with limited frequency response and plenty of distortion.  The first thing I would do would be to rip it out and connect the shielded mic cord directly to the crystal element.  Of course, that would be a problem with a rig with low impedance mic input.  I use a  homebrew mic pre-amp, with audio input directly to the grid of the 12AX7, and a 5-meg grid leak resistor to load  the mic.

That's what I did with the KW-1. In fact, when I had the audio deck out in February for recapping and repairs, I checked the resistor value. 4.7 megs, which was what I thought but couldn't remember.

Agreed on the 'power mic' D-104s, I've actually heard that the newer or later crystal heads (70s or 80s on, I think) don't sound as nice as the older crystal heads. I think I have one 'newer' type from the mid 70s, the rest are older. The one I use on the air came with the transmitter when I got it in 1988, and it was old then. Still seems to do the trick.

Stu, AB2EZ has played with his D-104 a bit recently and posted some info about his tests. Perhaps he'll chime in.

The best I could ever do with the D-104, to add good bass response that maintains the proper phase of the asymmetry, was to mix in some bass from a separate microphone.

It always sounds good up here, Don. My set up lacks bass as well, but I don't have a whole lot in my voice either. At some point I'd like to try experimenting with different mics, but time is always an issue.

At the end of the day, the D-104 seems to be a good baseline to go by, you can add or subtract from there.

[KF1Z]

Hi Don,

Your comment as to boosting the bass reverses the phase of the audio at the lower frequencies is interesting.

Can you clarify for me as to why this phenomenon occurs?

Thanks & 73,

Bruce

I can't explain it well ....

But this thread may help....

http://amfone.net/Amforum/index.php?topic=9129.0


Bruce

[Steve - WB3HUZ]

From the ESP web site:

However, if you are working SSB, or if you are talking AM or SSB on a professional HAM radio, then you want a FLAT response.

What is a professional HAM radio (and why must the word ham be shouted)? Too funny.

The part about wanting a flat response is just plain wrong, as Don pointed out very clearly below.

The sound files are rather meaningless too. Since amateur radio use will not be a channel with wide frequency response and a high SNR. Let's hear how the mics sound through 3-4 kHz channel with a 30 dB SNR.

[Todd, KA1KAQ]

From the ESP web site:

However, if you are working SSB, or if you are talking AM or SSB on a professional HAM radio, then you want a FLAT response.

What is a professional HAM radio (and why must the word ham be shouted)? Too funny.

Well, they couldn't very easily call it a 'professional amateur radio', now could they? That would confuse even the toughest good buddy.

Yelling is a good TEST of yer wall to wall audio. The purpose is to see how high you can make the needle on the gauge go. Obviously these modified D-10-4s (the CB designation) have the gauge booster built in.

[Steve - WB3HUZ]

It's probably more complex than this but, almost any eq or RC-based frequency shaping circuit is going to have some phase shift associated with it. By boosting certain frequencies, and not others, the phase relationship between the two frequency ranges will be altered.

Hi Don,

Your comment as to boosting the bass reverses the phase of the audio at the lower frequencies is interesting.

Can you clarify for me as to why this phenomenon occurs?

Thanks & 73,

Bruce

[AB2EZ]

Hi!

With respect to the effect of equalization on the asymmetry of the audio waveform... i.e. the peak of the waveform in one direction (e.g. positive-going) is larger than the peak of the waveform in the other direction:

This is easiest to understand if you think of the special case of a periodic signal that has no DC value (e.g., it has passed through an AC-coupled amplfier). For example: a series of identical pulses, each pulse having some shape, that repeat themselves every millisecond. Any periodic signal is made up of the sum of a DC component and a bunch of sine waves; where each of these sine waves is at a frequency which is a harmonic of the repetition frequency of the periodic signal. Since this periodic signal is assumed to have no DC value, and its repetition frequency is 1/(1 millisecond), it is therefore the sum of a bunch of sine waves, at multiples of 1 kHz, having various amplitudes and phases.

Now, each sine wave that makes up a part of the above periodic signal is symmetrical. The asymmetry of the aggregate signal, which is the sum of all of these sine waves, is a result of their phase relationships to each other (and also their relative amplitudes).

If you pass an asymmetrical signal through an filter that changes the phase relationships of the various frequency components, you will change the asymmetry (even if the filter is an all-pass filter that introduces no relative amplitude changes among the various frequency components). A filter (including an all-pass filter) can make an asymmetrical signal symmetrical; it can make a symmetrical signal asymmetrical; and it can also reverse the "phase" of a asymmetrical signal.

Best regards
Stu

[Bacon, WA3WDR]

On this board in the upper left, there is a selection of tools.  Click on 'Online Calculators' and then scroll down to 'Fourier Synthesis' and you can make waveforms out of a fundamental and its harmonics.   The sin and cosine sliders for each are side by side, and you adjust phase by sliding them in various combinations.

By combining the fundamental and the second harmonic in various phase and amplitude combinations, you can produce various composite waveforms with various asymmetries.  By extending this to higher harmonics, you can produce a waveform that is more like a real voice.

As a beginning exercise, leave everything at 0, except set A1 cosine (left side) to +4, and set A2 cosine (left side) to -2.  This adds a fundamental and its second harmonic in a particular phase relationship.  The average is zero, and you can see the asymmetry of the resulting waveform.

[W2XR]

Thank you, Bacon, Stu, Don, and Steve for all of this great supporting information relative to my inquiry.

Bacon, as suggested, I will take the Fourier Synthesis calculator through a few iterations relative to what I see with my rig at and below 50 Hz. Should be interesting. I little cut (3 dB) in the octave beween 40 and 80 Hz did the trick originally, with regard to some asymetry issues I was experiencing in that part of the spectrum.

Much oblige!

Best 73,

Bruce

[k4kyv]

The most phase distortion free equalisation circuit I have ever come up with was suggested by George W2WLR.  Use a triode amplifier like the 12AX7, with cathode bias.  Use a low value of cathode by-pass capacitance, so that at low frequencies there is natural degeneration from the unbypassed cathode resistor.  At higher frequencies, the low value of cathode bypass capacitor becomes effective in by-passing the cathode resistor and thus blocks the degeneration, bringing up the gain of the amplifier.  The low frequency response still extends to the limits set by the coupling circuits, but once the cathode bypass kicks in, at a frequency determined by the time constant of cathode resistor and bypass capacitor, the gain increases until a frequency is reached where the cathode is completely bypassed, and then the gain once again levels off and remains flat out to the limit of the coupling circuitry in the amplifier.  With two stages of a 12AX7, the difference between unbypassed and bypassed cathode resistor is about 10 dB.

[WBear2GCR]

Afaik, anything at all that is not flat DC to light introduces "phase shift" somewhere.
Similarly, anything that provides a "boost", "cut" or "shelf", or even an "extension" to the frequency response will introduce "phase shift".

Fwiw, a 2nd order filter, or 12dB/oct filter (slope) causes the pass band to be shifted 180 degrees out from the stop band. Along the slope the phase is shifting...

The only response modifying circuit that might not introduce phase shift, if programmed to do so would be a DSP filter, afaik.

          _-_-WBear2GCR

[Bacon, WA3WDR]

I don't think that we hear the phase shift.  I think we hear the frequency response effect.  The phase shift affects asymmetry, though, and this can result in a change in modulation level, or overmodulation and distortion.  It affects dynamics if we have fast peak limiting, too.

I have noticed that AM audio can sound muddy if it is out of phase (big peaks modulating downward), even if it did not appear to be overmodulating.  I think this may have to do with detector nonlinearity at low signal levels.  I can't say that I ever noticed speaker polarization except on my own voice, and that was because of direct vs. speaker sound combination.

[N3DRB The Derb]

hey bacon,

remind me of the story behind yer avatar. That's the work of Crumb, right?

[Bacon, WA3WDR]

Hi, Derb!

Yes - that's R. Crumb's Mr. Natural, from a cartoon of him whistling.  I cleaned up a beat-up jpg copy I found on the web, shrunk it down to avatar size, gave it a black border and made it into a 'transparent' gif so it sort of pops out of the page.

Mr. Natural was a feature of Zap Comics from the 60s, the work of madman cartoonist R. Crumb.  Mr. Natural was a cranky old anachronistic wizard in one phase, and a perverted old lunatic hippie guru in others.  It seemed like just the avatar for me...
Google search on Zap Comix
Google search on R. Crumb

On QRZ.com I use a similar avatar with Mr. Natural on a little foot-powered scooter.  That one is from a cartoon where a wide-eyed Flakey Foont (another R. Crumb character) is freaked out and asking, "Mr. Natural... what does it all MEAN?" To which Mr. Natural replied, "Don't mean (expletive deleted)."