triangle waves and ribbon microphones - QST July 2014

[Opcom]

I am gratified to see the use of triangle waves to check linearity being written about. After having used this technique for decades to test audio amplifiers, modulators, and other linear amplifiers but not seeing very many other people use it themselves, it's good to see it get some exposure. It's so simple, inexpensive, and attractive, and is very helpful for those without spectrum analyzers, although not a perfect replacement for them. I've found it useful for getting things to the point where the spectrum analyzer is needed. It will also show up nonlinearities in a way that spectrum analyzers can not, or in a way that is more familiar to those working with linear amplifiers.

Comments:
The author shows in figure three that a 2% 2nd harmonic distortion can be seen by eye as curves in the triangle. An additional level of granularity can be had by comparing the input triangle to the output envelope or waveforms. With the generator's triangle superimposed upon the envelope, much smaller defects are visible by comparison of the waves.

The author has written some nice PC software to generate a 1500Hz tone modulated by a 30Hz triangle. This is OK for RF amplifier checking, but for audio amplifiers and modulators, the experimenter is better off with a signal generator capable of making a triangle wave at various frequencies. That is because nonlinearities can vary with frequency.

In any case I hope that more hams will be encouraged to check their RF amplifier's linearity, especially on SSB where it is very normal to use linear amplifiers and unfortunately common to operate them in a less than linear manner. It would also be useful for those using other analog modes like SSTV.

--

The other article in the issue shows how to build a ribbon mike. That subject hasn't been touched on in a long time and I'm glad to see articles on making basic things at home. There are three other older articles on QST about making ribbon mikes but they are called velocity mikes. FEB 1933 (two), and MAR 1938. The new article takes advantage of modern magnets, but the older ones may offer additional insight into the ribbon itself.


I have a couple of questions:
What is the difference between "ribbon mike" and "velocity mike"?
What is the advantage of using two wires from the top terminal of the mike in the July 2014 microphone article? It was not explained or I missed it.

[AB2EZ]

Opcom

Re: ribbon microphones / homemade ribbon microphones

See also: http://mysite.verizon.net/sdp2/id16.html

In a ribbon microphone, the output voltage is produced when the total magnetic flux passing through the area enclosed by the electrical circuit... consisting of the combination of: the ribbon, the wires connecting the ribbon to the transformer, and the winding of the transformer... varies with time.

The movement of the ribbon increases and decreases the area enclosed by this electrical circuit... causing the total magnetic flux passing through that area to increase and decrease.

One can show (with a little bit of physics) that the amplitude of the signal produced, when the ribbon moves, is proportional to the velocity (speed) with which the ribbon moves in the direction that increases or decreases the area enclosed by the electrical circuit... through which the fixed magnetic field is passing.

One can also show (with a little calculus and a little physics... like F=ma) that the velocity of the ribbon will be inversely proportional to the audio frequency of a sound wave that impinges on the ribbon on one of its two flat sides. Therefore, the response of a simple ribbon microphone will fall off as 1/f... where f is the audio frequency of the sound wave that is impinging on the ribbon.

[Note, the comment in the QST article regarding why this is called a velocity microphone is, in my opinion, not correct]

To flatten out the frequency response, one can place the ribbon assembly in an acoustic cavity, that allows the incoming sound wave to strike one flat side of the ribbon, and a reflected sound wave to strike the other side of the ribbon. If the delay (at the speed of sound) between the incoming and the reflected sound waves is adjusted properly (by the design of the acoustic cavity)... the net result is that the amplitude of the velocity (v. time) of the ribbon will be about the same regardless of the frequency of the incoming sound wave. This is what RCA did in their ribbon microphones.

Time varying magnetic fields from nearby transformers and 60Hz wiring will pass through the cross sectional area of the ribbon circuit, and add interference to the output of the ribbon microphone... resulting in a large amount of superimposed hum. The extra wiring forms a second loop... through which these background magnetic fields pass... but, whose enclosed area is not affected by the movement of the ribbon. The hum in this second loop will cancel the hum in the ribbon loop, if the two loops are connected out of phase.  This is called a "hum-bucking" configuration.

Stu

[Opcom]

Thank you for that explanation. It is very clear how this works and why the extra wire on the top end. Considering the same definition of velocity applies to a moving coil in a magnetic field, ribbon seems clear.

[W3NE]

An excellent video clip demonstrating construction of a ribbon mike, essentially identical (but in greater detail) to the procedure described in the July QST article, has been on YouTube for many months. 

http://www.youtube.com/watch?v=_tWbwcS_9sA

Bob - NE

[W3GMS]

An excellent video clip demonstrating construction of a ribbon mike, essentially identical (but in greater detail) to the procedure described in the July QST article, has been on YouTube for many months. 

http://www.youtube.com/watch?v=_tWbwcS_9sA

Bob - NE

Really enjoyed watching that video Bob.  Its the first time I have seen a ribbon microphone being built with that amount of detail.    Some nice craftsmanship but at the same time very doable for the average builder.  The only Ribbon Mic I have is the RCA 77 and it still works well. 

Joe - GMS

[AB2EZ]

Opcom
et al.
As a refinement to my post, above

In the QST design, each of the two loops includes the ribbon.

If we call one flat side of the ribbon "the front", and if we call the reverse side "the back", then:

One of the loops includes a wire that passes in front of the ribbon. The other loop includes a wire that passes behind the ribbon.

The ribbon moves forward and backward, the area enclosed by one of the loops gets smaller and larger, while the area enclosed by the other loop gets larger and smaller.

As a result, both loops produce voltages as a result of the movement of the ribbon (proportional to the velocity of the ribbon). Perhaps somewhat non-intuitively... the two voltages are of the same polarity (in phase, if they are sinusoids).

Therefore, the wires (one from the top of the front loop, and one from the top of the back loop) are attached to the same side of the step up transformer winding.

Meanwhile, the hums produced in each of the loops, by time varying magnetic fields in the vicinity of the microphone... again, perhaps somewhat non intuitively...  are of opposite polarities; and will cancel (approximately) if the two loops are mirror images of each other (one passing in front of the ribbon, and the other passing behind the ribbon)

Again, this is often referred to as a "hum bucking" design.

If you make a ribbon microphone, it is important to minimize the area enclosed by each loop (to minimize hum produced by time varying magnetic fields in the vicinity of the microphone). I.e. keep the wire that passes in front of the ribbon and the wire that passes behind the ribbon close to the ribbon (less than 0.5 inches away from the ribbon); and twist the three wires that extend below the ribbon, on their way to the transformer.

Also note that you can expect the frequency response of the microphone to follow a 1/f characteristic (much higher low frequency gain)... so be prepared to use appropriate electronic equalization after the high gain (70dB) preamplifier that will be required when using the microphone.

The thicker the ribbon material, the more mass it will have (for a given width and length)... and the less the velocity it will achieve when the sound wave impinges on it. [The ribbon's acceleration = the force produced by the sound wave impinging on the flat face of the ribbon / the mass of the ribbon]. Therefore, it is important to use a very thin aluminum ribbon (preferably less than 1 micron thick).

My experience in building a ribbon microphone, a few years ago, was that it was a fun project... but the achievable sound quality was not anywhere as good as with any of my commercial microphones. Everyone enjoyed listening to it... and everyone enjoyed it even more when I switched back to my RE-27.

Stu

[W3GMS]

Very interesting detailed explanation Stu. 

When I use and listen to my RCA 77, other than the output being low, the response does not seem to be "tilted" to the low end but more of a smooth flat sound.  Are you aware of anything unique RCA did in that design of those microphones?  Yes, on air some presence rise is a requirement very similar to what would be expected with any flat response microphone.   

Here is the published response curve for a RCA 77D and I do see some bumps in the low end response around 200 Hz or so but nothing terrible dramatic.       

https://www.google.com/search?q=frequency+response+of+rca+77d&tbm=isch&imgil=mPlUG3He-gdGOM%253A%253Bhttps%253A%252F%252Fencrypted-tbn0.gstatic.com%252Fimages%253Fq%253Dtbn%253AANd9GcSlTMI_9W9sXTfZs_7-U2dTHQu-_EXXsYVrHEi59CYLmIOKG0DX%253B640%253B356%253BgCbwx98_L2iBtM%253Bhttp%25253A%25252F%25252Fwww.coutant.org%25252Frca77d%25252F&source=iu&usg=__sUYDlT6qSkw3OBfRcIfveKWR6oc%3D&sa=X&ei=bYSgU8z4B-y0sATZ7oGwBg&ved=0CCgQ9QEwAQ&biw=1280&bih=862#facrc=_&imgdii=_&imgrc=aaZOv1sotiLLEM%253A%3B6UY4nD9RH5HwGM%3Bhttp%253A%252F%252Fwww.coutant.org%252Frca77dx%252Ffreq77.jpg%3Bhttp%253A%252F%252Fwww.coutant.org%252Frca77dx%252F%3B559%3B938

Joe-GMS     

[AB2EZ]

Joe

RCA (and others) employ an acoustic cavity that works as follows to flatten the frequency response:

1. The acoustic cavity produces a reflected acoustic wave (travelling, of course, at the speed of sound) that impinges upon the flat side of the ribbon that is opposite the side that the incoming acoustic wave impinges upon.

2. Therefore, both the incoming and the reflected waves push on the ribbon... but in opposite directions.

3. The extra path length travelled by the reflected wave (i.e. from the front side of the ribbon, to the wall of the acoustic cavity, and back to the reverse side of the ribbon) causes a phase shift. So the reflected acoustic wave impinging upon the back side of the ribbon is out of phase with the incoming acoustic wave impinging upon the front side of the ribbon.

4. If the two acoustic waves (incoming and reflected) were exactly in phase, they would push on the ribbon with approximately equal force, but in opposite directions. Therefore, there would be no net force to push on the ribbon, and to cause it to accelerate.

5. But, since the incoming and reflected waves are not in phase, there is a net force to accelerate the ribbon.

6. One can show that since the phase difference between the incoming and reflected wave is caused by a fixed amount of delay, the phase difference will be larger at higher acoustic frequencies (phase difference = 2 x pi x delay x frequency).

7. One can also show that: the higher the frequency (and therefore, the larger the phase difference between the incoming and the reflected wave), the larger is the net force on the ribbon... and, therefore, the higher the acceleration. I.e. the net force ~ sin (the phase difference) ~ the phase difference, if the phase difference is not too large.  The acceleration of the ribbon = the net force pushing on the ribbon / the mass of the ribbon.

8. Therefore, as a result of the reflection of the acoustic wave in the microphone's acoustic cavity... incoming acoustic waves with higher acoustic frequencies produce more acceleration of the ribbon. Acceleration ~ f

9. But, the velocity of the ribbon is the integral, over time, of the acceleration.

10. Therefore, for a given amount of acceleration, higher frequencies produce less velocity. I.e. the integral of cos (2 x pi x f x t) dt = (1/f) x [1/(2 x pi)] sin (2 x pi x f x t)

The combined effect of 8. (i.e. acceleration ~ f) and 10. (i.e. velocity ~ (1/f) x acceleration) results is an (approximately) flat frequency response if the acoustic cavity is properly designed (e.g. RCA 77dx).

Stu

[W3GMS]

Thanks Stu.  So in the end, it largely depended on the acoustical cavity into which the element is placed to keep the response relatively flat. 

I have been noticing a lot of imported relatively inexpensive ribbon microphones appearing on the market but I have yet to work anybody using them. 

I as well like my RE-27 although I have been considering purchasing a Sure SM-5 that has quite a good reputation.  Its long out of production but you see them come up for sale from time to time.     

Joe, GMS