Ribbon material (a silly question!)

[PRR]

> good ribbon transformer have 6 dB noise figure with 0,5 Ohm ribbon. There is 1600 ohms of transformer noise resistance. You can make lower intrinsic noise resistance of ribbon mic with thick ribbon, but only with special transformer you can improve overall noise.

Clearly you know what you mean. But when you type so fast it is hard to know what you are saying.

It should be perfectly possible to wind a transformer, for almost any ribbon resistance(*), with losses around 10%. That implies a noise figure (before the amplifier) of about 1dB. The primary is just a few turns and very fat wire will fit: Beyer used wire not much smaller than the power lines inside my walls. The secondary resistance when wound for about 200 ohm nominal (over 1K actual) line impedance will generally be under 50 ohms.

(*)Below maybe 0.01 ohms, it may be hard to wind such a low-impedance winding: it works out to less than half a turn and you can not have less than half a turn on an E-I double-window core (one turn on a single-window core). But it seems easy enough to build the ribbon to more than 0.01 ohms.

Indeed, ribbon mikes with well optimized transformers can be the lowest noise microphones of any type. It is one of the points that Olsen kept emphasizing when he was defending "his" ribbons against encroaching condenser and dynamic techiques. Common dynamics and condensers have to be loaded with resistance (acoustic resistance) to damp their resonance and flatten their response. The self-noise of a condenser in the 6KHz range is often dominated by acoustic resistance noise. The ribbon mike tends to be flat without any damping resistance.

 

[SilverhammerNZ]

Yeah, I'm the "aluminium leaf" guy. It's horrendous to work with but I have had really good results when I've got it right. I have recently tried some 2micron material but I'm afraid I'm still sticking to my leaf.

Once in place and if you haven't incurred any lateral cracks in the ribbon then it will last if it's protected with silk/guaze etc.

How do I do it Marik? well I've tried to explain on my website but I do keep it "wet" with isopropol so it doesn't flutter all about in the breeze.

By "wetting" the ribbon I find I can then pick it up with a cotton bud (the wet ribbon just sticks to it). I can't use gears or a crimp to impart corrugations, I place the wet ribbon down on a corrugated surface such as around a camera lens (the grip part). I then gently press the ribbon into the corrugations, sort of roll the cotton bud over the ribbon.

Pick it off then gently try and place it into position, successfully clamping it in is a halleluja moment for sure.

Much easier if you have slightly thicker material such as 2 micron, but so far I have yet to find any that sounds as good as my leaf. I think the 2 micron material I have tried is the Lebow stuff, perhaps the other more expensive material is better? more pure perhaps? Someone else I know has had better results than me with the Lebow material.

Larry

 

[frede]

Larry, have you ever had the problem with the aluminium leaf loosing the tension over time?

I have just bought a old Bang & Olufsen BM3 microphone with a defect ribbon. I will try to make a new ribbon using your method.

 

[Guest]

[quote="PRRIt should be perfectly possible to wind a transformer, for almost any ribbon resistance(*), with losses around 10%. [/quote]
Dear PRR,
My mails can be unclear , it is not caused by fast
typing. My english is poor.

Problems of noise figure of input ribbon transformer
is not simple.
Olson in his paper Microphone Thermal Agitation Noise (JASA vol.51
No 2 part 1 p. 425-432. He has RESISTANCE (not impedance!!!
marik sometimes this not respect)
of ribbon 0.25 Ohms and transformer 0.05+0.05; it is 0.1 .
Olson uses very low inductance transformer to improve noise
figure (transformer noise amount was stull 40 percent and not yours 10)
After this low-inductance transformer, microphone LF corner was
rather high- 50 Hz.
Olson have some simplifications. i.e. Barkhausen noise of
transformer was omitted, but parameters of his mic (noise 1.2 dB)
are very good.
It is problem to make transformer with several turns (i.e. 20 -
this is value normally computed for ribbon and mumetall)
Wire have critical diameter and under it wounding is uneffective.
You can not wound 1,5 mm wire onto 5 mm skelet.
It is possible to wound primary with copper ribbon wire, but
where it buy?
Some producers wounds many wires in paralel.
On toroid is simplier to wound thick wire than to skelet.
But problems are with secondary.
xvlk

 

[Guest]

[quote author="Marik"]Dear XVLK,
The transformer I am using is a Lundahl LL2911, which is a very good iron, specifically designed for ribbon microphone use.
[/quote]
Yes, but for what ribbon this is specifically designed?
If I buy integrated circuit, I know recomended connection;
If I buy ribbon transformer I want to know recomended ribbon.
Or there are some ribbon normalization?
Why ribbons from RCA can be used in bayerdynamics and so on?
What are prescribed dimensions of normal ribbon?

PRR, please, translate it for Marik.

xvlk

 

[sismofyt]

frede, I've worked on a few of those BM-3s. Let me know if you need help (mostly mechanical). I also got some spares, I think.

 

[sismofyt]

Maybe we should try to get Steven Sank on board? :thumb:

 

[SilverhammerNZ]

To frede,
No I haven't observed any slackening of the ribbon over time. The material I use is similar in thickness to the Coles ribbons.

I guess because the material is very thin it may be more suseptable to wind blasts, but normally good sheilding should stop this.

Larry

 

[Marik]

[quote author="xvlk"][quote author="Marik"]Dear XVLK,
The transformer I am using is a Lundahl LL2911, which is a very good iron, specifically designed for ribbon microphone use.
[/quote]

PRR, please, translate it for Marik.

xvlk[/quote]

XVLK,

No translation necessary, as I see what you are saying. It seems that you are very much into theoretical aspects of ribbon coupling, which I respect and appreciate.
On practical level in ribbon mics there are however, whole bunch of other problems and compromizes to think about, first place. Although the noise and carefull transformer matching are very important aspects of the mic design, as you could probably see, at this point I am still concentrating and experimenting with acoustical and mechanical properties of the ribbons. Until I exhaust all the ideas I have at this point, DIY'ing mic transformer for "perfect" match are not visable, as I am still working on transducer itself. So at least for now, I will stick with LL2911.

for what ribbon this is specifically designed?

When I talked to Per Lundahl a few years ago about this trafo, he told me that it was designed for "RCA large type" of ribbons (i.e. ~0.18"x2").

Why ribbons from RCA can be used in bayerdynamics and so on?

They cannot because they are of different dimensions, but if you talking about material then..... if it works and sounds good--by all means use it. In fact, as far as I know, Stephen Sank (whose father designed RCA ribbons) re-ribbones Beyers with RCA foil with excellent results and gets very good reviwes from such respected sound engineers as Harvey Gerst and others.

What are prescribed dimensions of normal ribbon?

There are no "prescribed" dimensions, as well as there is no "normal" ribbon.
Once again, everything depends on mic's application, what do you want from the mic, and which compromises you are willing or have to make.
For example, if you need a mic with good vertical plane directivity, you will need to make it shorter, but at the same time the sensitivity will be lower. Since the shorter ribbons tend to have more resonanse modes, you will need to use acoustical dampening, which will affect low end and sound in general.... and so on. But I am sure you know all these.

BTW, I would be very interested to see the Olson article you have mentioned. I have his articles from AES and JASA which were compiled into RCA microphones book, but not the one you are talking about.

He has RESISTANCE (not impedance!!!
marik sometimes this not respect)

Olson has resistance, Rosen and Robertson have impedance......

 

[PRR]

> He has RESISTANCE (not impedance!!!

Yes, the difference is often confusing.

Ribbons come in many different sizes and thicknesses. There isn't really any "typical" ribbon.

Say we have built a ribbon with DC resistance of 0.1 ohms (a nice even number, easy math).

The impedance will be 0.1 ohms over most of the audio band. It is too small to be loaded by air except at the top of the audio band. We go to great trouble to avoid resonance in the middle of the audio range. There is always a resonance at the bottom of the audio band, and impedance will rise there. We see the same thing in loudspeakers. And for the same reason as in loudspeakers, we can mostly ignore the impedance rise at bass resonance.

0.1 ohms is the impedance, so for maximum power transfer we would load it in 0.1 ohms. However what we really want is maximum signal-to-noise. That happens when we run the transducer "un-loaded".

0.1 ohms is an awkwardly low impedance for a cable. Cables typically have 1 ohm to 10 ohms resistance, which will add to the 0.1 ohm source impedance.

0.1 ohms is a very awkwardly low impedance for an amplifier. Practical amplifiers have noise sources like 200 ohms to 20K ohms.

So we need a transformer.

The transformer should have a "nominal impedance" of 0.1 ohms, "matching" the ribbon. However it must really have a very low copper resistance, and a very high un-loaded impedance. For a 0.1 ohm ribbon, we would like 0.01 ohms copper resistance, and unloaded impedance well above 0.1 ohms (preferably over 1 ohm) all across the audio band. Good transformer design may not be easy. But this is not a very hard specification. You may have to use quite fat wire, or parallel strands, or get some sheet copper and slice it into strips. Compared to the problem of making the acoustic ribbon, the problem of winding the transformer is "easy".

The secondary of the transformer will be nominally 150 ohms, but the standard loading at a mike-amp input is 2K or more. The transformer impedance ratio is 0.1:150. The transformer winding ratio is 1:39. If you stand at the mike-amp and "look in to" the mike, you see the ribbon resistance transformed up to 150 ohms, plus transformer copper resistance and cable resistance. The total source resistance will be about 180 ohms, of which 150 ohms is the ribbon signal, and 30 ohms is dead copper loss. Noise figure looks like 1dB at this point. We can find mike-amps with 1dB noise figure in 150-200 ohms, so the total noise figure is a couple of dB.

Note one thing: because we don't normally have an amplifier at the mike, and often run long cables, we end up with two transformers. One steps 0.1 ohms up to 150 ohms for the cable; the other steps 150 ohm cable signal up to 1K-10K to work well with the noise of a simple tube or transistor. (We can use a transformerless mike amp, but that just gives different problems.) So we have two cores and four copper winding losses. While we can not avoid having a transformer very close to the ribbon, we could put an amplifier inside the mike body and raise the signal power enough to swamp noise in the next stage. This also isolates the ribbon and transformer from the rest of the world, so if there does turn out to be some special loading condition that improves the sound, we can set the load in the design.

> at this point DIY'ing mic transformers for "perfect" match are not viable

Perfection is hard to get, but changing the turns ratio should be possible for anybody clever enough to make and mount a ribbon. Take any mike transformer. Peel away all the shielding (for experimentation... for a finished product you will need shielding). Leave the 200 ohm winding alone. Wind several turns of wire around the windings. Use a milliVoltmeter to measure the turns ratio, and compute the nominal impedance. Use wire tables to know the resistance. In modern audio interfacing, you want the ribbon resistance to reflect as 150 to 300 ohms (same as a dynamic mike voice coil is built or transformed to 150-300 ohms). You want the copper loss to be much less than the ribbon impedance.

Just winding around the outside will not give best high frequency response. Leakage inductance will be high. But with normal 2K loading, a little leakage inductance is not a big problem. And windng capacitance will be no problem at all. Measure the transformer with 0.1 ohm source and 2K load, note how un-flat it is and allow for that in testing. When you get something you like, have a custom transformer wound with the primary tucked in the middle of the 150 ohm winding.

 

[Marik]

PRR,

:thumb: :thumb: :thumb:
as always.

Do you have any comments here:

http://www.groupdiy.com/index.php?topic=1641&postdays=0&postorder=asc&start=30

Message on the top of the page.

 

[crazydoc]

Marik, PRR, or anyone else with an insight into this:
On the subject of ribbons, is there any reason besides fitting it into the narrow gap of a strong magnetic field that the ribbon takes the shape that it does? If a ribbon type transducer did not generate electrical signals by cutting through a magnetic field, but rather, say, by reflecting optical signals, would another shape be more efficient or more desireable from an acoustic standpoint? Possibly a circular membrane with concentric circular corrugations? Or a wider/shorter linear ribbon?

 

[PRR]

> If a ribbon type transducer did not generate electrical signals by cutting through a magnetic field

Then it would not be a "ribbon mike".

The ribbon mike is an elegant combination of an electromagnetic system that has a rising response with frequency, and a acoustic-mechanical system that has a falling response with frequency, to give a transducer with flat response over a wide bandwidth.

Take out the electromechanical side, and none of it makes sense.

> but rather, say, by reflecting optical signals, would another shape

Optic systems generally sense displacement. The simple constant-displacement acoustic-mechanical system is the stretched diaphragm as used in omni condensers.

 

[Marik]

[quote author="PRR"]>

The ribbon mike is an elegant combination of an electromagnetic system that has a rising response with frequency, and a acoustic-mechanical system that has a falling response with frequency, to give a transducer with flat response over a wide bandwidth.

Take out the electromechanical side, and none of it makes sense.
[/quote]

PRR,

It is hard to argue with person like you, but at least I will give a try. :grin:
As far as I know, the electromagnetic system is linear. Of course, as most of the things in ribbon mics, it is up to a sertain point, after which magnetic strength (saturation) will affect LF region.

The whole beauty of the ribbons is that mechanical system, since it is a mass controlled one, has a falling response with frequency, starting from tuning resonance, but acoustical system has a rising response with frequency. Because of nature of pressure gradient in biderctional system, force on ribbon doubles with each octave increase, and as a result,
acoustic-mechanical system is solely responsible for flat response.

 

[Guest]

[quote author="Marik"][quote author="PRR"]>
As far as I know, the electromagnetic system is linear.
[/quote][/quote]
Yes, linear is, derivative is linear operation.
[quote author="Marik"][quote author="PRR"]
Of course, as most of the things in ribbon mics, it is up to a sertain point, after which magnetic strength (saturation) will affect LF region.
The whole beauty of the ribbons is that mechanical system, since it is a mass controlled one, has a falling response with frequency,
starting from tuning resonance, but acoustical system has a rising response with frequency. Because of nature of pressure gradient in biderctional system, force on ribbon doubles with each octave increase, and as a result,
acoustic-mechanical system is solely responsible for flat response.[/quote]
[/quote]
It is not correct. There is no effect of acoustical system (magnets)
to microphone properties in the pass band other than velocity transformation. (it is rise of sensitivity caused by ratio of mabnet width
and ribbon width).

In the high band there is fall caused by magnets width.
Some microphones compensate this by kind of horn.

PRR only say, that acusto mechanical systhem is mass controled.
It is frequency fall and transducer (mechanic-electric) is derivative,
it is frequency rise. Thank you, PRR.

Crazydoc not use frequency- rise transducer. His microphone will
not have frequency - constant transfer. I had explained him it.


xvlk

 

[Marik]

[quote author="xvlk"][quote author="Marik"][quote author="PRR"]>
As far as I know, the electromagnetic system is linear.
[/quote][/quote]
Yes, linear is, derivative is linear operation.
[quote author="Marik"][quote author="PRR"]
Of course, as most of the things in ribbon mics, it is up to a sertain point, after which magnetic strength (saturation) will affect LF region.
The whole beauty of the ribbons is that mechanical system, since it is a mass controlled one, has a falling response with frequency,
starting from tuning resonance, but acoustical system has a rising response with frequency. Because of nature of pressure gradient in biderctional system, force on ribbon doubles with each octave increase, and as a result,
acoustic-mechanical system is solely responsible for flat response.[/quote]
[/quote]
It is not correct. There is no effect of acoustical system (magnets)
to microphone properties in the pass band other than velocity transformation. (it is rise of sensitivity caused by ratio of mabnet width
and ribbon width).
[/quote]

xvlk,

Unfortunately (or fortunately) I am correct here.
The confusion comes from terminology use, so lets define it:

Electromagnetic system--strength of magnets.
Mechanical system--ribbon tuning=mass controlled system.
Acoustical system--the way sound waves reach rear side of the ribbon.
It may depend on magnet size, size of pole pieces, or any wave obstacles on the way to the rear.
And now please re-read my message, where (I think) I clearly separated these three different systems.

it is rise of sensitivity caused by ratio of magnet width
and ribbon width

Ribbon width (ribbon area) has its own affect on sensitivity and is independant of magnet width affect--contrary to what your statement may suggest.

 

[Guest]

Electromagnetic system--strength of magnets.

... and the ribbon itself {current path perpendicular to magnetic field}
voltage=B*l*velocity = B*l*diff(x,t)
where B is induction in the gap {you can measure it
by Hall probe from old videorecorder}
l is efffective length of the ribbon
diff(x,t) is ribbon velocity

This is one part of equations of the electrodynamic conversion.
Second part {you uses it, if you control ribbon via transformer inductance
and so on...}

F=B*l*I
where I is currrent trough the ribbon.

Is elegant to form this equations in matrix form, where it forms
matrix of gyrator. It is important in general circuit theory as
example of nonreciprocal passive circuit.

This gyrator changes circuit elements and topology.
i.e. we see ribbon, which is controlled by mass as
capacitor on electrical side. But with series resistor {of ribon ohmic loss}

Mechanical system--ribbon tuning=mass controlled system.

Yes, but for ribbon mass control is dominant not his mass, but mass of air surrouds
it. And this mass is normally acoustical.

To separate acoustical and mechanical system you can use vakuum.

Acoustical system--the way sound waves reach rear side of the ribbon.
It may depend on magnet size, size of pole pieces, or any wave obstacles on the way to the rear.
And now please re-read my message, where (I think) I clearly separated these three different systems.

it is rise of sensitivity caused by ratio of magnet width
and ribbon width
Ribbon width (ribbon area) has its own affect on sensitivity and is independant of magnet width affect--contrary to what your statement may suggest.

Only if you have big gap between ribon and poles.

xvlk

 

[crazydoc]

[quote author="xvlk"]
... and the ribbon itself {current path perpendicular to magnetic field}
voltage=B*l*velocity = B*l*diff(x,t)
where B is induction in the gap {you can measure it
by Hall probe from old videorecorder}
l is efffective length of the ribbon
diff(x,t) is ribbon velocity

This is one part of equations of the electrodynamic conversion.
[/quote]

So the faster the ribbon moves, the greater its output, (voltage or current, assuming a constant impedance and a constant displacement of the ribbon.) So intuitively (to me) this means that for electromagnetic portion of this system, output is proportional to frequency.
So something must be damping this frequency dependent increased output, which I assume is a decreased displacement (excursion) of the ribbon as frequency increases, because of either the mass of the ribbon or the mass of air it has to displace.

Please correct me if I'm wrong.

Second part {you uses it, if you control ribbon via transformer inductance
and so on...}

F=B*l*I
where I is currrent trough the ribbon.

What is F? Frequency?

 

[Marik]

Electromagnetic system--strength of magnets.
Mechanical system--ribbon tuning=mass controlled system.
Acoustical system--the way sound waves reach rear side of the ribbon.
It may depend on magnet size, size of pole pieces, or any wave obstacles on the way to the rear.

xvlk,

This was in regards to PRR?s message (at least to how I understood it).

Instead of getting buried in details right now, taking things out of context, and confusing apples and oranges, for sake of clarity of discussion, please get back to original PRR?s message and then to my initial response, keeping in mind original points. In a meanwhile, I will wait until PRR gets back to clarify things (if he wishes).

it is rise of sensitivity caused by ratio of magnet width
and ribbon width
Ribbon width (ribbon area) has its own affect on sensitivity and is independant of magnet width affect--contrary to what your statement may suggest.

Only if you have big gap between ribon and poles.

xvlk[/quote]

How would you identify "big gap"?

 

[Guest]

[/quote]
How would you identify "big gap"?[/quote]

Gap between ribbon and poles bigger than 50 micrometers.
(!!!!!!!!! you must have microscope for reribboning !!!!!)
It is not my calculations, it is calculations of mr. Anderson,
man from Olson s laboratory.

It is very thin tolerance to ribbon dimensions and if it is not
obtained, velocity transformation via poles can not occurs
(there is leak in that gap)
(in the middle frequencies) and sensitivity of the mic can be
reduced to half.


What are you on the bridge measurements ?
Some values of your mic for me?

xvlk