Ribbon mic (DIY) - Bigger gap to shorten front to back path.

[w0rf]

Hi Guys,

A lot in interesting ribbon threads in here
(I have been reading quite some of the list in this thread: http://groupdiy.com/index.php?topic=24012.0)

Now I am actually wondering why the magnets are always placed next to the ribbons?
Ok, this will give the highest magnetic field, but is that the only reason?
In the early days this would have been the only option, due to magnet strength.

But it seems beneficial the shorten the front to back path in order to have more high frequencies,
Shorter patch is less HF cancelation.

But what if you remove the front to back path, or better said, reduce it to the thickness of the ribbon

So instead of leaving a 6mm gap for the ribbon (just to give an example) why not have a 10mm gap?
(in inch that must be something like 0.4" instead of 0.23" )
2mm opening for the sound, 6mm ribbon and then 2mm air again.
This way the sound could pass directly to the back of the ribbon.

With todays stronger neodymium magnets I don't see why this wouldn't be possible.
You will ofcourse widen the motor by factor 2 or 3 (bigger magnets + maybe a pole to focus the magnetic field)
And weight and magnet costs will increase.

You would need to focus the magnetic strength, which could be done by a triangular iron pole or something else with high permeabilty.
The flat site of the triangles toward the magnets, and the pointy side towards the ribbon.
This should make a 10mm gap much easier with less leakage (I would think)

And then put a magentic return path arround the whole thing (to the back sites of the magnets.)

Just thinking out loud.

 

[Matt Nolan]

I have seen older mics with triangular pole pieces tapering towards the ribbon. Also with a number of holes drilled through them from front to back.

 

[PRR]

Short air path = low output.

Yeah, magic modern magnets... but excessive field strength makes motion resistance-controlled instead of mass-controlled. I just swept the garage and am too tired to think which way the slope will go.

Olsen made 2 or 3 mikes of the same relative proportions but 2:1 difference in absolute size. The little one had extended treble but low output; the big one vice versa. By comparing them he got an absolute (not relative) Calibration without other references.

 

[ricardo]

Having bigger gaps between the ribbons & the poles would only give smaller path length at HF where the wave lengths are small enough to squeeze through.

For any sensible effect, you would have to lose LOADSA sensitivity.

The LF path length would still go round the long way.

BBC Engineering Monograph No 4 describes the design of the 4038 in detail.  It uses different path lengths to tune the response.

The 4038 is probably still the best ribbon in the known universe though present ones are not good as the originals.  It's main fault is it's about 10dB softer than a modern ribbon.

It would be good to see a modern revision which addressed this without losing it's good qualities.

 

[emrr]

Ricardo, you should hear some of the modern ribbons.  The Samar I know in particular, and there are a few others which boast much flatter overall response with extended treble, having shorter paths and higher output.  There are shoot-out samples of the Samar against a Coles in particular somewhere, Recordinghacks I think. 

 

[ricardo]

The Samar I know in particular, and there are a few others which boast much flatter overall response with extended treble, having shorter paths and higher output.

Marik is in fact the guy who pointed out that 'modern' Coles 4038 don't have as extended HF as the original STC/BBC 4038s so I can well believe the Samars are one of the few modern ribbons with extended HF.

Which other ones are flat?

I've only measured one Royer and it is certainly not as flat as the 4038s I measured circa 1980 (one 'new' STC/BBC and one very early one used by Gilbert Briggs & Raymond Cooke at Wharfedale in the early 80s/ late 70s.  I was pleased to find the 2 matched very well though they were made 10yrs apart)

The Royer had some of the nice features of the 4038 but not all.  It had less extended LF and quite a sloping response.

 

[emrr]

Nice to know.  I believe at least Audio Technica and Mesanovic are making similar claims.  I've not heard either.  I think there are one or two more new companies in there as well. 

 

[w0rf]

From the sample I have heard from of those newer ribbons, some of them sound really nice.
I like the ones from Samar & mesanovic 2..
They seem to have a better extended frequency response as the older ribbons.

In mean time I've read some interesting stuff. The front to back path would be needed to have the lower frequencies, or it would be all HF. So some path is needed, and can be estimated based on the frequency response that we want.

I have found some explanation to calculate the flat response range and -3DB drop point (not invented myself).
up to 180° there is a full response, and then for ever 45° -3DB
-So if you want a flat range up to a certain frequency, then the front to back distance would need to be halve the wavelength of that frequency.
-For the -3DB drop point  that would be 5/8 of the wavelength.
-If we take 16kzh as an example, the wavelength would be +- 21.25mm right?
In this case front to back patch would need to be 10.625mm long.

So imagine the ribbon is 6mm and the poles are like 5x3mm then you have a path of 6/2 + 5 + 3 = 11mm ?

11mm x 2 = 22mm -> flat freq range is up to +-15.45khz
for -3DB we are at 5/8th of the wavelength: 11mm/0.625 (5/8) = 17.6mm -> 3DB drop would be at +-19.32khz

But this is not achievable with neodymium as poles,
Even if you could find them in a size like this (which might need some custom order) the magnetic field would not be strong enough. So I would need some iron/steel poles to do something like this. and even then they can only be thin in the middle of the ribbon. Towards the magnetic source, the poles need to grow considerably bigger, in order not to saturate the poles.

Also this doesn't take into account the strength of the magnetic field, stronger field would have less Hi Freq.
Then there are the resonator plates that are used for more HF.. anyone an idea if some usefull ones can be bought?

 

[w0rf]

If one would use low carbon steel as poles and for return path.. it would require some work to get them into shape, which means they will be work hardened.

Does anyone one know if basic annealing is enough to restore the high saturation point?
Or does it need to be done in a hydrogen environment?
Things like mu-metal are done in hydrogen environment to have a high permeability, but in this case I do not care about a very high magnetic permeability.. I just want a high saturation point

The BBC Engineering Monograph was pretty interesting in explaining several items that need to be taken into account..
I'm beginning to think that make something "nice" yourselves might be not so easy. Not sure if if its worth to put some money into something that might not even work..

One could always buy one of those DIY ribbon motors, with a +-14mm path, but then you get up to 12khz flat, and 15.17khz as -DB point, 18.22khz as -6DB.. not really wow.
Some resonator plates could get that response a bit better maybe, but the waffle plates ones on those chinese mics, like RB500/apex 205 don't seem really great, so it seems all about finding good ones..

 

[ricardo]

-So if you want a flat range up to a certain frequency, then the front to back distance would need to be halve the wavelength of that frequency.
-For the -3DB drop point  that would be 5/8 of the wavelength.
-If we take 16kzh as an example, the wavelength would be +- 21.25mm right?
In this case front to back patch would need to be 10.625mm long.

So imagine the ribbon is 6mm and the poles are like 5x3mm then you have a path of 6/2 + 5 + 3 = 11mm ?

11mm x 2 = 22mm -> flat freq range is up to +-15.45khz
for -3DB we are at 5/8th of the wavelength: 11mm/0.625 (5/8) = 17.6mm -> 3DB drop would be at +-19.32khz

That's for -3dB @ 45° compared to on-axis and better than ANY RIBBON MIKE IN THE KNOWN UNIVERSE.

It's even better than a 1/2" B&K omni measurement mike.

But this is not achievable with neodymium as poles,
Even if you could find them in a size like this (which might need some custom order) the magnetic field would not be strong enough. So I would need some iron/steel poles to do something like this. and even then they can only be thin in the middle of the ribbon. Towards the magnetic source, the poles need to grow considerably bigger, in order not to saturate the poles.
...
but in this case I do not care about a very high magnetic permeability.. I just want a high saturation point

NeFeB has the HIGHEST flux saturation of ANY commonly available material.  MUCH better than fancy steels or mu-metal.

My cost-no-object loudspeaker would have Alnico magnets with NeFeB poles in the treble units.

Then there are the resonator plates that are used for more HF.. anyone an idea if some usefull ones can be bought?

If you are serious about this, you need to match your 'resonator plates' to your mike.  The BBC Monograph has some clues.  But first you need to have a convenient way to accurately measure frequency response.

BTW, Raymond Cooke & Gilbert Briggs used a 4038 to measure speakers cos it was the flattest mike available to them at the time.  Loadsa caveats when measuring speakers with a particle velocity mike .. some of which are in Brigg's classic book, Loudspeakers.