Ribbon microphone

[jzagaja]

Let consider 4x50mm ribbon immersed between 10x5mm magnets and 1:37 low inductance transformer (RCA 1931 type) which is flat measured at secondary (no primary load). Electret microphone is as close as the centre of the transducer as possible. Stimulus (MLS, sine sweep) is connected to the secondary at level 0dBu. What causes rising top end response?

[cuibono]

The ribbon mics I used to work with always had a quick drop off above 10k. I have no idea why you're getting an increase....

Maybe a measurement artifact?

[bear]

cavity resonance?
reflection?

[jzagaja]

I thought it could be cavity resonance but it should be different in red and blue curve because those two have totally different slits and slit/gap parameters can roll up the response by 5-6dB only - see Verhagen book.

[bear]

dunno, looks a lot like a 6dB oct boost? If I recall what I saw last night... also check the mic against something else?

check how it measures at double the distance?

consider some effect of the presumed matching transformer also - possible HF resonance in it?

_-_-bear

Oh, also use the mic to monitor a pink noise source into a driver known to be reasonably flat - compare the measurement mic to the test mic??

_-_-bear

[Marik]

Originally posted by jzagaja
Let consider 4x50mm ribbon immersed between 10x5mm magnets and 1:37 low inductance transformer (RCA 1931 type) which is flat measured at secondary (no primary load). Electret microphone is as close as the centre of the transducer as possible. Stimulus (MLS, sine sweep) is connected to the secondary at level 0dBu. What causes rising top end response?

What do you mean by "low inductance transformer"?

You graph looks like inverted on the top end.
Judging from the picture and known dimentions, the path is about 40mm which translates into 8.6KHz null.
Sure, becasue of pretty big cavity and resonances the null can be much smoothed, but it still does not explain that raise and HUGE peak @ ~18KHz.

But wait a second... your ribbon is not corrugated, so it basically works like a marimba bar. Eyeballing the size and considering the weight of 0.6um ribbon it makes much sense of its self resonance of 18KHz.
Make your ribbon corrugated and the whole mic will turn into a perfect mass controlled system, with a smooth response over the whole bandwidth.

Best, M

[jzagaja]

Dear Marik,

Good to see you here. I've found your contribution to the ribbon topics at Prodigy Professional. People there really know this things.

Looking for core material, size, wire diameter, screen material and wound topology.

Now my ribbon is corrugated. Please notice that rising top end doesn't depend on foil thickness.

At the moment I moved back to microphone and amplifier. My OPA1632 trial was unsuccessful - output is highly unbalanced. Schematic appended.

[jzagaja]

What do you mean by "low inductance transformer"?

Low leakage inductance - important in ribbon-speaker

[Marik]

Originally posted by jzagaja

Now my ribbon is corrugated. Please notice that rising top end doesn't depend on foil thickness.

Anything changed with corrugated?

Originally posted by jzagaja


Low leakage inductance - important in ribbon-speaker

For ribbon mics the low leakage inductance (as capacitance) is not that important. True, the ratio of 1:37 seems like very big, but since we operate with such small source impedances the actual secondary turns number (to square of which the leakage inductance is proportional) is not that big.

Best, M

[jzagaja]

Anything changed with corrugated?

Can't measure - it's too silent. But wait a minute - 10k output impedance from souncard and 0.1k from ribbon. Should I use buffer?

From the other hand I have used my 1:8 tranny and Tripath amplifier. I couldn't measure with my laptop integrated sound card (noise, weird impulse response). With Digigram and it's balanced output I've got again low signal from power amp.

Rising top end is still visible.

Confused again