Active ribbon-mic

[ben_allison]

Nice! I'll be using it with an SM7. I don't think mic-noise will be an issue.

 

[ricardo]

Who's built PRR's circuit (and used the resulting mikes) on page 5 of this thread?

pic also at http://headfonz.rutgers.edu/no-tran-ribbon-boost-Phantom.gif

Don't think there have been any improved versions posted except Rossi suggesting & PRR agreeing to dispense with C1/2.

I rather think that R3/4 22R would increase noise compared to a nominal 300R ribbon straight into a THAT1510 preamp optimised for ribbons w/o P48V etc.

Anyone with comments from experience?  Also is the ribbon safe if one line is shorted.

 

[zebra50]

Hi Ricardo,

A few people have built it including Rossi and myself - best choice transistors etc are discussed above. It does work, it gives a very flat frequency response and it does give you around 10 to 12 dB gain.

If you have a good quiet preamp with plenty of gain, and a ribbon mic in good order, then you probably don't need one.

If you're struggling for level with an older 30/50 ohm ribbon mic then a matching transformer is the way to go, in my opinion.

> Also is the ribbon safe if one line is shorted.
If you omit C1 & C2, and short one side to ground, you could put a DC voltage across your ribbon mic's output transformer secondary - so use good cables!

Stewart

 

[PRR]

> I rather think that R3/4 22R would increase noise compared to a nominal 300R

Can you calculate it?
Using rough math I get 1.071:1 or 0.59db.

> a THAT1510 preamp optimised for ribbons w/o P48V etc.

If you have that THAT, why would you need a boost?

> is the ribbon safe if one line is shorted

Fix your cables.

Ribbons, bad cables, Phantom in the room..... BAD mix.

C1 C2 will limit long-term unbalance current but can dump a BIG thump, and that's all it takes to throw a ribbon out.

Spend 3 dollars and 20 minutes twisting one up, put a cheap 600:600 where the ribbon mike's secondary would be, 'scope that, and then drop a screwdriver on one Phantom line. Try-and-fry is really a better path than asking around for various opinions.

> If you omit C1 & C2, and short one side to ground, you could put a DC voltage across your ribbon mic's output transformer secondary

Asking Idiot Assistant, it says in the one-line-shorted condition there is "zero static voltage" across the mike secondary, 1.5 pico-Amp through it, with 0.3pA numeric error. The reason why is interesting but unintended. That's static.... if you poke a nail through the cable while-powered there will be a transient and I won't sim that because I would not trust it.

 

[ricardo]

> I rather think that R3/4 22R would increase noise compared to a nominal 300R
Can you calculate it?

About an additional 0.6dB NF for 300R

> a THAT1510 preamp optimised for ribbons w/o P48V etc.
If you have that THAT, why would you need a boost?

You don't. I'm on another forum discussing this and I've offered my own poor take on the subject.
http://tech.groups.yahoo.com/group/micbuilders/files/Ricardo/ActiveRibbon/
Please feel free to take it apart.  It's "virtual earth" so that's a start.  I'm just pontificating and trying to persuade someone to try it out. I haven't got a ribbon. I'm a capacitor guy.

> is the ribbon safe if one line is shorted
Fix your cables.
Ribbons, bad cables, Phantom in the room..... BAD mix.

All good points, Guru PRR.

C1 C2 will limit long-term unbalance current but can dump a BIG thump, and that's all it takes to throw a ribbon out.

Does this mean you recommend NOT having C1/2?  My instinct is not to have them.

Spend 3 dollars and 20 minutes twisting one up, put a cheap 600:600 where the ribbon mike's secondary would be, 'scope that, and then drop a screwdriver on one Phantom line. Try-and-fry is really a better path than asking around for various opinions.

Of course.  But what sort of zap can a ribbon stand?  Which ribbon? You can work out stored energy in caps and integrate the pulse on a storage scope to measure the zap.  But what does that tell you?  I'm not looking for opinions as much as experience.  OK, guru opinion is probably backed up by experience in similar fields so I'll have the opinion too.  But it's hard to get someone to try their vintage 4038 on an experimental circuit  ;D  But if someone has already done it ... 

.. Idiot Assistant.  Was is das?

 

[zebra50]

Hi Ricardo,

People can't see your circuit unless they have a yahoo account. 

Can you add the file? There's an attachment facility under "additional options".

Cheers!

Stewart

 

[ricardo]

my little circuit

 

[mrclunk]

I built a dual version of the prr circuit for my B&O BM5, works great!
I used the original transistors suggested by prr as i couldn't find the others.

I've also tried it with RCA 44, 77's and Altec 639's.
The biggest improvement at that particular studio was with the Altec's which seemed to be picking rfi from somewhere, with the ribbon booster, zero.
I've pcb layout that fits into a standard Hammond type guitar pedal box if anyone's interested?

 

[PRR]

Your R5 R6 3K have somewhat less noise disadvantage (small) than my 22r resistors.

Your NFB connection shows a low impedance to the mike. Olson advocated hi-Z loading, and this has been the custom until recently. Olson wasn't perfect, and custom changes.

If your C1 C2 are not WELL matched, hot-plugging will thump the ribbon. (Yes, "don't do that!", and true my plan may also have startup unbalance.)

No, I do not know how much thump is bad. (I've bunged a few ribbons but never while a meter was watching....)

 

[Marik]

Olson advocated hi-Z loading, and this has been the custom until recently. Olson wasn't perfect, and custom changes.

I never read Olson advocating hi-Z loading. The manual for RCA77 mentions ribbon unloaded as a measurement condition. That would be interesting to see a graph with the loaded ribbon. Because of the particular construction of that mic the ribbon is heavily damped and I would not expect the difference more than 1-2dB.

Best, M

 

[ricardo]

Thanks for your comments PRR.  This started in micbuilders. One ex-Shure member builds his own ribbons from scratch and is considering building it into a mike so he can damp the ribbon to suit the virtual earth.

The mechanical 'response', impedance plot and Ri of the preamp all interact.  But the efficiency of ribbons preclude significant 'electromechanical damping'.

I don't think my circuit is worthwhile unless for a 30R nominal ribbon like the original STC 4038, hence the high 30dB gain.
I've actually said that if you have a 300R ribbon, PRR's proven circuit should be used. Of course if you want to build a dedicated THAT1510 preamp for 300R ribbons ...

Having said all that, I feel you need 20dB gain for a 'normal' -55dBV/Pa 300R ribbon which is difficult with your circuit without excessive Ro.  In my experience, Ro in the tens of ohms is required for the very best RFI immunity.

Hmmm.mm!  The 300R Coles 4038 needs 30dB which means the 30R STC 4038 needs 40dB. I can make C1/2 even smaller & pose less danger to the ribbon.

 

[Marik]

The mechanical 'response', impedance plot and Ri of the preamp all interact.  But the efficiency of ribbons preclude significant 'electromechanical damping'.

Except the tuning resonance the impedance plot of the ribbon is pretty linear up to the highest frequencies, so the efficiency is less relevant. There is no real need for the significant (whatever quantity it is) damping.

Best, M    

 

[ricardo]

But the efficiency of ribbons preclude significant 'electromechanical damping'.

You can see the relative significance by looking at the impedance curve of the mike.

BBC Engineering Monograph No.4 "THE DESIGN OF A RIBBON TYPE PRESSURE GRADIENT MICROPHONE FOR BROADCAST TRANSMISSION" - Shorter & Harwood
Fig 19 impedance at the nominal 300R output of 4038, shows a LF peak of 340R with 'DC' of 275R.
The difference between 'DC' and the peak 340-275=65R shows the relative importance of

Mechanical (including acoustic) damping Rm
and
Electromagnetic Damping (shorted output)
Rem = (Bl)^2 / re  B: Tesla  l: metres  re: ohms

Units are mechanical ohms. Ns/m

Rem / Rm = 65R / 275R = 0.236 so Electromagnetic Damping is less than 1/4 of Mechanical Damping in 4038.

This is exact for dynamic mikes & speakers.  A big peak means either little Mechanical Damping or a big magnet, loadsa Electromagnetic or both.  A small flattish peak means the device is mostly mechanically / acoustically damped.

For ribbons, the transformer tends to complicate things but it's still useful to get an idea of relative importance.

Thanks for posting how you do your transformers, Guru Marik.  I confess my pontificating on Cu tape windings was a guess but I now have a transformer guru to support me.

Could you comment on HF performance using Cu tape along the lines of PRR (for sectioned windings) in
http://www.groupdiy.com/index.php?topic=5743.20

 

[Marik]

Edit: deleted as an irrelevant information.

 

[leswatts]

Thanks for your comments PRR.  This started in micbuilders. One ex-Shure member builds his own ribbons from scratch and is considering building it into a mike so he can damp the ribbon to suit the virtual earth.

Well, that would be me.

I'm dealing with (acoustic) resistance controlled ribbon pressure microphones among other things.

Active would give me some options with fairly easy complex electronic EQ as well as increased output.
More options than waffle plate resonators and irregular baffles anyway.

Ricardo mentioned the best bet of all might be a very high ratio transformer providing all the voltage gain with
FET/bipolar followers providing current gain/impedance matching . One manufacturer does this.
Noise and fault protection could be very good...depending.

I do have my transformer guy working on high ratio transformers...they would be greater than 100:1 TR.
Lundahl makes a 110:1. I'm sure Marik could too.

Of course the circuits discussed here and on micbuilders have greatest utility with existing microphones/transformers as a booster.

Les

 

[ricardo]

Gentlemen,

I've just entered the 21st century and downloaded LTSpice.

But I've difficulty finding transistor models.

I believe some of you have SPICE models of the various circuits in this thread eg PRR's on page 5.

May I ask for a copy of these including the transistor models used?  Please PM to avoid reducing the S/N on the forum.

Many thanks

 

[PRR]

Isn't it quicker to build it than to SPICE it??

For a model, use any good-gain transistor of the right polarity. The differences between them are not quantifiable in SPICE.

 

[ricardo]

Isn't it quicker to build it than to SPICE it??

I'm trying to work out why my offset adjustment dun wuk.  And I wannabe a SPICE guru too.

For a model, use any good-gain transistor of the right polarity. The differences between them are not quantifiable in SPICE.

I sorta realise that but ANY transistor model is new to me.

 

[PRR]

> why my offset adjustment dun wuk

With any two instances of the same SPICE transistor, at the same bias, there is NO offset because the two "parts" are identical to the 13th decimal place. If you want to trim offset you probably have to insert a 0.005V battery in one Base. (Some libraries allow parameter tolerances; this is probably more trouble than it's worth.)

I don't recall an offset trim; your invention?

 

[ricardo]

With any two instances of the same SPICE transistor, at the same bias, there is NO offset because the two "parts" are identical to the 13th decimal place. If you want to trim offset you probably have to insert a 0.005V battery in one Base. (Some libraries allow parameter tolerances; this is probably more trouble than it's worth.)

Surely SPICE allows hfe & Vbe (vs Ic) to be varied in the model?

I don't recall an offset trim; your invention?

Reply #246 18may11