Active ribbon-mic

[Rossi]

damn, the taxes... that reminds me..... oh well!

Yes, would be great if you came up with something. But of course, there's always other stuff to do. Maybe we're lucky and PRR chimes in while you're away. Anyway, take care.

Andreas

 

[clintrubber]

from bcarso:

For one cap to go away, just connect the tranny output between the base of the input Q and its bias network---a microamp or so of base current shouldn't hurt the tranny too much.

That's been suggested before in this thread, but could we always be absolutely sure there isn't something bigger happening at power-on/off ? The 'static'/operating-situation seems fine indeed though.

Hey, and in general, this is a nice thread ! Thanks for the informative messages.

Bye,

Peter

 

[clintrubber]

from Rossi:

But I would actually prefer using the active circuit as an option, so an external box would be preferable.

I thought it was in this thread that the small Switchcraft-XLR-XLR-enclosure was mentioned.
Likely it'll be too small though, but it's easy to make a variation on it with two XLR-connectors & a piece of metal pipe in any length you want. That's how I housed the ESP#93 for some boundary mics (electrets).
It makes for a compact enclosure and could be clamped to the mic-stand.

On the other hand would a bit larger enclosure give you all options for components of course, but as said then you're actually/almost just pulling your mic-pre towards the ribbon-mic. I did fit a few film-caps in the tube i.s.o. electrolytics, but it's a puzzle.

Bye,

Peter

 

[Rossi]

As far as I'm concerned, I have no particular enclosure in mind, so I don't have a rigid space limit. I would prefer a small circuit, though, in case I do want to put it into the mic itself.

Does anybody think an instrument amp such as INA163 would be a good idea? Or regular opamps in an instrument amp configuration?

Another chip that interests me, and not just for the purpose at hand, is the OPA1632, a fully differential opamp. It's targeted mostly as a buffer amp for high performance AD converters with differential inputs. But this chip looks like you can do a lot more with it. It would be hard get it here in Germany, though. It's fairly new, and not even the better German online stores carry it. Well, looks like it would require too much current anyway. Still, an interesting chip.

 

[PRR]

> Does anybody think an instrument amp such as INA163 would be a good idea?

Uh.... for what?

I mean, what is the point of this thread? "Many ways to make a simple thing complicated"?

> potential phantom-damaging, why not make the ribbon-mic active?

Why not just be careful? In fact, since most ribbons need a rather different preamp than dynamics or condensers, and there is no reason to screw around with ribbons except "a sound", it really seems to me that you should pick a suitable preamp (withOUT phantom) and plug your ribbons ONLY into that preamp.

Ribbon mikes work GREAT if you just run them through less than 300 feet of plain mike cable into a good low-noise high-gain preamp.

Perhaps your problem is the Phantom Mike Infection? You have so many of these new-fangled German-style high-output needs-power mikes that it is dangerous to have any ribbons around?

FWIW: a properly wired ribbon and Phantom should not cause problems. But if it does, it is expensive.

As for the vaguely defined other issues:

A ribbon NEEDS a transformer. Extrapolating transformerless input designs to 0.2Ω leads to input device currents about 1 Amp for BJT, 10-100 Amps for tubes, and FET somewhere between. And even at that extreme, I don't think it would be practical. You gotta have a transformer.

However you don't transform to line impedance; you transform to whatever is convenient for an amplifier. Simply transforming 0.2:10K will allow a tube to run just 1mA with near-ideal noise performance. Even a selected TL072 will work with a manageable secondary impedance. BJTs can work great with lower impedances, reducing transformer compromises. Plain old 5534, or a single BJT with 100uA bias, will be very low noise with a 0.2:2K transformer.

Given that everything is together inside a metal case, and floating and transformered, I don't see ANY point in any differential-input amplifier.

For any phantom-powered mike, the output stage is the hard part. The maximum power we can take is about 24V and 7mA. This is about 3.4K. The load is two loads which may be as low as 100Ω each. Driving 50Ω from 3.4K means either a steep step-down transformer or not a lot of output voltage. The later approach may be tolerable because we do NOT want a large output voltage: anything with Phantom on the jack will max-out with less than 3V input. DC power available: 24V*7mA= 160mW, audio power out: no more than 3V^2/200Ω= 45mW. So the power is there, but not a lot to spare. It has to be used carefully. For example, a push-pull Class AB output making 3V in 200Ω without transformer will need 10VDC at 8mA. That's for unbalanced operation: simple balanced output needs 16mA. So we are looking at an unbalanced output, an audio transformer, or a nasty power conversion transformer. Or less than 3V max output, which is a very honorable choice (dynamics don't make more than 300mV in "musical" situations).

 

[clintrubber]

from PRR:

I mean, what is the point of this thread? "Many ways to make a simple thing complicated"?

It originally started as a simple quest for (1) preventing the phantom-accident and (2) giving such mics a bit higher output to make the use with preamps with less high gain possible. I guess since those Royer-active ribbons exist there seemed at least some sort of validation for an 'active ribbon'.

But indeed, it looks like it all comes with too many drawbacks (so one becomes more curious about what's inside the existing actives) and/or it's just putting more of the actual preamp towards the mic. Was hoping to get away with the ESP#93 that's already built but that turned out to be less suited. Haven't read all of this thread yet, but I must say it evolved into an interesting read.

it really seems to me that you should pick a suitable preamp (withOUT phantom) and plug your ribbons ONLY into that preamp.

In principle this is the safest idea. In practice I fear many people would be temped to want to use that nice preamp for other mics as well. And then the phantom-requirement might pop up again.

But I figure other things could work. While this might be cumbersome for certain situations, for me it'd not be impractical to give that attached ribbon-mic-wire a say, a XLR-4 plug and give that 'nice' DIYed mic-pre a ribbon-dedicated XLR-4 input as well. And make sure the XLR-4 can't ever get phantom-power (by an input-selection switch for instance: select between usual XLR-3 or XLR-4).

Regards,

Peter

 

[Rossi]

Thanks for stopping by, PRR.
I agree that a dedicated preamp would be the best and mot professional solution. My main concern, however, was improving the performance of the new budget ribbons (Nady, Thomann et al), especially when used with budget preamps. The budget ribbon (BR) requires about 10 dB more gain than a SM58 on an average preamp. Many budget preamps don't offer that much gain, become overly noisy and/or don't sound good when pushed to the limit. So the question is: can we make the BR a little louder with a phantom powered circuit or would the drawbacks (noise, sonic degradation, cost) outweigh the benefits (higher output, no phantom power accidents).

The circuit was meant to be post the microphone's transformer, I didn't mean to replace it.

Btw: Yes, the BR does survive phantom power (like any other ribbon wired correctly), I accidentally tested that. Still, there's always a moment of uncomfortable suspense & heartbreak, when stuff like that happens. Plus, on a lot of budget preamps you can't turn off phantom power per channel. So even while the BR *most probably* won't be fried, you don't dare to use it in combination with a condenser.

 

[clintrubber]

I accidentally tested that.

Great quote :grin: :wink: :thumb:

The budget ribbon (BR) requires about 10 dB more gain than a SM58 on an average preamp.

I had a look at the T-Bone & SM-57 figures and the were both some -55 (insert correct units). I guess the '58 is then also around this, but I understand from your message that the T-Bone must be giving less output. Understandable, but the the quoted sesitivity is wrong. Could imagine that - the quoted max SPL or what was it, 160 (insert correct units) seems a bit fantastic...

 

[Rossi]

I guess we all know by now that figures don't mean much, especially when it comes to budget equipment. The testing conditions for the respective figures may be different, or maybe it's simple inaccuracy. I have an SM58 and two Thomann RB-500s. Both Thomanns definetely require more gain than the Shure. As I said, about 10 dB. That's what you would expect from a ribbon anyway. A friend who has a Royer and now got two Thomanns as well tells me his Royer (a passive one) is a bit louder than the Thomanns.

 

[clintrubber]

Thanks Rossi, this hands-on info is indeed much more of use than those stated figures.

Thanks,

Peter

 

[PRR]

Getting a good noise figure, from 200 ohms, and good output drive, with only Phantom power, isn't an easy thing to do.

But if you allow a transformer it gets elegant:
{see below}

Adjust the resistor under the FET for about 2mA total current (1mA, 4mA, not at all fussy).

Gain with 200:2K transformer is 11.6dB. It might be expedient to use a 4-winding 150/600:150/600 transformer, one winding for input, 3 windings in series for output, for gain of 8.6dB.

In front of Full Orchestra, ribbon output will be like 20mV, and I estimate THD as about 0.01%, not rising rapidly at insane outputs. Yes, a ribbon will hang together at 160dB SPL, but I hope you are never in that studio, and if you are then gain is not your problem.

Noise will not degrade the performance of "budget", or even some "classy", mike preamps, though with a very-fine preamp there will be a little rise of noise.

Transitor type is not too fussy, though 100mA-400mA Switches might be a better choice than high-Beta low-current "low noise" transitors.

Performance is unchanged for 48V to 24V Phantom. It droops a bit at 12V behind 2*6K8 resistors, but that is not legal Phantom. (There is a P12 but it uses smaller resistors, and is not common in music use.)

 

[dayvel]

Or just use an instrumentation amp to build a dedicated preamp. Pretty easy when you don't have to fart around with the phantom power. The friendly folks at TI will give you the chips if you ask nicely.

 

[Rossi]

Thanks, PRR. I suppose that means we should forget about an active ribbon without an extra transformer (not counting the one already in the mic).

My impedance converter looked pretty similar but without the FET. I did try a step up transformer, too. But the only one I had on hand was a horrible POS and only 1:3. Not surprisingly, the sound was unusable, but the noise performance was good. I'll see if I can find an inexpensive transformer that is up to the task. If not, I'll better think about a dedicated preamp for ribbon microphones. No phantom power, higher than usual input impedance (or maybe two impedance settings), optimised for high gains, and a few goodies that often come handy with ribbons such as variable low cut.

 

[Rossi]

[quote author="dayvel"]Or just use an instrumentation amp to build a dedicated preamp.[/quote]

I don't think I can get free samples in Germany, but Reichelt carries the INA163 for about 5 Eurobucks. The through-the-hole INA217 would be easier to work with, but the 163 isn't too small to solder, I think. I was actually thinking about building a preamp with INA chips, anyway. I might as well add two ribbon optimised channels. The only problem I see is that the BB INA chips are meant for relatively low input impedances. If I remember correctly, optimal noise performance for INA163 should be @ about 1,2 K. The INA163 based Apogee Mini-Me is 1.4K, I think. But that may not really be a problem in practice.

 

[clintrubber]

Thanks PRR for the schematic.
You must have known that I had a pair of
these small Beyer 200:2k TR/BV-transformers around. :thumb:
They're 6H (@50Hz) i.s.o. 10 H prim inductance, but OK.

from PRR:

Transitor type is not too fussy, though 100mA-400mA Switches might be a better choice than high-Beta low-current "low noise" transitors.

Why would the switching-types be better suited ?

from Rossi:

Thanks, PRR. I suppose that means we should forget about an active ribbon without an extra transformer (not counting the one already in the mic).

I remember from another or this thread a remark about TXs in series (cascade).
Here we now could get three in a row: the internal TX, the added 200:2k and an eventual TX of the actual mic-pre.
Any funny things could happen ?

from Rossi:

But the only one I had on hand was a horrible POS and only 1:3.

As you'll know no problem here because of the 1:3 (ratio), since the suggested 200:2k is about just that.

Regards,

Peter

 

[Rossi]

Yes ratio / gain was quite okay. I would have liked a little more, maybe 1:4 or 1:5, but then again my impedance converter wasn't as well designed as PRR's and might have consumed some of the transformer gain.

But as I said, that transformer I have sounds horrible. I just saw that Neutrik offers small transformers for about 10 Eurobucks. There's a 1:4 (200:10K) that looks a bit more promising than the no name one I have. There's also a 1:3:10 Neutrik at about the same price as the 1:4. I'm not sure they're that good, though, as Neutrik offers better transformers at 35 bucks a piece. There must be a reason for that price difference.

 

[PRR]

> They're 6H (@50Hz) i.s.o. 10 H prim inductance

Ignore those H values. Simulated Henries are free so I use big numbers. The key thing is that it be designed as a good-response microphone (or 150Ω line) input and 1:3-1:5 ratio.

 

[PRR]

> about 10 Eurobucks. There's a 1:4 (200:10K)

That sounds a lot like the US$15 Radio Shack 1:7 XLR-1/4" tranny

I have seen the "same" thing under many brands in different lands; I'm sure they all come from the same factory with "Your Logo Here". Response is really quite good at low to medium levels.

If you want to be THAT "budget", try this:

Note that I even grounded one side of the transformer secondary, so you do not need to cut-open the XLR-1/4" adaptor.

I have not bothered to draw the output jack, but obviously the 6K8 resistors are in the mixing board or other mike-in.

Output is really unbalanced, but gain is almost 7, so line-crap should not be a problem.

Distortion models as some absurd number like 0.003%THD at 380mV out. The simulation is not good about transistor Vce. It will hard-clip at 0.6V peak output, 60mV RMS input, about 135dB SPL in a ribbon mike. Few sources really exceed this.

Noise should be very low. Build one with a socket, try a dozen $0.07 transistors, weed-out any that seem noisier than the rest.

 

[Rossi]

Thanks again, PRR.

The 1:4 neutrik transformer is not one that comes in a plug (as the name Neutrik might suggest), it's really just a transformer. Certainly not a great one, but I thought it might do the trick. It's not that I want to scrooge around as much as possible. I just think that at a certain point it's probably wiser to invest that money into a better preamp.

Here's the Neutrik 1:4
and the Neutrik 1:3:10

Hope those links work alright.

 

[bcarso]

Back from visiting my frail mother, and I see the exchanges have been lively.

The limiting voltage noise parameter in a bipolar is rbb', the so-called base spreading resistance. The other term in the voltage noise goes down for a while as you increase collector current (hence PRR's statements about how much current one would need to make something that is a satisfactory direct ribbon preamp input---yes, 0.2 ohms is insanely low to do better than source thermal noise in the preamp, without a transformer).

There are some transistors that have small rbb' sometimes as a consequence of making a good switch---the prime example being the old Motorola (and I hope still, On Semi) 2N4403. Toshiba 2SA1015's are about rbb' = 30 ohms, and there are some lower value parts. And then you can always parallel parts although to do that directly can require some tedious matching.

So-called "low noise" transistors are often for 10k-ish source Z's and optimize e sub n and i sub n together. They don't sometimes work very well for low source Z's.

Brad