Mic Preamp -- Is This a Silly Idea?

[Rossi]

There are no values in your schemo, but I'm pretty sure this is going to be too noisy. You'll get extra noise from the input resistor and most opamps don't have a spectacular noise performance at low input impedances, so this circuit will probably add more noise than is acceptable for a low output dynamic mic which is likely to require a lot of gain (which is why we are interested in a pre-preamplifier in the first place). PRR's ribbon booster (in the old active ribbon thread) seems more attractive. I've built it in some variations, and I can only recommend it to anyone dealing with low output dynamics and ribbons.

John, the 2SD786 is long out of production (as you probably know), the 2SC3329 was an more than adequate replacement, but Toshiba seems to have stopped production. The best I could find that is still available is the 2SC2546 from Hitachi.

 

[chris319]

PRR's ribbon booster (in the old active ribbon thread) seems more attractive. I've built it in some variations, and I can only recommend it to anyone dealing with low output dynamics and ribbons.

I've seen that discussed here but am I correct in saying that it only provides < 15 dB of gain?

 

[JohnRoberts]

The gain as drawn can be varied by value of total resistance between the two emitters of LTP, working into the collector load.

I would be careful about increasing gain and too much signal swing at collectors, while keeping emitter reistors small is useful for low noise.

Those values are a good starting point.

JR

 

[JohnRoberts]

[quote author="Rossi"]

John, the 2SD786 is long out of production (as you probably know), the 2SC3329 was an more than adequate replacement, but Toshiba seems to have stopped production. The best I could find that is still available is the 2SC2546 from Hitachi.[/quote]

Yup, I recall being told by the ROHM rep 8 years ago or so. That just shows how long it's been since I've done a low noise design, but the physics hasn't changed, and those old parts were still available in DIY quantities from web merchants. I still have a bunch of 2SB737Ss in my private stash, should I decide to dabble.

I am most interested in eliminating the electrolytic blocking capacitors and/or transformers from the low level path between mics and preamps (along the lines of Wayne's work but discrete with a few twists).

Upon reflection this may be moot as most phantom mics have such components in their audio path and non phantom mics don't need the supply at all, so such blocking caps "could" be bypassed, while they routinely aren't.

JR

 

[Guest]

[quote author="Rossi"]There are no values in your schemo, but I'm pretty sure this is going to be too noisy. You'll get extra noise from the input resistor and most opamps don't have a spectacular noise performance at low input impedances, so this circuit will probably add more noise than is acceptable for a low output dynamic mic which is likely to require a lot of gain (which is why we are interested in a pre-preamplifier in the first place).
[/quote]

It was my guess about "What is inside".

 

[emrr]

Anyone have a link to the noise specs on those devices? I recall seeing them for both the Sanken and the Rhode devices of this nature, and the noise performance was not so great. I'm remembering -110-115 db EIN numbers, but could be wrong. Maybe fine for the needs in this case.

 

[Rossi]

[quote author="chris319"]
I've seen that discussed here but am I correct in saying that it only provides < 15 dB of gain?[/quote]

I think it has 10 or 12 dB gain, which is sufficient IMO. You still have the gain from your preamp, and your preamp's noise performance is likely to be better at high gains. So a 10 dB booster + 50 dB preamp gain is likely to be lower noise than a 40 dB Booster and 20 dB preamp gain. Also, if your booster has a lot of gain, you have to remove it for loud sources whereas a 10 dB booster will still be usable.

 

[Rossi]

[quote author="JohnRoberts"]

I am most interested in eliminating the electrolytic blocking capacitors and/or transformers from the low level path between mics and preamps (along the lines of Wayne's work but discrete with a few twists).

Upon reflection this may be moot as most phantom mics have such components in their audio path and non phantom mics don't need the supply at all, so such blocking caps "could" be bypassed, while they routinely aren't.

JR[/quote]

I may be wrong, but I suppose the P48 blocking caps are not bypassed for dynamic mics because of voltage offsets at the input.

There was an early FET based version of PRR's ribbon booster that does not require input caps, maybe that would be interesting to you. I'm not sure you can get as low noise as from a BJT input stage at such low impedances, though.

 

[chris319]

your preamp's noise performance is likely to be better at high gains.

Funny thing about the mic trims on the FR-2LE. Listening on headphones with the mic trims full CW, there is a certain level of noise. Back the trims off slowly and at a certain point the noise drops abruptly, and I mean abruptly. This leaves the trims at about the 4:30 position. At that setting the background noise is substantially lower but it now requires 6 mV instead of 3 mV to reach 0 dBFS.

I think it has 10 or 12 dB gain, which is sufficient IMO.

With the gains backed off as described above and if you're shooting for peaks at -9 dBFS, 12 dB will increase the working distance of a typical dynamic mic on conversational speech from 1" to 4". If you're covering a news conference with an RE50 the audio will be down in the mud.

if your booster has a lot of gain, you have to remove it for loud sources

Not a problem.

 

[JohnRoberts]

[quote author="Rossi"]

I may be wrong, but I suppose the P48 blocking caps are not bypassed for dynamic mics because of voltage offsets at the input. [/quote]
That's why I always left them in my designs

There was an early FET based version of PRR's ribbon booster that does not require input caps, maybe that would be interesting to you. I'm not sure you can get as low noise as from a BJT input stage at such low impedances, though.

It has been years since I looked, but last I checked bipolar devices were 6dB or so quieter than JFET, and the few very low noise JFETs were obscure military parts.

------

I planned to use servos to deal with both DC error at input and DC offset at gain pot.

I still have questions regarding balance of real world phantom current draw and impedance balance from different mics. Blocking caps at the preamp input allow for a world of sins. I was glad to see Wayne looking at such things. This is mostly a mental exercise on my part.

JR

 

[Guest]

It's still a silly idea, to add one more inconveniently powered preamp to the main preamp that can be made without power and space limitations.

 

[Rossi]

[quote author="JohnRoberts"]
I planned to use servos to deal with both DC error at input and DC offset at gain pot.

I still have questions regarding balance of real world phantom current draw and impedance balance from different mics. Blocking caps at the preamp input allow for a world of sins. I was glad to see Wayne looking at such things. This is mostly a mental exercise on my part.

JR[/quote]

If you want to avoid both transformers and lytics, the easiest way would perhaps be to go for a relatively high input impedance (say 10 k) so you don't need that much capacitance, which would make polypropylen caps feasible. Still caps, of course, but hardly noticeable as such.

 

[JohnRoberts]

[quote author="Rossi"]

If you want to avoid both transformers and lytics, the easiest way would perhaps be to go for a relatively high input impedance (say 10 k) so you don't need that much capacitance, which would make polypropylen caps feasible. Still caps, of course, but hardly noticeable as such.[/quote]

Yes better for the caps, but now you still have to deal with relatively large physical structures (film caps) hanging off the input side, and the input noise current will be working into the higher capacitor reactance at LF in series with mic source impedance.

I guess it depends somewhat on the noise spectra of the input devices and cap will be OK at higher frequencies where our ears are more sensitive, but not optimal IMO.

I would still want to HPF the signal but preferably using that film cap after the high gain has been accomplished.

FWIW, it's not a bad idea to go ahead and make the resistors on the bases as large as you can tolerate for dc offset current and build your nominal termination impedance on the mic side of the blocking caps. That way the electrolytic aren't working as hard., and could deliver better CM at hum frequency if pole is lower.

JR


.

 

[chris319]

Now there is this:

http://www.sanken-mic.com/english/acce/pdf/had-48.pdf

Not such a silly idea.

 

[bcarso]

Is this Sanken affiliated with the semiconductor company?

 

[chris319]

[quote author="bcarso"]Is this Sanken affiliated with the semiconductor company?[/quote]
Doesn't seem to be. A lot of people like their lapel mics, BTW.

 

[Junction]

[quote author="chris319"][quote author="Junction"]I have one of these at the studio, haven't used it, but I can have a look at it next week and let you know what on the PCB, it is a surface mount IC, thats all I remember at this stage.[/quote]
That would be great, thanks![/quote]

The Rode D Power units have an Analog Devices AD620AR 8 pin surface mount single op-amp, can find a data sheet at the following;

http://www.analog.com/UploadedFiles/Data_Sheets/37793330023930AD620_e.pdf

The PCB is 1.25" x 9/16" or 32mm x 15mm, and hard soldered directly to the xlr insert so I'm quite sure it would fit inside just about any microphone, even a pencil mic. There are 4 largish caps 2 x 22uF 63V and 2 x 100uF 16V (surface mount). The rest of the components are miniature surface mount, can't even read the values with a magnifying glass.

Rode don't seem to provide any decent specs, other than 20dB gain that's it! I'm not sure how they achieve a balanced output from a single op-amp???

Michael

 

[clintrubber]

Thanks for the info.

[quote author="Junction"]I'm not sure how they achieve a balanced output from a single op-amp???

Michael[/quote]
Perhaps it's impedance balanced ? (so no neg-polarity signal, just providing equal source impedances for 'pos' & 'neg')

 

[Junction]

Oh, I found some specs with the D Power Plug, the external plug in version;

Gain: +20dB
Input Level (nominal): 1mV
Output Level (nominal): 10mV
Freq Resp: 20 - 20KHz (-3dB)
Input Z: 100K Ohm balanced
Output Z: 100 Ohm Balanced
Noise: -116dB A-Wt RTI
Max Input: 0.8 Vp-p
Max Output: +11.4 dBu @ 1% THD (8Vp-p)
Current Consumption: 3mA @ 48V

Michael

 

[Guest]

What about an average temperature in the clinic, ... sorry, THD curves? How low impedance do they get from 6.8K resistors powering this device to provide low enough distortions? And what about some more stages inconveniently powered between a mic and a mic preamp that was already designed for such a high gain?

I would never design such a thing, even if it is well marketed commercial project...