Microphone boosters?

[JohnRoberts]

Reading what I wrote earlier "For me, it shows that your Claret is either poorly designed or duff.", one may think it is statistically not very likely.
Actually many mic preamps do not perform as they should because their input transistors have been subjected to zenering of their base-emitter junction. This happens when connecting an unbalanced source when phantom is on. The preamp seems to perform normally after that, except that noise has suddenly increased.
There are well known solutions, but many designers are not well aware of them, particularly because they don't understand the problem.

I don't know how common this is, but the well known practice of using clamp diodes to prevent the b-e junction reverse biasing, is fairly widely seen. According to Motchenbacher and Fitchen's text on low noise design (where I first read about this phenomenon on the 70s) a low noise junction made noisy because it zenered, can also self-anneal (get quiet again) over time.

Accidentally shorting mic inputs with fully charged phantom voltage blocking capacitors can dump amps of current into inputs, causing more harm than a little hiss.

Now, back to your question "If I have a good quality pre-amp that provides 55db of gain, but I need 65db of gain in order to achieve the signal level that I need, what would you suggest?"
There's only one simple answer: add a line-level gain stage. I don't know of any commercial product doing just that, so you would have to build one, or use a standard mic pre that accepts a line level. It would not need to be particularly low-noise.

Alternatively you may use a 1:3 step-up xfmr inserted between the mic pre's output and the converter's input. Assuming an input impedance of 10-20k on the receiver, this would reflect as 1-2k on the primary side, which most mic-pre's can handle reasonably.
I just mention this possibility in order to leave no stone unturned, but don't advocate it.

Another alternative is to modify the 55dB gain stage to deliver 65dB. Depending on the topology this may introduce performance compromises, or not. YMMV.

JR

 

[zmix]

Back in the early 80s I did a quad gate limiter (LOFT/Phoenix audio lab). In that circuit I put the OTA in the negative feedback path of a simple inverting op amp. Again the OTA was only affecting the audio path when limiting or gating, so the un-gated audio path was as clean as a unity gain op amp (very clean).

I do not recommend OTAs for serious dynamics work, but even then the sonics will typically be dominated by the side chain manipulations that effectively multiply the audio signal even when using a perfect gain element.

I have a rare guitar pedal from the early 1970s called the "Low Frequency Compressor", made by Electro-Harmonix, it uses a CA3080 in the feedback loop of an opamp. At first glance this seems to be the best way to reduce the inherent gremlins of the 3080, and I have always been intrigued by this design, as it's the opposite of the implementation in the MXR Dyna-Comp, which uses the 3080 inline.
This requires the sidechain to work inversely to the MXR, and it has a characteristic that you elude to above, it seems to only produce noise when limiting, which gives the guitar / bass an interesting puff of fairy dust on each note.

Might perform better with a new layout, though..

 

[JohnRoberts]

I have a rare guitar pedal from the early 1970s called the "Low Frequency Compressor", made by Electro-Harmonix, it uses a CA3080 in the feedback loop of an opamp. At first glance this seems to be the best way to reduce the inherent gremlins of the 3080, and I have always been intrigued by this design, as it's the opposite of the implementation in the MXR Dyna-Comp, which uses the 3080 inline.
This requires the sidechain to work inversely to the MXR, and it has a characteristic that you elude to above, it seems to only produce noise when limiting, which gives the guitar / bass an interesting puff of fairy dust on each note.

Might perform better with a new layout, though..


View attachment 86181

using a 3080 OTA for a guitar compressor is probably a premium design, it was more common to use simple JFET limiters. Using an OTA provides a better defined gain control law, than crude JFET limiters.

JR

 

[abbey road d enfer]

What Id like to know is how do you design an op amp based output stage that takes its power and output signal over the same two (or three including ground)wires

That's the easy part. See attachement.

, it would also need to pass power upstream on xlr input pins etc.

Not too difficult either, but it could only transmit what is left after P48 has passed through the resistors. Whatever the circuit draws is what may be missing for teh mic.

. I see on certain op amp spec sheets 'rail to rail' capable , is that the type we need ?

R2R is important when voltage is scarce, typically opamps operating under single 5V rail or less. Most R2R opamps do not operate under +/-15V.

If the gain between the head amp and op amp driver could be adjusted remotely from the PSU over the same three wires we'd have our cake and eat it . Ive heard Bo talking about this remote idea before with control signals transmitted over the audio cable . We only need two functions , volume up/down , could the DC supply voltage itself be momentarily ramped up a couple of volts to send a signal to increase volume and down a few volts down for a decrease ? What different devices could we use to control gain and how many steps would we need ? Seeing as were likely to get fine gain control over the source at the mixer we connect it to, maybe even five steps of 10db each would suffice , going to the trouble of making a fully adjustable VCA based thing is over the top, could a simple switched resistive network with solid state relays maybe 5 or 10 positions of attenuation be adequate?

Basically you're describing what "digital" mics do under AES42 protocol, which uses simultaneous upstream and downstream digital flux. Don't forget you want to be able to monitor the mic status, so you need some feedback.
There have been some attempts at gain remote control via P48, but the necessity to conform to the standard (48 +/-4) does not leave much latitude.

What happens if the mic is accidently disconnected and reconnected I hear you say ?, no worries the we could just make so the power supply tells the mic the config of the relays on power up ,and bam your gain structure is preserved ,

I alaway have a smile when I see a sentence with something like Boom! or Hey presto! Usually ends up in a can o' worms...

 

[abbey road d enfer]

I don't know how common this is, but the well known practice of using clamp diodes to prevent the b-e junction reverse biasing, is fairly widely seen.

Nevertheless, there are a number of older mixers that are not adequately protected, and I suspect a number of interfaces too. Not all designers have read the seminal papers by Gary Hebert et al.

According to Motchenbacher and Fitchen's text on low noise design (where I first read about this phenomenon on the 70s) a low noise junction made noisy because it zenered, can also self-anneal (get quiet again) over time.

I recently read that serious studies showed that they never recover their previous performance, at least on a human time scale.

 

[Bo Deadly]

I have a rare guitar pedal from the early 1970s called the "Low Frequency Compressor", made by Electro-Harmonix, it uses a CA3080 in the feedback loop of an opamp. At first glance this seems to be the best way to reduce the inherent gremlins of the 3080, and I have always been intrigued by this design

The "Engineers Thumb" compressor does that. From merlin who posts here once in a great while.

as it's the opposite of the implementation in the MXR Dyna-Comp, which uses the 3080 inline.

I actually have one of these (one with a CA3080). It's is noisy as heck. But it's only one of five pedals I own (that, the Rocktek Chorus and 3 Germ fuzzes) so it must have kept it for a reason.

 

[JohnRoberts]

Nevertheless, there are a number of older mixers that are not adequately protected, and I suspect a number of interfaces too. Not all designers have read the seminal papers by Gary Hebert et al.

I recently read that serious studies showed that they never recover their previous performance, at least on a human time scale.

I was just repeating what I read decades ago, I have used clamp diodes across such low noise junctions since learning about their utility.

I have never experienced semiconductor junctions that get quieter over time myself, but I have heard phoolish sounding anecdotes about exotic SKUs that must be burned in for several days before serious listening.

I always look to mine out the tiny nuggets of reality behind popular audiophool memes. I rarely find anything of substance.

JR

 

[Newmarket]

I actually have one of these (one with a CA3080). It's is noisy as heck. But it's only one of five pedals I own (that, the Rocktek Chorus and 3 Germ fuzzes) so it must have kept it for a reason.

Ha! I love my simple Rocktek Chorus pedal on bass. In fact I have two Sort of a one sound thing. Depth Up / Tone Up / Rate Slow. But I love it !

 

[Newmarket]

. I see on certain op amp spec sheets 'rail to rail' capable , is that the type we need ?

You don't want to be looking at R2R devices (of any flavour and there are various flavours) for audio applications if looking for fidelity.

 

[JohnRoberts]

You don't want to be looking at R2R devices (of any flavour and there are various flavours) for audio applications if looking for fidelity.

I will assume you are not talking about R2R ladders used for old A/D conversions.

There are some modern CMOS opamps with rail to rail output and modest current consumption.

Read the data sheets.

JR

 

[Tubetec]

Thanks for the op amp diagram Abbey ,


Just as a test of concept with off the shelf gadgets I was thinking of putting together what you see below ,

Head amp ,designed for electret mics but might have enough gain for higher-z dynamic voice coils or Ribbons with step up , runs off 48volts on the XLR , looks like only a few fets or transistors involved , plug in power for the electret capsule might need to be disabled

Volume/gain control not nessesairily a pot .

OPA 1622 Op amps

Coupler to house the electronics

Off the shelf phantom supply modified for extra current or other voltage as required .

 

[Cranehazard]

I was taught that the noise floor of any system is dominated by the first stage. The fets in the boosters have some really low noise. If you use it with a dynamic mic you are exploiting this effect. If you use it with a condenser, the noise will still be dominated by the mic circuit ( it's a mic preamp) in the condenser. I also belive the tonality and response of a system is dominated by the first stage. When you add the fets in the booster to a instrumentation preamp in the interface you are making a preamp with a discrete front end farther down the cable (which doesn't seem to matter). Those fets are super quiet that why it works. If the preamp already has a discrete front end my hypothesis is the mic booster won't help with the noise much.

 

[Newmarket]

I will assume you are not talking about R2R ladders used for old A/D conversions.

There are some modern CMOS opamps with rail to rail output and modest current consumption.

Read the data sheets.

JR

Boom Boom (Basil Brush style + maybe a British thing ?...)
Yes CMOS rail to rail OpAmps also tend toward Low Power. Just not optimised for low frequency AC (aka Audio) precision.

EDIT / Addendum

As it happens, in the day job I just came across this Rail to Rail OpAmp that goes to +/-20V.

https://www.ti.com/product/OPA2991

 

[Newmarket]

Many audio interfaces have limited gain, because they are so often used for close mic'ing. With current 24-bit converters, a weak recording can be bumped up digitally without too much harm.

Right. Got it! tbh I wasn't thinking of interfaces with mic pre. Although I do have a TDIF interface with mic inputs but I've never actually used it.

 

[abbey road d enfer]

As it happens, in the day job I just came across this Rail to Rail OpAmp that goes to +/-20V.

https://www.ti.com/product/OPA2991

Yes, my comment was addressing the majority of R2R opamps, but tehre are some oddities.
I noticed the rather high input noise voltage (10nV/sqrtHz), which usually goes with low power drain (<600uA).
I'm not sure I'm tempted to use it. The 42V technology is not enough to tempt me.

 

[ruffrecords]

I was taught that the noise floor of any system is dominated by the first stage. The fets in the boosters have some really low noise. If you use it with a dynamic mic you are exploiting this effect. If you use it with a condenser, the noise will still be dominated by the mic circuit ( it's a mic preamp) in the condenser. I also belive the tonality and response of a system is dominated by the first stage. When you add the fets in the booster to a instrumentation preamp in the interface you are making a preamp with a discrete front end farther down the cable (which doesn't seem to matter). Those fets are super quiet that why it works. If the preamp already has a discrete front end my hypothesis is the mic booster won't help with the noise mucrh.

There is really no such thing as 'super quiet' The bottom line is most mic pres get within 3dB of perfection when fed from a 150 ohm source. This means that no matter how 'super quiet' a cloudlifter may be, even if it had no noise at all it would only improve the noise by 3dB. Mic pres these days are so good that the dominant noise source is the microphone itself.

Cheers

Ian

 

[Newmarket]

Yes, my comment was addressing the majority of R2R opamps, but tehre are some oddities.
I noticed the rather high input noise voltage (10nV/sqrtHz), which usually goes with low power drain (<600uA).
I'm not sure I'm tempted to use it. The 42V technology is not enough to tempt me.

Indeed. Previously I've mainly used them in 5V single supply mode for industrial controls with DC levels.
There's often a difference in input bias currents when operating toward the rails. I wasn't aware of any operating above the 'standard' 36V voltage span but then again I haven't been looking for it.
As it happens it came to my attentioin through an EDN email link for AB amplifier circuit.
https://www.edn.com/class-ab-20-w-c...EDNAnalog-20211111&oly_enc_id=2682G4238956J8W

 

[JohnRoberts]

There is really no such thing as 'super quiet' The bottom line is most mic pres get within 3dB of perfection when fed from a 150 ohm source. This means that no matter how 'super quiet' a cloudlifter may be, even if it had no noise at all it would only improve the noise by 3dB. Mic pres these days are so good that the dominant noise source is the microphone itself.

Cheers

Ian

I fell down the rabbit hole of quantifying ein (equivalent input noise) numbers as voltages decades ago. I found it more comparable to quantify mic preamp noise performance as NF (noise figure). Simply stated NF is a simple comparison of a real preamp to a perfect (theoretical) preamp. 0dB NF is perfect, no added noise to the source's noise.

To echo Ian's observation, most modern preamps will deliver a NF of 3dB or better. It is increasingly difficult and expensive to engineer out those last remaining dB. My best efforts were probably in the 1-2dB NF range. Then consider how audible would a fraction of dB noise floor improvement be (not very IMO).

JR

 

[Cranehazard]

There is really no such thing as 'super quiet' The bottom line is most mic pres get within 3dB of perfection when fed from a 150 ohm source. This means that no matter how 'super quiet' a cloudlifter may be, even if it had no noise at all it would only improve the noise by 3dB. Mic pres these days are so good that the dominant noise source is the microphone itself.

Cheers

Ian

I have heard this myth about mic impedance dominating the noise floor but I just don't believe mic pres in interfaces are as good as you say and that's why so many people are buying mic boosters.

 

[emrr]

I have heard this myth about mic impedance dominating the noise floor but I just don't believe mic pres in interfaces are as good as you say and that's why so many people are buying mic boosters.

I tend to think the preamps are generally that good, but polluted by their environment - USB bus power, etc. Net result the same.