Boosting dynamic mics with phantom power

[airon]

Has anyone used one of both of these two products, the Cloudlifter CL-1 Mic Activator and the TritonAudio Fethead ?


The Fethead is half the price of the Cloudlifter unit. I can see the Cloudlifter CL-1 being more rugged, but for static environments, I'm leaning towards the TritonAudio unit.

I'm thinking about acquiring an SM 7b with one of these units, since the mic is barely above 1mV/Pa. But it's terribly practical and sounds good for speech. My preamp is a Fireface UC, so I'm not 100% sure I even need a booster like this.

I'm currently researching both. What is the experience with these ?

[cyrano]

A Fireface has plenty of gain for an SM7b...

That said, all these pre-preamps are the same design. There are differences, but you'll find way more difference because of the impedance of the preamp that follows.

And coupling them to a preamp is like a game of dice. Nobody seems to understand exactly why there's so much difference sometimes in S/R ratio and even frequency chart. Impedance is a strange fellow...

[numberthirty]

I bought a Cloudlifter just to use with an SM7. To me, it was absolutely one of those "Why in the world didn't I do that sooner?" purchases. The enclosure has also held up really well in situations where bonehead moves were not a rarity.

[SonicAxiom]

thread title is somewhat misleading, airon. I already considered coming up with a similar title on april 1st



.

[multibody]

Originally Posted by airon

Has anyone used one of both of these two products, the Cloudlifter CL-1 Mic Activator and the TritonAudio Fethead ?


The Fethead is half the price of the Cloudlifter unit. I can see the Cloudlifter CL-1 being more rugged, but for static environments, I'm leaning towards the TritonAudio unit.

I'm thinking about acquiring an SM 7b with one of these units, since the mic is barely above 1mV/Pa. But it's terribly practical and sounds good for speech. My preamp is a Fireface UC, so I'm not 100% sure I even need a booster like this.

I'm currently researching both. What is the experience with these ?

I have the exact same signal chain (sm7b-fethead-fireface uc)

I am quite pleased, the signal has low noise and is a charm to work with. I also like the sound but never compared it much to anything else.
I remember that some people were talking about the importance of preamp-impedance for the sm7b. So the variable impedance of the cloudlifter might be an advantage.

[karbomusic]

It's the CL-Z that has variable Z and I can't find the input impedance for the CL-1. The long and short of it is that as long as the input impedance is high enough, that's all that is needed. FWIW, the fireface already has a 2k input impeadance which is the max of the CL-Z. I don't see any reason the UX can't handle the SM7b by itself but the extra gain is nice.

They throw JFET around like it's magical but AFAIK that's really just about the fact JFETs naturally have a higher input impedance (vs something that is BJT based) and work better for such circuits (aka if it were to saturate some, it's prettier to the ears than BJT). I have a CL-1, it's handy and I sometimes use it for my Royer 121s; more for extra protection against phantom power than actually needing the gain. I've also used my Royer's without the CL-1 and I don't remember any difference that caught my attention since my existing pre's are up to the job already.

I think it's a fine piece of gear and the extra gain is nice, but if the preamp is already good enough it's a nice to have more than critical component.

[Icchan]

In my understanding these products are mainly for live-use where long cable lengths can be an issue and you want more gain as close to the source as possible so that there won't be signal to noise ratio problems when the signal finally reaches the mixing console?

Though, these days that's not an issue anymore since most things use digital signalling anyway and the AD-converters and mic pre-amps are really close to the stage even when the FOH-mixing desk is 100m away ;D

It's true, that some sources might be somewhat low sensitivity, like ribbon mics, and there I can understand that not everyone has high end mic-pres that have the gain to work with them properly.

since they can't take loud sources without breaking, you need a lot of gain and very high impedance microphone pre-amps...

[EpicSounds]

I have the Triton FETHead to use with SM7B. Works pretty well, super small, does the job but increases noise slightly. The cloudlifter is better but needs an extra cable.

I don't currently use it with my Arturia Audiofuse, but that has a built-in boost.

We reviewed both on the home recording show years ago

http://www.homerecordingshow.com/201...view-and-more/

http://www.homerecordingshow.com/201...using-livecut/

[cyrano]

Originally Posted by karbomusic

It's the CL-Z that has variable Z and I can't find the input impedance for the CL-1. The long and short of it is that as long as the input impedance is high enough, that's all that is needed.

That's what I figured too.

But it isn't. There's a thread on GroupDIY that's been going on for 15 years about the concept. And some tests have shown that typical low noise BJT's with a high beta don't work too well. FET's are better. And to my amazement, the old venerable AC187/AC188 pair fares very well. You'd expect an old germanium power transistor not be suited because of it's noise character, would you?

There's a lot of difference between mics with a transformer and trannie-less mics. And both source and destination impedance influence S/R and even gain.

I only found this thread a few days ago, but it explains why I got very different results when I tested a few of these pre-preamps years ago. But then, I only tested with two preamps.

It's hard to tell what the outcome will be in these circumstances.

They throw JFET around like it's magical but AFAIK that's really just about the fact JFETs naturally have a higher input impedance (vs something that is BJT based) and work better for such circuits (aka if it were to saturate some, it's prettier to the ears than BJT). I have a CL-1, it's handy and I sometimes use it for my Royer 121s; more for extra protection against phantom power than actually needing the gain. I've also used my Royer's without the CL-1 and I don't remember any difference that caught my attention since my existing pre's are up to the job already.

I think it's a fine piece of gear and the extra gain is nice, but if the preamp is already good enough it's a nice to have more than critical component.

That's where the design on GroupDIY started. Protect ribbons from phantom power and provide some extra boost. But it never came to be a universal solution, because nobody understands the behviour of the circuit. And not all FET's are suited. Some unexpected candidates behaved well, while others that are commonly used in condenser mics dailed. You need a transistor with low rBB and there aren't that many around these days.

[pipelineaudio]

We had a argument about these as an aside on the Reaper facebook group yesterday and I did try and cover it on the livestream last night

https://youtu.be/n7yIXjgxZ8Q

I was convinced for years that these were scams, and in theory they still are AFAIK, but in practice, they might actually be of benefit given current mic preamp technology

Also, in the case of line noise itself, I could see the cloudlifters and such doing good things, even when compared the the old school blasphemy of daring to put one amplifier after another

I'm going to order one of these or the fethead and see how it is

[karbomusic]

Originally Posted by cyrano

That's what I figured too.

But it isn't.

In relation to what? What I mean is that I'd expect 10-100 times the mic impedance to be more than ideal from an impedance matching perspective (minus other complications). However, I'm sitting here staring at a $2000.00 mic pre I own, it's variable Z only goes to up 1.5k. So "in relation to what means" we've obviously been recording pretty well over the last 60 years, seems like that'd be hard to miss or just give it a high enough impedance to begin with - or we are splitting gnat hairs which is where I lean on this.

There's a thread on GroupDIY that's been going on for 15 years about the concept. And some tests have shown that typical low noise BJT's with a high beta don't work too well. FET's are better.

I'm not sure where the magic is there though. It's pretty common knowledge JFETs usually give you Hi-z where BJT doesn't; JFETs vary, they get measured and selected if it matters. None of this is a disagreement btw, just thinking out loud but... I just don't really buy into the variable Z unless our goal is modify the response, sufficiently high, isn't going to give someone something that doesn't actually exist.

[karbomusic]

Originally Posted by pipelineaudio

I'm going to order one of these or the fethead and see how it is

I have no problem with them and I like my CL1; I think they can help solve problems, if the problems actually exist (noisy pre etc.). Not so sold on needing one just because you have a ribbon or low output mic and angels will sing, not my experience anyway but I do also already have great preamps (UFX, NEVE, API, A-Designs) so no real chance to test on something cheap.

I do have an $89 M-Audio card I use in my electronics lab, I could always hook it up and play with my ribbons to see if the CL-1 changes anything.

[vdubreeze]

Originally Posted by Icchan

In my understanding these products are mainly for live-use where long cable lengths can be an issue and you want more gain as close to the source as possible so that there won't be signal to noise ratio problems when the signal finally reaches the mixing console?

Though, these days that's not an issue anymore since most things use digital signalling anyway and the AD-converters and mic pre-amps are really close to the stage even when the FOH-mixing desk is 100m away ;D

It's true, that some sources might be somewhat low sensitivity, like ribbon mics, and there I can understand that not everyone has high end mic-pres that have the gain to work with them properly.

since they can't take loud sources without breaking, you need a lot of gain and very high impedance microphone pre-amps...

I wouldn't say they're mainly for live use. I would say mainly for studio use, where you can't get enough level without raising the noise floor due to a low output mic and/or a very quiet source with/or a preamp that gets too noisy at the top of its range. If the mics aren't low output and the source is not soft and the preamp is quiet enough at high settings for the kind of music being recorded then they're not needed. But otherwise they're very useful. I don't personally know anyone using them live but maybe folks do.

[karbomusic]

where you can't get enough level without raising the noise floor due to a low output mic

Can you help me out a little... I'm at a loss as to how that doesn't bring the noise floor up with it and is one of my qualms about what are essentially OpAmp preamp circuits in little metal box. That's not knocking them since most every preamp on the planet is just that, just that the noise floor comes up with it unless the noise is in the line (which is unrelated to the mic), which it shouldn't be there either for reasonable runs and good cables.

[pipelineaudio]

Thats my issue too. In theory, the cloudlifter would ALWAYS be noisier for a given output level at the mic preamp output itself

But I think people are claiming that at the top end of some mic pres, the mic pre gets noisy in a way past linearity and that the cliudlifter can boost it ahead of time enough that even though it adds its own noise, it keeps it out of the linlinear noise area of the mic preamp and therefore is less noise

That at least, is the closest I can make to what I hear being claimed

Also, there is the separate issue of possibly lessening the transmission line noise by having a higher original signal, but I dont know

[karbomusic]

OK, I pulled my apart, it's basically what I expected (no shocker really, like I said, much of this is all the same science)....



So the preamp is two Linear Systems LSK389 opamps. That choice is likely due to being small, inexpensive, low current and ultra low noise. They run about $3 USD each so the majority of the expense is the enclosure, connectors and construction which is usually the case. As you can see it's a handful of tiny SMD parts, no rocket science, just a low noise pre with 20 something dB of gain. So for that PCB and the parts on it, we would be lucky to hit $6.00 USD if buying in bulk - the connector and properly shielded enclosure costs a bit more.

Again, not knocking it - I love mine. But if I sat down and designed something similar, I'd be under the same constraints and end up at similar prices. I just wanted to point out that amplifying a signal at it's heart is not a mojo process.

[karbomusic]

Originally Posted by pipelineaudio

Thats my issue too. In theory, the cloudlifter would ALWAYS be noisier for a given output level at the mic preamp output itself

If you mean it will always add it's own internal noise, that's a good point except it is probably low enough that it's still an improvement over whatever needs fixing, if something needs fixing.

But I think people are claiming that at the top end of some mic pres, the mic pre gets noisy in a way past linearity and that the cliudlifter can boost it ahead of time enough that even though it adds its own noise, it keeps it out of the linlinear noise area of the mic preamp and therefore is less noise

Right, that's what I meant earlier by fixing a problem if it exists. Some preamps may very well not do as well at the top 1/3 of their gain, especially an interface, simply because it is surrounded by so much other circuitry including digital (which can be difficult to keep from leaking noise across the ground planes). This is a problem worth solving, if you have that problem.

An aside is that they have an advantage the interface makers don't have. They don't have that noisy circuitry and tiny real estate to deal with so they get some of this for free by default.

Just trying to kindly buffer expectations, it's great to have but still just a simple opamp pre, though a better one than some interface companies are going to choose to save money.

[karbomusic]

One last thing, it has 4 jumpers on it. I didn't take the time to trace the circuit so I wonder if those allow something like passing through the phantom power or lifting a ground or similar. Could also just be something simple like changing jumpers or them becoming connectors when they are using that same board in other CL products.

[vdubreeze]

Originally Posted by karbomusic

Can you help me out a little... I'm at a loss as to how that doesn't bring the noise floor up with it and is one of my qualms about what are essentially OpAmp preamp circuits in little metal box. That's not knocking them since most every preamp on the planet is just that, just that the noise floor comes up with it unless the noise is in the line (which is unrelated to the mic), which it shouldn't be there either for reasonable runs and good cables.

karbomusic, cool that you cracked yours open and gave us a run down of the innards. Thank you for that

It's not so much about apples to apples noise floor. If your preamp of choice gets noisier once in its uppermost input range, and not just because it's being turned up but because the electronics are such that the noise ratio goes up unpleasantly with the input up past, say, 80%, it wouldn't be uncommon for a low output mic to record with a better s/n ration backed down some. If backing it off to where the extra noise doesn't ramp up yet sounds better (and even noise aside some preamps just don't sound as good for clean sources nearly maxed out) the recording level might be too low. The additional clean gain from one of these inline pres can put the mic signal into the sweet spot of the other pre. We got along without these things for so long that unquestionably we can still get along without them For sure a normal output mic and a clean gaining pre won't benefit.

Also, what I meant about in live use, I can see their value for long cable runs with certain combinations of gear, I've just never seen them in use at a live show or found that someone I knew who did live recordings was using them on stage. It does seem that by now almost everyone I know has them

[cyrano]

Originally Posted by karbomusic

In relation to what? What I mean is that I'd expect 10-100 times the mic impedance to be more than ideal from an impedance matching perspective (minus other complications). However, I'm sitting here staring at a $2000.00 mic pre I own, it's variable Z only goes to up 1.5k. So "in relation to what means" we've obviously been recording pretty well over the last 60 years, seems like that'd be hard to miss or just give it a high enough impedance to begin with - or we are splitting gnat hairs which is where I lean on this.

In this case, it's not that simple. And nobody seems to have cracked the physics behind it.

What we suspect is that impedance is the major factor. But it's clearly not linear. So there must be some other factor.

The circuit that was used, wasn't an opamp, but a simple one or two transistor stage. Should be fully known...

And that was the surprise. It was very hard to make a low noise circuit with the usual suspects, eg the typical low noise, high gain transistors, because in this circuit they operate at a relatively high bias point.

To stay within phantom power limits proved to be a problem from the start.

That's when it was found out that some switching transistors performed better. And fet's. And finally even old germanium ones.

It was also dependent on the preamp that followed the circuit. Nobody expected that either, considering most mic preamps have an impedance of 1 to 5 kOhms.

I'm not sure where the magic is there though. It's pretty common knowledge JFETs usually give you Hi-z where BJT doesn't; JFETs vary, they get measured and selected if it matters. None of this is a disagreement btw, just thinking out loud but... I just don't really buy into the variable Z unless our goal is modify the response, sufficiently high, isn't going to give someone something that doesn't actually exist.

That's exactly what I'd figure too. High-Z is more sensitive to noise pickup from the environment, but we were looking at noise production in the circuit itself. And that's where the fets faired better than the transistors you'd usually use. So it was new, uncharted territory when this all started over 15 years ago.

Nobody's gone forward and "published" a circuit yet, as there is no understanding of the rules involved.

I just remembered more than usual from reading that thread because "impedance matters" has been bugging me for years now. Just like "balanced to unbalanced" isn't always straightforward either.

I suspect we can still hear a lot more than we can measure. But (white) noise is easily quantifiable and that was the main parameter in all of this. If there is an audible, annoying hiss, you don't need to measure. And that's what happened. The combo mic/CL/preamp produced more hiss in some cases than what was added by the CL. And less in other cases.

My take is that the spectrum of the noise could influence these results. But that would require that some of the noise produced is out of polarity with noise produced in the mic preamp itself. And that's a hard one to swallow as that noise would be random. A quantum effect?

So where is the noise coming from, or disappearing to?

[ashcat_lt]

The only time that ANY boost is better than Take Volume in Reaper is if it also improves S/N somehow. That's literally all that matters. If using the CL or whatever adds noise, but doesn't somehow put it in a range where the preamp has less noise, it's a waste of time and money. Similarly, if your preamp gets noisier when you turn it up, then don't. Why would you? Just so the meters look good during tracking? Who cares? S/N is all that matters. Reaper has an infinite amount of clean, silent gain available once you get your signal inside.

[Icchan]

Well... I'm an electronics designer by trade so... bear with me.

First of all, every single component on your signal path has a self noise, and in simple terms (so don't take this as a gospel) higher the impedance of that said passive component, higher the noise voltage. Larger the physical size of that component, higher the noise etc.

That's why to get the lowest noise, you want to minimize amount of any electronics between the source and the destination (e.g AD converter).

That's why very high impedance mic pres would be highly susceptible to noise. And this is why designing a good mic pre is quite a complicated undertaking since the source impedance can vary wildly depending on what mic gets attached to the cable, and if you just make the preamp very high impedance thinking it'll work with any mic, you'll add too much noise... so you need other solutions and you need compromises.

That being said, every single amplifier stage (each operational amplifier if you will) will add some input noise of its own, then it will amplify the original signal, and any noise that's in that signal already, so your STNR will be the same as originally+added input noise

Then it'll add output noise and in the end your output signal has more noise then to begin with.

This is fundamental, you can't avoid this in real world circuits. You can just have amplifiers that have more noise than the fundamental level... the issue is how close can you get? (spoilers, not very close cheaply with bandwidth of 20khz)

This is why you don't necessarily want to have multiple gain stages, or if you do, have your first stage be as much gain as possible (or as much as you need) so that any noise added by next stages is low in comparison to your amplified signal level.

And there are physical limitations as to how low the noise added by these stages can be, and in my experience, we've fought against that fundamental limit in audio for decades already. You really want to have as much signal straight from the source as you're able to get.

Dynamic mics are inherently less sensitive, and condensers can be made incredibly sensitive. Different materials, different mechanics, electronics (condensers have a pre-amplifier built in) and all that will affect the microphones self noise and sensitivity, thus you really need to have the right mic for the right source and environment.

And the right pre-amplifier for your microphone.

Mic with a high output impedance should be more sensitive, if it isn't you'll run into trouble with needing very low noise and high impedance mic-pre with high gain that can be costly to design and thus to own.

Further more, high output impedance will wreak havoc to your signal if you have longer cable between the mic and the pre-amplifier... I'd avoid anything that's over ~300 ohms in live situation. Though as I stated earlier, today mic pres can be close to the stage and then everything is digital so audio cable lengths have collapsed from the early days when you could run the multi-core all the way to the FOH mixer.

At the other extreme would be low output impedance, high sensitivity microphone with a loud source... adding passive attenuation in line with that microphone is way cheaper and better regarding noise than trying to add gain to low level signal.

[ashcat_lt]

It seems like some preamps have relatively high output noise compared to input noise such that turning up the gain will help quite a bit with S/N as it "lifts" the signal up out of that noise floor.

OTOH, some designs actually do add more proportionally more noise as you increase gain. That sort of makes sense for an opamp stage where gain is proportional to the resistance in the NFB path. More resistance = more gain but also more noise.

I very often wonder if all-in-one preamp/interfaces actually add much gain. The digital part certainly cannot handle the sorts of levels you might get from a typical analog preamp. Seems like "gain" must sometimes mean "less attenuation", which opens its own set of issues.


Edit - Though honestly much of the time, the S/N is limited more by the noise in the source and its environment and any noise added by subsequent gain stages is after low enough to be ignored.

[airon]

That‘s a lot of good information folks. Thank you very much. I‘ll likely stick to my RME Fireface UC preamp for now.

If I do need extra gain, these units will still be there.

[SonicAxiom]

Originally Posted by Icchan

Well... I'm an electronics designer by trade so... bear with me.

First of all, every single component on your signal path has a self noise, and in simple terms (so don't take this as a gospel) higher the impedance of that said passive component, higher the noise voltage. Larger the physical size of that component, higher the noise etc.

That's why to get the lowest noise, you want to minimize amount of any electronics between the source and the destination (e.g AD converter).

That's why very high impedance mic pres would be highly susceptible to noise. And this is why designing a good mic pre is quite a complicated undertaking since the source impedance can vary wildly depending on what mic gets attached to the cable, and if you just make the preamp very high impedance thinking it'll work with any mic, you'll add too much noise... so you need other solutions and you need compromises.

That being said, every single amplifier stage (each operational amplifier if you will) will add some input noise of its own, then it will amplify the original signal, and any noise that's in that signal already, so your STNR will be the same as originally+added input noise

Then it'll add output noise and in the end your output signal has more noise then to begin with.

This is fundamental, you can't avoid this in real world circuits. You can just have amplifiers that have more noise than the fundamental level... the issue is how close can you get? (spoilers, not very close cheaply with bandwidth of 20khz)

This is why you don't necessarily want to have multiple gain stages, or if you do, have your first stage be as much gain as possible (or as much as you need) so that any noise added by next stages is low in comparison to your amplified signal level.

And there are physical limitations as to how low the noise added by these stages can be, and in my experience, we've fought against that fundamental limit in audio for decades already. You really want to have as much signal straight from the source as you're able to get.

Dynamic mics are inherently less sensitive, and condensers can be made incredibly sensitive. Different materials, different mechanics, electronics (condensers have a pre-amplifier built in) and all that will affect the microphones self noise and sensitivity, thus you really need to have the right mic for the right source and environment.

And the right pre-amplifier for your microphone.

Mic with a high output impedance should be more sensitive, if it isn't you'll run into trouble with needing very low noise and high impedance mic-pre with high gain that can be costly to design and thus to own.

Further more, high output impedance will wreak havoc to your signal if you have longer cable between the mic and the pre-amplifier... I'd avoid anything that's over ~300 ohms in live situation. Though as I stated earlier, today mic pres can be close to the stage and then everything is digital so audio cable lengths have collapsed from the early days when you could run the multi-core all the way to the FOH mixer.

At the other extreme would be low output impedance, high sensitivity microphone with a loud source... adding passive attenuation in line with that microphone is way cheaper and better regarding noise than trying to add gain to low level signal.

valuable info right there, thanks!

.

[Gilligan8]

Going to dig this one up from the grave as I'm looking for more info on the cloudlifter circuit.

Anyone care to take a pic of both sides?

I'm 99% sure this is 95% of the same circuit. Just different jfets but looks almost like a straight replacement. Only thing is there are a few more/different resistors in the CL.

I think Alctron just copied their circuit and used a "cheaper" jfet pair.

[Unox]

Originally Posted by Gilligan8

Going to dig this one up from the grave as I'm looking for more info on the cloudlifter circuit.

Anyone care to take a pic of both sides?

I'm 99% sure this is 95% of the same circuit. Just different jfets but looks almost like a straight replacement. Only thing is there are a few more/different resistors in the CL.

I think Alctron just copied their circuit and used a "cheaper" jfet pair.

Hi!

The schematic is basically the same. R3 and R4 are 30K in the cloudlifter. Can be that the LSK389C has a better noise spectrum, dunno.

But I made a funny mod in my cloudlifter.

Had a Carnhill VTB9045M mic input transformer laying around. Changed the input resistor on the cloudlifter PCB from 3k to 47k which -together with R3 and R4- translates to a bit under 2k at the transformer input. Transformer wired inputs parallel, outputs in series, adds about 11 dB. And some Carnhill Neve-like color of course.

Fets are cool and the LSK389 is dead quiet with only 1.5nV noise (and very low current noise). So the extra input resistance the Carnhill adds doesn't translate in excessive noise like it would in a bipolar input stage design.

I did this cos I have a very low, soft voice and my SM7B needs almost the full 60 dB of gain a preamp has to offer, and 60dB's of gain doesn't bring out the best in most pre's.

SSL used to have three stages in their mic pres, first a Jensen trans, then two NE5534's (kinda biased in class-A) taking care of the rest.

This cloudliftermod does about the same. And sounds pretty ok. Great noise rejection, royal, clean gain without extra caps in a small box.





Best,

Nick