New: A Direct-Coupled Input-Capacitorless Active Mic Preamp

Where we discuss new analog design ideas for Pro Audio and modern spins on vintage ones.
olafmatt
Posts: 32
Joined: Wed Oct 22, 2014 11:33 am
Location: Europe

Re: New: A Direct-Coupled Input-Capacitorless Active Mic Preamp

Post by olafmatt » Sat Jun 09, 2018 2:49 am

mediatechnology wrote:
Fri Jun 08, 2018 6:48 pm
From experience, what value do you see in the technique? Lower noise, increased transparency? What are the audible benefits realized in practice?
This is tricky to answer because I never compared the exactly same preamp with flying frontend and no caps to how it sounds with flying disabled and caps added. And because it used some sort of level shifting between the gain stage and the diff-amp (to avoid caps in the direct signal path altogether) several impedances had to be higher than I'd normally make them (to reduce power dissipation in the resistors with an extra up to 48V DC offset). So noise performance (at least at low gains) was worse than in a traditional preamp.

The advantage I saw was that you could switch phantom power on and off without any audible plops and clicks (unless the microphone made them). Also there's no rising noise at LF. But of course you still get LF rolloff, which is something most people overlook if they hear "no caps in the direct signal path". I'm not so sure whether I noticed any improved sound due to "no capacitors", since obviously there are still caps in the microphone itself. And those are usually the worst caps anyway due to the small space available inside the mic body. - Removing the cap in series with Rgain is a far greater improvement.

I built that thing almost 10 years ago, before I came up with the new servo circuit. Maybe I should try again... but wondering whether I should try again made me think whether the selected preamp circuit is the best one to use here.
Now with the new servo I simply use (non-electrolytic) caps in the input and place the Fc very low. Previously there was always the need to have it higher in order to have it not interact with the servo's moving Fc. With 22u caps I can get the 1/f noise corner to below 20Hz, so even the LF noise contribution of the caps can be ignored.

But still, being able to say "no caps in the direct signal path" is a nice thing to have (and the phantom blocking caps plus the caps in the microphones are currently the only caps left in my signal chain's direct signal path). Just that I'd only do it again after coming up with a preamp circuit hat doesn't move the DC conditions on the input around.

User avatar
mediatechnology
Posts: 3276
Joined: Sat Aug 11, 2007 2:34 pm
Location: Oak Cliff, Texas
Contact:

Re: New: A Direct-Coupled Input-Capacitorless Active Mic Preamp

Post by mediatechnology » Sat Jun 09, 2018 5:52 am

I didn't realize that your original version "burnt off" the common mode DC as heat. I recall Chalupa did that.

Thanks for sharing your observations!

User avatar
mediatechnology
Posts: 3276
Joined: Sat Aug 11, 2007 2:34 pm
Location: Oak Cliff, Texas
Contact:

Re: New: A Direct-Coupled Input-Capacitorless Active Mic Preamp

Post by mediatechnology » Sat Jun 09, 2018 9:39 am

Just that I'd only do it again after coming up with a preamp circuit that doesn't move the DC conditions on the input around.
What do your instincts tell you about which topology would be affected the most by moving DC conditions?
What would be the audible effects?

What about common mode wiggles?

The common mode DC potential of the mic output often varies significantly with operating current/SPL.
The DC Q-point varies with acoustic signal to produce an AC common mode signal.
(Even-order half-wave IIRC.)
Some of it is LF and infrasonic.

Common mode to differential conversion of those wiggles would seem to occur in any active input preamp - more so if AC-coupled.
So I'm wondering if this is something that isn't already present in conventional topologies.

Your thoughts?

olafmatt
Posts: 32
Joined: Wed Oct 22, 2014 11:33 am
Location: Europe

Re: New: A Direct-Coupled Input-Capacitorless Active Mic Preamp

Post by olafmatt » Sat Jun 09, 2018 12:35 pm

mediatechnology wrote:
Sat Jun 09, 2018 9:39 am
What do your instincts tell you about which topology would be affected the most by moving DC conditions?
What would be the audible effects?
With microphones with caps on the output you can get interaction between the servo and those caps. You can look around and find the mic with the worst-case cap size and scale your servo accordingly. Until somebody plugs in a mic where the manufacturer went cheap and put in even smaller caps than you thought they ever would.

IIRC you once pointed me to an article where somebody made a phantom powered microphone that was drawing current only on one signal line. Then it had a servo in there that was drawing an equal current on the other line to not produce any output offset. In such a case, which servo will win the fight?
mediatechnology wrote:
Sat Jun 09, 2018 9:39 am
What about common mode wiggles?
Haha, now you got me... I was actually thinking about asking for your motivations for making an input-capacitorless mic preamp. Like, which "problem" do we actually want to solve and do we actually know that the cause of the problem is the input caps? - For my part I believed the gospel that input caps are bad and need to be removed. And it is a nice little chalange to build something that many people think "can not work".

But seriously, what's broken with normal mic pres with phantom power on them? Somehow I felt that many people were complaining about phantom supply voltage (on the mic end) gowing lower for high SPL signals because the mics draw more current. If you look at Sennheiser MKH 20/30/40 series mics it doesn't happen there since they have a CCS and always sink the same amount of current. Maybe that's why these mics sound better than some other mics? (Actually Shoeps mics are reported to sound "better" when you decrease the phantom resistors from 6k8 to 5k6.)

With mic current consumption changing you get a common-mode signal. This modulates the Ccb of the input transistors. So I went and made a preamp that uses input transistors in a bootstrapped cascode. So essentially the voltage (Vce) across the input transistors is held constant, no matter what the input signal is. This means that common-mode input signals don't modulate Ccb anymore.

Then you have many modern mics with impedance balanced output. If we do regular calculations of clipping levels and required current drive capability for opams we assume symmetrical signal swing. But unless you have your preamp set at very high gain it is not symmetric with such a microphone. This needs to be taken into consideration when deciding on required headroom (voltage and current wise) and has the biggest influence at very low gain.

There is that Gaskell AES paper about cap sound in micpres... and unless I'm reading it incorrectly, it shows that for a PET cap as DC blocker the rise in distortion caused by the cap starts to move up where the frequence reponse is roughly -1dB. So just make the Fc of the DC blocking caps low enough (which you should do for other reasons anyway) and those caps will moste likely not have any influence on your signal. - So if that paper is correct, at least distortion contribution of these caps is not a reason to get rid of them.

User avatar
JR.
Posts: 2418
Joined: Sat May 24, 2008 7:21 pm

Re: New: A Direct-Coupled Input-Capacitorless Active Mic Preamp

Post by JR. » Sat Jun 09, 2018 12:51 pm

I won't attempt to speak for Wayne, but my interest in DC coupled mic preamps was purely to create a marketing hook, for all the capacitor haters, and an interesting mental exercise. I never attempted to do this for a mass production environment fearing unintended side effects. Some of the cost overhead for elevated front end rails could be relatively cheap when spread across tens of console channels, but problems could be problematic to remedy in a larger audience.

No need to repeat how many other capacitors are in the full sound path. :o Kind of like back in the 70/80s when the phools would complain that they could "hear" the sound of audio passing signal through a switch. :roll:

Indeed we can hear "bad" components of any flavor.....

JR

User avatar
mediatechnology
Posts: 3276
Joined: Sat Aug 11, 2007 2:34 pm
Location: Oak Cliff, Texas
Contact:

Re: New: A Direct-Coupled Input-Capacitorless Active Mic Preamp

Post by mediatechnology » Sat Jun 09, 2018 1:16 pm

Haha, now you got me... I was actually thinking about asking for your motivations for making an input-capacitorless mic preamp.
I honestly don't know what they were. I think it was "because I thought I could."

That servo'd output paper IIRC is Dimitri Danyuk's which appeared as a Design Idea.
Who wins? Don't know.
Wouldn't "conventional" servo's at the bases also have an issue about who was in control even with AC-coupled inputs?
There are two servos operating against (or with) each other on opposite sides of the input coupling cap.

When you turn on a phantom-powered mic from a cold state with an active AC-coupled input how long do you usually wait before its quiet enough to record?

I think you're right about the conclusions in the Gaskell with regard to Fc vs THD.
I scanned it though and noted that they measured higher LF THD with applied voltage DC voltage regardless of capacitor type though PP and PET were much lower compared to conventional polar and bipolar electrolytics.

User avatar
mediatechnology
Posts: 3276
Joined: Sat Aug 11, 2007 2:34 pm
Location: Oak Cliff, Texas
Contact:

Re: New: A Direct-Coupled Input-Capacitorless Active Mic Preamp

Post by mediatechnology » Sat Jun 09, 2018 1:18 pm

JR. wrote:Kind of like back in the 70/80s when the phools would complain that they could "hear" the sound of audio passing signal through a switch. :roll:
You obviously never had to replace ITT Schadow switches by the thousands in MCI consoles because they distorted.
The ten poles in the EQ were particularly fun.
Sorry you missed that.
Tonelucks rule.

User avatar
JR.
Posts: 2418
Joined: Sat May 24, 2008 7:21 pm

Re: New: A Direct-Coupled Input-Capacitorless Active Mic Preamp

Post by JR. » Sat Jun 09, 2018 2:27 pm

mediatechnology wrote:
Sat Jun 09, 2018 1:18 pm
JR. wrote:Kind of like back in the 70/80s when the phools would complain that they could "hear" the sound of audio passing signal through a switch. :roll:
You obviously never had to replace ITT Schadow switches by the thousands in MCI consoles because they distorted.
The ten poles in the EQ were particularly fun.
Sorry you missed that.
Hopefully you were compensated for that. :lol:
Tonelucks rule.
I had to deal with more than my share of bad insert jacks where the switch contacts would oxidize or otherwise become intermittent (and that wasn't my worst jack drama.... how about female XLRs that refused to release? (more than thousands... Neutrik had to change the design and that took months, so add a few zeros to the quantity. ) :oops: The insert jack problem was traced to a purchasing agent approving a change to the (switchcraft) plating media without rigorous engineering approval.

JR

User avatar
mediatechnology
Posts: 3276
Joined: Sat Aug 11, 2007 2:34 pm
Location: Oak Cliff, Texas
Contact:

Re: New: A Direct-Coupled Input-Capacitorless Active Mic Preamp

Post by mediatechnology » Sat Jun 09, 2018 2:31 pm

Hopefully you were compensated for that. :lol:
Yes I was. Better than I am now. :lol:

olafmatt
Posts: 32
Joined: Wed Oct 22, 2014 11:33 am
Location: Europe

Re: New: A Direct-Coupled Input-Capacitorless Active Mic Preamp

Post by olafmatt » Sat Jun 09, 2018 2:44 pm

mediatechnology wrote:
Sat Jun 09, 2018 1:16 pm
Wouldn't "conventional" servo's at the bases also have an issue about who was in control even with AC-coupled inputs?
There are two servos operating against (or with) each other on opposite sides of the input coupling cap.
That's a good question. - And it is not just about "conventional" servos... my servo is no different, it just injects the correction that moves the bases "from the rear".
mediatechnology wrote:
Sat Jun 09, 2018 1:16 pm
When you turn on a phantom-powered mic from a cold state with an active AC-coupled input how long do you usually wait before its quiet enough to record?
At least 30 minutes, but that's not because of the preamp, it is because of the mic and all the electrolytic caps in there. And usually it takes the musicians even longer to be quite enough.
mediatechnology wrote:
Sat Jun 09, 2018 1:16 pm
I think you're right about the conclusions in the Gaskell with regard to Fc vs THD.
I scanned it though and noted that they measured higher LF THD with applied voltage DC voltage regardless of capacitor type though PP and PET were much lower compared to conventional polar and bipolar electrolytics.
Yes, the figures with 48V applied were a lot higher. But stil, even with worst case electrolytic caps the rise of THD above 0.001% only started where the frequency response had already dropped by roughly 0.1dB. So sticking with PP / PET and making FC at say 1Hz should be "good enough".

Post Reply