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

Where we discuss new analog design ideas for Pro Audio and modern spins on vintage ones.
Locked
clintrubber
Posts: 3
Joined: Fri Jul 18, 2008 8:36 am

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

Post by clintrubber »

Hey, looks like a few familiar persons around here :D

Thanks for the welcome!

Have a good weekend all,

Peter
User avatar
mediatechnology
Posts: 5442
Joined: Sat Aug 11, 2007 2:34 pm
Location: Oak Cliff, Texas
Contact:

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

Post by mediatechnology »

Peter - so glad to see you here.

JohnR wrote:
A two gang switch for 10s and 1s could give more than enough range, putting the 1s on a second gain stage would simplify managing gain interactions.
Yes that is the best way to do it with coarse adjustment (10dB) steps upfront and fine (1dB) in the gain stage. Juggling the two as a single Rgain would be difficult indeed. I think that may be how it's done internally in the PGA2500 with fine adjustment in the second stage.
User avatar
mediatechnology
Posts: 5442
Joined: Sat Aug 11, 2007 2:34 pm
Location: Oak Cliff, Texas
Contact:

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

Post by mediatechnology »

Some of you may find this statement in the Benchmark MicMan Jr. interesting:
Microphone power [phantom] should be turned on 1/2 hour ahead-of-time, to allow formation in the dielectric in the coupling capacitors.
I remember a discussion about the SSL 82E149's use of back-to-back "non-polar" input capacitors which had the positive terminals (in the middle) biased with always-on 48V applied. The caps were always polarized as long as the console was on - regardless of the state of phantom. I vividly remember certain consoles not having this configuration taking very long times to become quiet after phantom was switched on.

I mention this because that leakage current noise while the dielectric forms is significant. We can't always plan 1/2 hour ahead. Another reason to eliminate the input caps...Dielectric formation time.
User avatar
JR.
Posts: 3700
Joined: Sat May 24, 2008 7:21 pm
Contact:

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

Post by JR. »

I feel like the old duffer repeating his stories, but I recall disapproving one vendor's capacitor in that application (blocking phantom) while at Peavey when the cap vendor made some process change to their part that made their leakage current audibly noisy. In a production environment nobody waits a half hour to retest. My recollection is it was a process change to shrink the part size, that caused the problem in only that one part number.

I've never experienced forming issues except for some really old parts sitting in a warehouse for too many years.

JR
Cancel the "cancel culture", do not support mob hatred.
User avatar
JR.
Posts: 3700
Joined: Sat May 24, 2008 7:21 pm
Contact:

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

Post by JR. »

We have a smoking gun, for my concerns about DC coupling every mic in the world.

Problems reported with TLM 103

http://www.prodigy-pro.com/diy/index.ph ... #msg407094

JR
Cancel the "cancel culture", do not support mob hatred.
User avatar
mediatechnology
Posts: 5442
Joined: Sat Aug 11, 2007 2:34 pm
Location: Oak Cliff, Texas
Contact:

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

Post by mediatechnology »

John - Thank you for bringing that to my attention. I think the comment made was for a topology that didn't provide correction of differential offsets from mic-induced errors. I saw that problem from the get-go and the posts here deal with that issue in great detail.

I wrote early on:
7) Because DC gain exists, offsets which occur within the microphone itself become significant. In many cases the input DC differential offset may be larger than the signal itself. Later in the post we explore the sources of microphone-induced offsets and deal with compensating it using servos. Without servos, the gain multiplication of the offset would cause the 1510 to clip. The servo samples input offset at the 1510s input transistor emitters (also known as Rgain) and correction is applied back to the input as an injected current. Further examination reveals that the phantom resistors and microphone internal resistors form a Wien bridge.
Olafmatt wrote:
Been there, done that... Actually it works pretty well. Used an opamp to create a 'differential virtual ground' (that only looks like ground to common mode signals, but not so to the differential audio signal). That way you can burn away the phantom offset after the 'flying' input stage. Problem is too keep everything low impedance you need high-power resistors. I did it with 10k (higher than I would usually use), which means 4.8mA DC flowing through a resistor that also carries your audio. Thats about 0.25 Watt to burn. Alternatively, try a zener diode in series to drop the offset down to something the next stage can handle.
There's no mention of any attempt to eliminate differential mic offset, just the 48V common mode offset.
I also have to note that my cap-free micpre never really liked a TLM103. It worked flawlessly with all mics I tried, except the TLM which seems to draw more phantom current on one leg than the other and thus confuses the floating mechanism and servos.
I'm surprised it worked with as many mics as it did. Transformer ones are a cinch. Emitter follower outputs are OK too. Purely capacitor-coupled ones I would think are going to be the worst. With a modest 40 dB gain mic offset is multiplied 100-fold. For 60 dB gain as little as 10 mV is going to drive the output to 10V.

And then there's this very good point:
There is one reason that speaks against doing it: you will need a filter after the micpre to filter out all the low-frequency rubbish. With the mics shock-mounted on a regular mic-stand I could see passing trains in the wave display of the DAW. And the train line is about 1km away. So unless you come up with a capacitor-free low-cut as well the whole effort is useless.
Which is why I've always wanted that film cap with variable load (HPF) after the preamp.

The comments regarding the TLM103 drawing asymmetric currents has been around for some time. I wrote Neumann about it two years ago. They responded:
For the feeding resistores there is a tolerance for the matching of <0.4% defined in the IEC 61938. This recomendation is missing for the microphone side at the moment. But most manufacturer use this for the microphone side as well.
When I pressed them about uneven current draw, citing the Behringer measurement microphone (though not by name) I received:
Yes, there one or two manufactutres sending all the current over one pin. All the others are more or less symmetric regarding current. In the old German standard it was a rule to be symmetric and it will probably be part of the next issue.
The Behringer measurement microphone, being unbalanced, would never work.

In the post that you cited olafmatt tried floating the 1512 supplies and then offsetted the output 48V to remove the common mode DC instead of AC-coupling the output. He did not specifically mention any differential servo being used to eliminate mic offset such as that of his TLM103. Even minute offsets, just from resistor tolerance in some topologies, when multiplied by the DC gain of the preamp, would indeed be a deal-breaker for that topology. But it's not a smoking gun.

I anticipated DC offset produced by the microphone and also used a differential servo similar to Menace. IIRC it can correct tens, maybe hundreds of millivolts of input DC error. There are different ways to do that now as posted here in another thread and more than a few things I would change from the original drawing.

http://www.picocompressorforum.com/foru ... ?f=6&t=256

I went back and looked. I wrote:
Now if 60 mV of offset is a realistic microphone-induced error, and that's a very big "if," then the KS could be made much larger. There's still plenty of compliance range in the OP07 servo. If 60 mV is a small real-world error, then with 10ks there's still plenty of correction available. Point is we're only using about 1.35 volts (max in my tests) swing from the OP07 output. There's at least another 9-10 volts left in each polarity.
With the 10k current injection resistors I used to correct differential error, a 60 mV "mic offset" only required 1.35V at the servo. With bipolar (floating) 15V supplies an OP07 can provide about 13V output or ~10X 1.35V. This means up to 600 mV of error can be corrected. For more correction, the 10Ks can be made smaller. Also, in the design shown, there's a 2k-4K7 termination resistor which will equalize the offsets.

I just did a rough calculation of the error without a termination resistor to equalize the error. A Wien bridge with an open middle (no load) makes a nice model.The TLM103 is spec'd to draw 3 mA. That's 1.5 mA/leg. Assume a 10% imbalance which is more than 10X the resistance error (0.4%*2) suggested by Neumann. That's a significant 2V offset. Reduce the bridge error to 1% and it becomes 200 mV. Connect a 2K-4K7 load across that bridge (the termination resistance) and the errors drop significantly from that. Too lazy to do the math right now but the numbers drop by what, 3-6X? So we go from 2V to about ~600mV with a 10% imbalance and with a 1% imbalance it's ~30mV.

(EDIT: I just noticed how "rough" my calcs were. I assumed 10% in each leg. The figures above represent a 20% current mis-match.)

Both are within the correction range of the differential servo. Without that servo the error times the DC gain is going to produce an output pinned to the rail. I'm surprised that a lot of the non-transformer output mics olfamatt tried worked at all.

(EDIT2: I was going to check the AT mic I have here but it appears to be transformer coupled. It measured about 2 mA per leg. Since it was transformer coupled with a 10 Ohm DC resistance I built a bridge with phantom pullups 2x 6K81 that matched within ~0.1%. I then used two 15K 5% resistors to simulate the mic that measured within ~0.5% arranging the pair so the greatest offset occurred. That was around 55mV without a termination resistor. With a 4K7 Rterm the offset dropped ~3X. With a 2K2 Rterm ~6X. We're talking easily-correctable offsets even at 10X the error and 10X the IEC spec.)

A differential servo is required but I don't recommend the use of the Menace servo shown earlier in this thread. Although the 1510 demo board shows it, it may not be in the next version.

If Menace is used with AC-coupled inputs to reduce capacitor value and allow films there are additional issues. The open-circuit current noise across 100k bias resistors of Menace are significant. When shunted by the source impedance they appear to be swamped by the low value. If the goal however is to use 100k bias resistors so that low-value film caps can be used then the issue becomes the reactance, at low frequencies, of the caps appearing in series with the microphone. They become larger than the source impedance and, if too small, can increase 1/f noise.

In "wet" applications (e.g. phantom on) the 1/f and popcorn noise of the electrolytics themselves are significant due to leakage currents. Even after several minutes of electrolyte formation and being left on they can cause up to 1 uA of Ios and 1/f and popcorn noise greater than the 1512. Fortunately film caps don't have that problem - they're just too big in the values needed.

Fortunately the servos, either Menace, output to ref, or output to input acts as a high pass filter for 1/f noise.

Anyone from other fora wanting to discuss this interactively is certainly welcome to join us here. Thanks John for posting the links to us here so they can find us.
User avatar
JR.
Posts: 3700
Joined: Sat May 24, 2008 7:21 pm
Contact:

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

Post by JR. »

I believe I've shared this before but in my experience the noise contribution from leakage in the electrolytics are variable between brand/series of capacitors. I encountered one while at peavey that was so bad we had to black ball it for that application. My understanding from talking to the engineer involved was that it was a cost shrink by the manufacturer that was too aggressive. FWIW this leakage current is working into the source impedance (for audible noise) so should be modest.

JR
Cancel the "cancel culture", do not support mob hatred.
User avatar
mediatechnology
Posts: 5442
Joined: Sat Aug 11, 2007 2:34 pm
Location: Oak Cliff, Texas
Contact:

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

Post by mediatechnology »

I found the 1/f component to be quite high in a number of caps I tested even with the input terminated in 150R and Rbias at 1k. These samples were Panasonic SU-series so as you often say YMMV. For new never-before-powered caps it took them 30 minutes or more to get "quiet."

I did find the four-cap with 48V through 100k always on and applied to the middle did make them stabilize from phantom on/off events much quicker.
User avatar
mediatechnology
Posts: 5442
Joined: Sat Aug 11, 2007 2:34 pm
Location: Oak Cliff, Texas
Contact:

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

Post by mediatechnology »

Roger - I haven't had a chance to re-explore the Murata modules.

I don't know for certain that AT3525 is transformer out, but the DCR was ~20 Ohms. I think I may have written 10 Ohms in the earlier post.

See what offset your 4050s have when you get the chance.

Wayne
User avatar
mediatechnology
Posts: 5442
Joined: Sat Aug 11, 2007 2:34 pm
Location: Oak Cliff, Texas
Contact:

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

Post by mediatechnology »

Thanks Roger. Hardly a smoking gun. :D

You might check to see what the DC resistance is across pins 2 and 3. That should tell us if there's a transformer there.
Locked