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
User avatar
mediatechnology
Posts: 5444
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 »

1. There's a 1M resistor right after the 1uF/250V cap. In the other circuit it is 100K instead. What value is correct?
2. In the filter section there are two optional caps. If they have to be omitted, should they be short-circuited?
3. I've noticed that the trim stage is inverting. Does it mean we get the inverted signal at the output (I don't see any other inverting stages)?

I was looking for a phantom power switches, and most miniature switches are rated for 5A @ 120V AC or 28V DC. I wonder if it's safe to use them @ 48V and small phantom current?
Tha value of the resistor after the 1 uF determines the LF cutoff. IIRC in my second stage I have a switch there with 1 M having a 0.2 Hz cutoff for ultra low sq wave tilt* and two much lower values to permit a 6 dB/oct HPF for the real world at 20 and 80 Hz. Note that the 2k value to produce 80 Hz is the minimum load the 1510 will drive.

*(Tell me how that one sounds on kick drum please.)

Those optional caps are there in stage two just in case they're needed. I expect that stage is best made a follower without gain but I also placed an optional resistor from the inv input to ground in case someone wanted a bit more post-gain. Post gain offers higher values of Rgain in the 151X stage and can also allow the 151X to drive a lower impedance termination to move the HPF upward. (The 151X minimum load is limited by drive current.)

Good catch. Yes, the trim stage does invert. It's not obvious, but I flipped the 1646 outputs (redefining them) so that the 1646 effectively re-inverts the inverter. Overall it should follow correct polarity.

Have you considerded a relay or dual photoMos relay for switching phantom? A PS7113-2A (dual) works very well.

http://www.necel.com/opto/en/pdf/PN10269EJ02V0DS.pdf
ilya
Posts: 88
Joined: Tue Dec 29, 2009 2:10 pm
Location: Russia
Contact:

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

Post by ilya »

Wayne,

Thanks for comments. I was distracted by optional caps and thought that you did the opamp HPF circuit, and I tried to figure out how that worked!
Now I see that the HPF switch parallels 2K and 8K8 with 1M and everything is much more clear to me now!

Thanks for pointing out the reversed 1646 output pins. I thought that it should be wired in reverse polarity, but haven't checked the pinout at the time.

One question that's still there is what's the reason for the first (non-inverting) buffer? Can't we just omit it and go right into the trim opamp after the RC filter?

Thanks for suggestions on solidstate relays btw. They're very convenient, although this may be not a cost-effectie solution since we're adding the second switch in the circuit where the mechanical switch should work just fine. What I need to do is find the switch with suitable specs though :)

Here's the second PDF: http://www.dontsov.ru/DIY/DCPreamp/Auxi ... ircuit.pdf
User avatar
mediatechnology
Posts: 5444
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 »

Ilya -

The reason for the first buffer stage is to permit the use of a high quality film cap and allow a low value with low stored charge with ultra-low LF cutoff. Thus the 1M and bifet op amp follower.

You also might want to read in the middle of the first post about Jung's use of "balancing" resistors around the bifet. There's some links to ADI. It really improves the THD performance of the bifet.

Wayne
rmodeste
Posts: 1
Joined: Sun May 16, 2010 12:15 pm

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

Post by rmodeste »

Wayne,

I have been reading your posts and I am really impressed and happy to see this project and all the progress that everyone has made.

I will give you a little history. My good friend Don Pearson had been bugging me for years to work with him to do this same project. Don was an audio professional for at least 40 years. He founded UltraSound in San Rafael CA and was one of the Greatful Dead's principal touring sound engineers.
In the early early 80's we had built a bank of 16 mic preamps that used a transformer coupled front end and a servo-ed Jensen 990. These were extensively used by the Dead and many other bands for years. Even though these mic preamps worked quite well for that time Don was always in pursuit of audio perfection and thought we could do much better. He thought that a direct coupled mic preamp was the advancement necessary to get there.
We had a different idea for realizing a direct coupled mic preamp design. The more I read your posts the more I realized that you are on the right track. Our idea was to use a cascoded front end to shift the DC level of the phantom power down to a more manageable level and then servo out the remaining DC. I never got deep enough into it to realize all the practical problems and issues that you uncovered during your design process let alone build a prototype. Don started me on this thought about 20 years ago. We would pick this up and put it down every year or two although never seriously. It just turned into a good reason to stay in contact.
Unfortunately, about 5 years ago Don died. He would really have enjoyed talking to you about this project. I am sure he is smiling somewhere and trying to whisper a few of his opinions into your ear.
For my part, I am going to prototype up this design and work on a PCB layout. Like most people, I am way too busy to make progress in an expedient manner but I will make the time to read up on your progress and post whatever tidbits I discover.

Thanks,

Rob
User avatar
mediatechnology
Posts: 5444
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 »

Thank you Rob for the kind words and history. You may have inspired me to dust off that protoboard and for that matter my entire workbench. I haven't done much in there other than change batteries since November.

As I mentioned to Ilya earlier up in the thread I would change the servo to the one here: viewtopic.php?f=6&t=256

It just works better and reduces the DC requirements of the servo amp. I think it will work pretty good in the DC preamp too - if not better - than the one that senses the emitters. I should lash that up.

I've only heard of two microphones that only pull current out of one pin and are unbalanced. They are the Octavia (don't know which model, OC112?) and Behringer Measurment Microphone. The servo wouldn't be able to correct for that. If you know of any others that are unbalanced I'd be curious to learn of them.
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. »

Yup, Don's work and reputation is pretty well known, at least to us old farts.

One of my favorite scribbled approaches is DC coupling the input transistors of the typical "discrete + opamp" preamp topology , and letting them float up to whatever voltage the input demands. IIRC the very low noise input devices are a bit marginal for full breakdown voltage so cascode is useful there to keep them within safe operating voltages.

Another area that I have thought about but never melted solder over is the phantom supply. This 48V voltage source fed through 6.8k build out resistors is an archaic approach, attractive for it's simplicity, but not cast in stone. If we think about how this supply is used in active microphones and direct boxes, the voltage is not as useful as the current. I am not aware of this being a polarizing voltage for any popular mic designs.

I can imagine several alternate approaches. Why does the phantom R need to be 6.8k? The AC impedance at that node is probably closer to 1k to ground or less, so making this lower impedance the DC path, means X mA of phantom supply could be delivered from a much lower phantom voltage rail. Probably still hotter than conventional IC process voltages but not the current 48V+15V and change we need to survive. Adding current sources in parallel with the Rs might help even more, but this may deliver diminishing returns as the input probably needs to rise to at least 15-20v to keep the mic/DI circuitry happy.

Another thought as we evolve into a progressively more digital world, is to float the A/D convertor up to 48V with the preamp and then level shift the digital output with an optical loop or whatever is handy.

I appreciate Wayne's investment in time and labor to identify so many real world gotcha's, this has been a several decade long mental masturbation for me too. I came across Wayne's work several years ago, but remain apprehensive about the real world complications imposing this upon customers, while i suspect a small esoteric product could get away with it.

At a minimum, conventional mic preamps could inexpensively bypass their input blocking caps when phantom power is not being used, but AFAIK this is not done, certainly not widely if done at all. Customers would probably be more turned off by the switch noise, than any subtle sonic benefit they might receive. Customers are funny that way. The best sounding audio path will be dismissed if it suffers switch clicks and pops and turn on/off transients, as evidenced by all the cap coupled gain pots to prevent scratching. Certainly just as bad sonically as the input blocking caps if not more so.

JR

PS: I did have one correspondence with one of the lads, back in the 'late 70s about dielectric absorption in tantalum caps. Maybe it was Healy, I don't remember which one, but he was a kit customer of mine, way back then. FWIW the kit was a DJ mixer that doesn't sound much related to their day job, but most bench tweakers messed around with phono preamp designs back in those days.
Cancel the "cancel culture", do not support mob hatred.
User avatar
mediatechnology
Posts: 5444
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 »

JR. wrote:Customers would probably be more turned off by the switch noise, than any subtle sonic benefit they might receive. Customers are funny that way. The best sounding audio path will be dismissed if it suffers switch clicks and pops and turn on/off transients, as evidenced by all the cap coupled gain pots to prevent scratching. Certainly just as bad sonically as the input blocking caps if not more so.
+1. With the tweeters blown out in the monitors it's sometimes hard to make good comparisons. ;)

This was recently a major point of contention where I designed a circuit that clicked, could not be prevented from clicking, told everyone that it clicked and to just not use it. Sometimes you have to "Just Say No."

No one believed me that it really did click. I responded that I've built one, it clicks, you haven't built one, try it and please get back with me when you do. Sure enought it clicked too. I had a good guess from a scientific standpoint why but I'd actually built one and the science later backed me up. I'm not particularly fond of switching tip and ring in mic preamps either.

Don't do this: http://www.thatcorp.com/datashts/dn140.pdf#page=5 (schematic figure 4)

Here's why: http://www.thatcorp.com/datashts/dn140.pdf#page=7 (DC Blocking Capacitors and Clicking)
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. »

Not to insult the theme of this thread... if you added a second pair of blocking caps after the switch wiper would it get silent? Ugly and not very elegant, but not very expensive in the grand scheme of things. Some cap whackos might even favor biasing polars, + lead out away from the switch in both directions and biasing the switch DC operating point slightly negative. Even if this doesn't do much if anything useful, you could spin it into a hypothetical advantage (excuse) to argue that solution with. :lol:

FWIW I've done quite a bit of polarity switching at the mic level front end without customer complaints, while IMO anything you can do later in the signal chain with no compromise is preferable. In console design, layout and signal flow matters so when pad and polarity functions are grouped together, it is a tighter layout operating on the mic level signal for both.

Note: a recent poster in another forum recently complained about oscillation in a console where he had made modifications. It appears his lead dress was the source of oscillation. Putting an output polarity swap switch, right next to an input pad switch, is flirting with path complications. Customers think of pad, phantom, and polarity as "input" functions, so logically like them located close together.

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. »

http://www.datasheetcatalog.org/datashe ... Xwvssz.pdf

While looking for replacements for my old and well loved 2sb737 low noise devices. I notice the Hitachi 2sa1084 is good for 90V breakdown and 2sa1085 is better at 120V. Either of these could be used in a DC coupled design with the 48V phantom. In my old scratching using 2sb737 I had to cascode the input devices to deal with peak voltages.

Just a little food for thought as modern parts offer solutions not available when I was actively looking at this.

JR
Cancel the "cancel culture", do not support mob hatred.
User avatar
mediatechnology
Posts: 5444
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 »

Well, there may be a problem using those:

http://documentation.renesas.com/eng/pr ... 1084ds.pdf

I was afraid of that when I noticed your link was to datasheet archive and the datasheet was 2001. So many parts are getting killed off. Maybe ON or Fairchild will pickup the 2SA1084.
Locked