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.
User avatar
mediatechnology
Posts: 3208
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 » Mon Jun 11, 2018 6:46 am

I just wanted to make it clear to those joining us late: LF servo resonance is not unique to an input-capacitorless design.

I'll draw the analogy that a servo is a subtractive, derived, HP filter. Any servo filter order greater than "1" is prone to peak.

Though the preamp's servo filter order may be 6 dB/octave, when combined with the capacitors that are potentially in the mic output, or the series combination of mic output's and preamp input caps, the filter order can be >6dB/octave if they're close in frequency.

Having the servo Fc known and unchanging with gain - as your servo does - makes the job easier.

ngineer
Posts: 3
Joined: Tue Apr 21, 2015 12:21 pm

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

Post by ngineer » Sun Aug 19, 2018 10:47 pm

Hi, Just wondering if this is still under development? I'm in beta test stage of a unit that uses a single mic pre with a 1580/5171/1606 combo and am interested to see if this approach might be better than the typical capacitor approach.

Thanks!

User avatar
mediatechnology
Posts: 3208
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 » Mon Aug 20, 2018 4:30 pm

Thanks for joining us recently and posting ngineer.

I've taken a break for the summer but do intend to revisit this project and nail down the input and servo as a next step this fall.

The 1580/5171 are great parts but almost require auto-assembly to be soldered.
In the past I've used DIP adapters and they are the best option for using these parts in a DIY project.

Rather than use adapters I'm leaning to the ZTX851 discrete input for both the MC phono preamp and the DC-coupled preamp along with manual gain control.

There are multiple reasons.
One is the aforementioned difficulty of soldering QFN parts.
The second is the need to access nodes for the servo that aren't brought out in the 1580.
Manual gain control eliminates the requirement to galvanically-isolate the SPI control lines.
I want to keep the first iteration simpler.

I proved that it could be done using the 1570/5171 and you're welcome to run with what I've already laid out.
The one thing I did discover was the need to calculate and tweak servo Q.
The values I originally showed didn't peak badly but IIRC could be optimized without drastic changes.

ngineer
Posts: 3
Joined: Tue Apr 21, 2015 12:21 pm

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

Post by ngineer » Mon Aug 20, 2018 8:39 pm

Thanks for the reply. I've already dealt with the soldering problem by building a small reflow oven and have had good results with my first beta versions. Also the small circuit size is an advantage in my particular application. I'm looking hard at this idea for the next version and will let you know if I go forward. I spent much time in discussion with the factory about the phantom menace and was not completely satisfied with the solution. This is one of the reasons this idea appeals to me, it approaches the problem from the other direction, where impedances are higher and capacitors can be so much smaller. I'm wondering if the external differential servo gives enough of a performance boost to justify it's existence, was that your only reason for going that way? I need to go back and read all of the post, and will do that before moving forward.

User avatar
JR.
Posts: 2377
Joined: Sat May 24, 2008 7:21 pm

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

Post by JR. » Tue Aug 21, 2018 1:44 pm

For onsey-twosey prototyping I've soldered QFN packages with hot air rework wand...

BGA is on my too hard list, but I know people who say they've done them.

JR

User avatar
mediatechnology
Posts: 3208
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 » Tue Aug 21, 2018 4:42 pm

There were several reasons I used an external servo.

Originally I did my proof of concept using the 1510 so I had to use an external servo anyway.

I decided that if I was going to eliminate electrolytics on the input I should eliminate them in as many places as possible in the signal path.
That includes the servo.

Another goal was to have a passive input pole on the servo (thanks JR) to relax servo op amp bandwidth requirements and allow its selection to focus on DC characteristics.

The differential Deboo met both requirements in that it has a passive input pole and uses a single film cap.

The Deboo has an advantage over a conventional "two capacitor" differential integrator in that the response to step changes is monotonic and it uses one less film C.

Loose absolute tolerance electrolytics from the same lot usually ratio match pretty well so that hasn't been a factor using the 5171/5173 servos.
For a small form-factor design the 517X internal servo may be the way to go.
However, in a DC-coupled preamp we may need to use an unavailable node or be limited to servoing the input or input stage emitters so it may not be the best option.

As to soldering with either hot air or an oven I wish you guys the best of luck.

Last I heard THAT sends their QFN work out.
I worked with a manufacturer that was very experienced doing SMT prototyping and their hand-soldered yield was about 25% soldering the 5171.
I had Cimmaron give it a go repairing a 1571 adapter. That didn't work out which was OK because I didn't expect it to.
We had posts here on the Altoids mixer about half were intermittents due to soldering.

If you're eventually going to produce these in mass or just for yourself go for it.
If you're going to have it as a DIY project QFN soldering will create support hell.

ngineer
Posts: 3
Joined: Tue Apr 21, 2015 12:21 pm

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

Post by ngineer » Tue Aug 21, 2018 6:46 pm

Thanks for the reply. Your reasoning is very well thought out and I won't waste too much time re-inventing the wheel. I realize that I need to do a thorough review of the existing info and then ask questions instead of asking questions that have already been answered.

So far I've had good success with the smd devices (I did learn to never use a 401 size cap). When I started this development, I just assumed that I would have to learn to deal with the parts that were available, it does makes prototyping un-fun though.

User avatar
JR.
Posts: 2377
Joined: Sat May 24, 2008 7:21 pm

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

Post by JR. » Wed Aug 22, 2018 7:38 am

My first smd nightmare was some tiny x8 resistor arrays that the contract manufacturer couldn't even solder effectively (lots of solder bridges).

Some parts are only available in SMD package and just try to breadboard components with pads on the bottom. :oops:

But in general SMD works well..(I use lots of x4 resistor arrays with no problems).

JR

weroflu
Posts: 40
Joined: Wed Apr 19, 2017 9:19 am

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

Post by weroflu » Thu Sep 06, 2018 3:32 am

uhh just checking in re wayne's comment. I would not let my fail rate influence anyone, mainly because this was my first attempt at qfn. the 1580/5171 were not the cause of any problems. tldr for me was that i made the opposite error as most people in not putting enough solder on the pads. usually noobs are warned to not make the center qfn pad too high. once that was done my success rate was pretty high. also i reflowed these chips an ungodly amount of times with no apparent damage - of course this might reduce chip life but last i tested things were working fine.

if your boards are very small and ground plane not very large then hot air is good, otherwise you're going to need an oven or a preheater underneath.

bga was actually easier than qfn, using a hotplate i was 2/2

Post Reply