Discrete Transformerless Preamp?

GroupDIY Audio Forum

Help Support GroupDIY Audio Forum:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.
abbey road d enfer said:
CJ said:
tubes are discrete,  Western Electric made some low noise stuff,

what is the final product gonna be, MP3 for reverb nation or virgin vinyl?
Listened to on earbuds in urban transportation?...

Well fellows, he did say it was for recording "acoustic ensembles".  Perhaps there is still some audiophillia in classical recording.
 
I'm not really sure why we are giving the OP a hard time about his design goals?  He seems to have laid out a perfectly reasonable set of objectives to me.  His more subjective criteria are listed very last.

I've owned two of the "premium" transformer less mic pres listed in the thread, the Millennia Media HV3D and the GML 8304.  Both are reasonably low noise, low distortion circuits and yet one sounds musical and one doesn't IMO.  I have an inkling as to why the poorer performer is such but  I would need to dig in pretty deep to measure the differences.

Not every subjective assessment or goal is audiophile grade hand waving, sometimes it's just shorthand for imperfections or lack thereof that complement a given musical style.  The 1176 is full of distortion and IMO is a very musical tool, the Millennia mic pre has much much lower levels of distortion and is not musical (as I said I have a hunch and maybe I'll borrow one again to put it on the dScope).

Cheers,
Ruairi



 
ruairioflaherty said:
I'm not really sure why we are giving the OP a hard time about his design goals?  He seems to have laid out a perfectly reasonable set of objectives to me.  His more subjective criteria are listed very last.
I don't think we're giving him such a hard time; I think if it was the case, he would have raised a holler. We're also in this for having a laugh from time to time.
Now, if he has been offended, I do apologize.
 
JohnRoberts said:
You lost me at subjective, but got me back at flying rails. I did a lot of late night scribbling about that.

...

In a raise the bridge or lower the water exercise, it might be simpler to leave the mic preamp grounded and tweak the phantom power.  Namely drive the phantom power resistors with modest voltage, then drive pin 1 to a negative voltage using a servo to keep inputs at 0V.

Sorry to backtrack, but this concept from John Roberts is subtle and brilliant, but perhaps hard to deploy in a commercial device. The big problem is that a box of preamps might be asked to work in a context where the pin 1 voltage of each mike can't be reduced to achieve the 'P48 drive' to the attached mike, simply because pin 1 is common to a whole lot of other pin 1 connections in the room; each pin 1 can't be servoed like that.

However, for a DIY or specialized application where the user of such device might be able to arrange for each mike line to have its own isolated pin 1 connection, this is actually a brilliant solution to the P48 coupling problem. Further, servos like this are not tough to design, since their noise output will appear as common mode to the mike channel, easy to reject, and easy to 'finesse' by over-design.

Again, a tip of the hat for sharing this gem, which I know you could probably not use in your previous commercial works without generating endless support calls. However, here in this polite land, I think it's worth examining this idea, understanding that (DC) isolating pin 1 to each mike input is all that you need to make this work.
 
Monte McGuire said:
JohnRoberts said:
You lost me at subjective, but got me back at flying rails. I did a lot of late night scribbling about that.

...

In a raise the bridge or lower the water exercise, it might be simpler to leave the mic preamp grounded and tweak the phantom power.  Namely drive the phantom power resistors with modest voltage, then drive pin 1 to a negative voltage using a servo to keep inputs at 0V.

Sorry to backtrack, but this concept from John Roberts is subtle and brilliant, but perhaps hard to deploy in a commercial device. The big problem is that a box of preamps might be asked to work in a context where the pin 1 voltage of each mike can't be reduced to achieve the 'P48 drive' to the attached mike, simply because pin 1 is common to a whole lot of other pin 1 connections in the room; each pin 1 can't be servoed like that.
Not brilliant, but outside the box thinking.

Indeed pin 1 needs to be a low AC impedance to ground for effective shielding, even if biased at some negative DC voltage. This could be a nightmare for general applications where pin 1 grounds are rarely segregated.  Further a singer touching a mic body could likely get energized to that negative voltage potential, perhaps causing issues elsewhere.  The voltages and currents are not dangerous to human safety but could interfere with other sensitive inputs (phantom power involves mA level power).

However, for a DIY or specialized application where the user of such device might be able to arrange for each mike line to have its own isolated pin 1 connection, this is actually a brilliant solution to the P48 coupling problem. Further, servos like this are not tough to design, since their noise output will appear as common mode to the mike channel, easy to reject, and easy to 'finesse' by over-design.

Again, a tip of the hat for sharing this gem, which I know you could probably not use in your previous commercial works without generating endless support calls. However, here in this polite land, I think it's worth examining this idea, understanding that (DC) isolating pin 1 to each mike input is all that you need to make this work.

If I had a better handle on rolling my own very high performance A/D convertor, I would float the mic preamp and A/D up at whatever voltage it needs to be to keep the mic happy, and then optically couple the digital output.

Again I see this as mostly cosmetic (perception bias), since modern caps can be pretty clean when well used (i.e. keep signal currents low).

I scribbled plenty but never melted solder because I do not expect a significant audible improvement. There is probably a niche market among cap haters.  ;D

JR

PS: It was so long ago I don't even recall what I settled on. Probably two variable power supplies. A positive rail feeding the two 6.8K phantom resistors (0V to maybe +30V), and then a negative rail (0 to -48V) that when added to the positive rail sums to 48V total. So this would all be transparent to the mic. 

The servo would regulate the preamp input to be 0V, by varying either the + or - rail, while they would track each other with the 48V spread.  A non-phantom mic wouldn't drop any voltage in the phantom resistors so the + phantom rail would be 0V and pin 1 would be at full -48V DC. 

Have fun, this is mostly a science fair project. 
 
I've thought this some time ago, even once sorted the insulation of pin 1 to each mic the problem is that cases of the mics are connected to it's pin 1, I guess you see where this is going...

If a case of a mic touches the case of another mic with different phantom currents, they will be at different potentials, so you would be shorting them out. Let's say a dynamic touches another mic which takes rather high current, dynamic is taking nothing, the other will be having twice the current. If a mic get's in touch with a case connected to ground you run on a similar problem again.

So unless you paint all your mics on some insulating coating or modify them so the case isn't DC coupled to pin 1 it's becoming more usable. Even then (AC coupled case) when two mics touch you may expect some pop noise at the output.

I do like the idea of a floating input, there are some around. The idea of the floating ADC is really nice and the first time I hear about it.

JS
 
joaquins said:
I do like the idea of a floating input, there are some around. The idea of the floating ADC is really nice and the first time I hear about it.
It's quite common in instrumentation/measurement.
Studer, NTP and  several other nordic manufacturers made experiments in that domain.
 
abbey road d enfer said:
joaquins said:
I do like the idea of a floating input, there are some around. The idea of the floating ADC is really nice and the first time I hear about it.
It's quite common in instrumentation/measurement.
Studer, NTP and  several other nordic manufacturers made experiments in that domain.
It sounds logical, especially if the end point is digital.

JR
 

Latest posts

Back
Top