Print

Please login or register.
1 Hour 1 Day 1 Week 1 Month Forever
Login with username, password and session length

[donnie7]

Welcome Dan! Would like to discuss the signal path, Both at microphone level and line level. Also could you cover proper gain staging and and its importance in optimizing audio quality. Thank you all!

[danlavry]

“Welcome Dan! Would like to discuss the signal path, Both at microphone level and line level. Also could you cover proper gain staging and and its importance in optimizing audio quality. Thank you all!”

This is the first answer on this forum. It is long.  Let me know if it is too long, too basic, too complicated… So here we go:

Microphones put out little signal energy:

Dynamic mics (not powered) have the task of picking up air vibration over a small area (the membrane) and converting that to electrical power. With so little mechanical energy, the electric energy can not be any higher…

The condenser mic does receive some DC power (typically 48V phantom for pro mics). The DC is used for charging the “capacitive membrane” (which makes the condenser work), but it is also used for amplifying the signal output power.
But even the condenser mic is very limited. The energy supplied to the condenser mic is in the order of  magnitude of 100mW, and most of it is for DC biasing of some active circuitry. The signal output power peaks are rarely above few mW (milli Watts)

So far I am talking power, not voltage, because power is what counts. But we can now switch to the “normal” language – voltage. Say you get 1mW signal out of a condenser into a 250 Ohms AC load. It amounts to 0.5V signal. Of course getting a 1V signal out of a condenser takes loud sound and close mic placment. At feet away, soft sounds perhapse a dynamic mic, and the signal may be a few microvolt…

So we need obviously need to amplify the signal. Tiny analog signals are too delicate, and there are 2 reasons for that:

First, the circuits handling our tiny signals generate noise. Such self generating noise is “there all the time”. It is random noise. We want the noise to be very low, much lower than the signal. Lower noise is how we get closer to “noise free” signal.
Unfortunately, we can not amplify the signal only. The signal present at the input of a microphone preamplifier is made of both music and noise. The amplifier circuit does not know music from noise. It all gets amplified (treated the same way).
So the input stages of the micpreamp is a “specialty” circuit. It has to be designed to generate as low noise as possible.

Second, we need to deal with noise and undesirable electromagnetic pickup into cables, circuit board traces and all other physical components. A given noise interference has some power level, say 1mW. It can be “manifested” in many ways. A 1mW on a 1 KOhm yields 1V noise signal.  But 1mW on say 10 Ohms yields only 0.1V noise (20dB less voltage).

So we want low resistance for immunity from interference, high resistance for more voltage signal, low resistance for less self generating noise, proper resistance for maximum power transfer and so on… the “typical engineering set of tradeoffs”….

Another factor is: Balanced vs. unbalanced signal handling:

Unbalanced signal transmission is about sending some signal on some wire from a source driver to a load. The voltage level (signal) goes up and down with respect to some ground wire (or path). That is how most of electronics works. But long wires (an antenna in fact) and/or a high source and load resistance, pick up noise. You can help the situation some buy shielding the wire, and that is a “part way solution”. Picking say 10uV when the signal is 10V is OK (120dB signal to noise). But the same 10uV noise added to 100uV mic signal is only 20dB signal to noise… So what is good for line level may be terrible for mics.

Balance signal transmission is the method used for mic preamps. How does it work? You have 2 wires, one carries the signal, the other carries the inverted signal. If wire A (pin 2 of the XLR) goes up to +1mV, wire B (Pin 3 of the XLR) goes to -1mV and so on. The signal receiver subtracts the 2 signals so in our example we get +1mV – (-1mV) = 2mV signal. If we make sure the wires are parallel to each other (or even twisted around each other). Occupying the same space, they will act as “identical antennas” and pick up the same unwanted noise.
Say they both picked up 10mV noise at some instant of time. The wire A now has 1mV+10mV=11mV. Wire B will now have (-1mV) +10mV = 9mV.          
Again, the receiver circuit subtracts the signals A-B = 11mV – (-9mV) = 2mV.
So the “trick” worked. The noise disappeared. The outcome is the same. If we add the same unwanted signal to both wires, subtracting one from the other adds zero total. Yet the subtraction of a signal from its inverted mirror image yields doubling of the signal. Cool concept.

So back to your question: Mic signals vs. line level (a few volts), or we even AES/EBU analog signal levels (higher than line level): A balance transmission with a shielded cable is a must for tiny mic signals. The other particular issues for micpreamp signals handling (amplification) are extremely low self generating input circuitry, and a good differential amplifier (for the signal subtraction A-B). We call it CMRR (common mode rejection ratio). Common mode is the signal added to both inputs (A and B). Thus we call it named “common”.

Line level is easier to handle. With more signal, one can use noisier (self generating noise)    
circuits. Let me stop here for now. Give me some feedback (I can continue with more talk about line level issues, or get into more fine detail regarding what I touched on so far).

And lets start another thread on proper gain staging…  That is also a very good subject.    

BR
Dan Lavry

[donnie7]

Dan, Bravo and APPLAUSE! Very comprehensive answer. Thoughtful, precise and by all means not too long. Glad to have you answer the first question on your board. We welcome you with open arms. Keep up the wonderful work you are doing! Wishing you all the best that life has to offer! With Regards and Respect, Donnie Dixon.

[plughead]

Wow,

I gotta thank you as well, Dan: that explanation was so clear and informative, I'll be coming back to this forum often. Hopefully to better understand what the heck it is I'm trying to accomplish in this business...

Hat's off!

[Roland Storch]

danlavry wrote on Sat, 02 October 2004 01:56

So back to your question: Mic signals vs. line level (a few volts), or we even AES/EBU analog signal levels (higher than line level): A balance transmission with a shielded cable is a must for tiny mic signals.Dan Lavry

I have a question to the shielding of balanced interconnestion.

Dr. Jahne from StageTec claims that with balanced cables and (important) a common mode rejection better than 90 dB unshielded cables are better than shielded ones. What do you think about that?

[ted nightshade]

That's one for the archives. Should probably become a sticky! Great reference material and thanks Dan.

[danlavry]

[quote title=Roland Storch wrote on Sat, 02 October 2004 14:54]

danlavry wrote on Sat, 02 October 2004 01:56

I have a question to the shielding of balanced interconnestion. Dr. Jahne from StageTec claims that with balanced cables and (important) a common mode rejection better than 90 dB unshielded cables are better than shielded ones. What do you think about that?

I really did not understand the question. Can you please try to re word it?

BR
Dan Lavry

[Roland Storch]

I was refering to that article:
http://stagetec.com/stagetec/e_company_know_how39.htm

Jahne describes how you can have less input noise with unshielded cables instead of shielded cables.

[danlavry]

Roland Storch wrote on Mon, 04 October 2004 00:14

I was refering to that article: http://stagetec.com/stagetec/e_company_know_how39.htm Jahne describes how you can have less input noise with unshielded cables instead of shielded cables.

Oh well, I am not the "cable expert of the universe", yet my basic EE knowledge, and that certainly includes careful reading of the "bible" on the subject: "Grounding and Shielding for instrumentation" by Ralph Morrison, plus what the Maxwell Equations (the 5 fundamental laws of electricity and magnetism) suggest otherwise.
I quickly read the article you pointed out. I did not see any science in it, just a report about some specific experiment. I was not there to see what they did.
So what do they suggest? Are they saying that the whole world is best to remove shields? Including Ma Bell with over 100 years experience?

In any case, they talked about a dynamic mic. It would be good to see what the specific setup was. Did they ground both ends of the shield (that can be a cause for problems).

With the more commonly used condenser mics things WILL NOT WORK AT ALL without a shield connected at both ends. For condenser mics the shield act as a DC current path (ground return).
Does their theory suggest you will be better off to to replace the shield with a wire? I can assure you it would not be so for condensers!!!

BR
Dan Lavry    
 

[Terry Demol]

danlavry wrote on Mon, 04 October 2004 00:45

Roland Storch wrote on Mon, 04 October 2004 00:14 I was refering to that article: http://stagetec.com/stagetec/e_company_know_how39.htm Jahne describes how you can have less input noise with unshielded cables instead of shielded cables. Oh well, I am not the "cable expert of the universe", yet my basic EE knowledge, and that certainly includes careful reading of the "bible" on the subject: "Grounding and Shielding for instrumentation" by Ralph Morrison, plus what the Maxwell Equations (the 5 fundamental laws of electricity and magnetism) suggest otherwise. I quickly read the article you pointed out. I did not see any science in it, just a report about some specific experiment. I was not there to see what they did. So what do they suggest? Are they saying that the whole world is best to remove shields? Including Ma Bell with over 100 years experience? In any case, they talked about a dynamic mic. It would be good to see what the specific setup was. Did they ground both ends of the shield (that can be a cause for problems).

Dan / Roland,

Dr Jahne's (stagetec) statements are quite true but they really
apply specifically to Nexus or a similarly engineered system.

Looking at the technical specs the IO's on Nexus are specced as
transformer isolated. This tells a big part of the story.
Some important issues are:
- Good transformers have very high common mode rejection and
that rejection extends to quite high frequencies.
- Good transformers have excellent RF (radio frequency)
rejection to 'protect' the following active stages.
- Stagetec appears to have a special IP with HUGE CMRR and very
low ground coupling capacitance. They mention up to 150dB CMRR
and 10pF ground coupling c which to me suggests further active
electronics supplementing the transformer. As an example, an
active common mode bootstrapping arrangement could give these
types of figures and could be made to extend well into RF range
if done correctly, there are many ways to do it.

So don't throw away your shielded cables just yet, unless you
are using stagetec equipment. The only exception I would
say is a decent OP dynamic mic into a HQ transformer input
mic pre might be worth experimenting with a single run of
twisted solid core CAT-5 and comparing to your fave shielded
pro cable for grins. But if you hear a difference don't tell
anyone as there should be none

Looking through their website and from feedback from
users about their products it appears Stagetec have some
pretty lateral thinkers on board.

Cheers,

Terry

[danlavry]

[quote title=Terry Demol wrote on Mon, 04 October 2004 03:32]

danlavry wrote on Mon, 04 October 2004 00:45

Roland Storch wrote on Mon, 04 October 2004 00:14 I was refering to that article: http://stagetec.com/stagetec/e_company_know_how39.htm   Dan / Roland, Dr Jahne's (stagetec) statements are quite true but they really apply specifically to Nexus or a similarly engineered system. Looking at the technical specs the IO's on Nexus are specced as transformer isolated. This tells a big part of the story. Some important issues are: - Good transformers have very high common mode rejection and that rejection extends to quite high frequencies....Cheers, Terry

I am not questioning the fact the one can get better results without a shield in a specific system. However there is some reason for it. In my opinion, if they did it optimaly, a shield would do better. If what they say were correct, you can be sure that the whole world of technology would be "dealing with it big times".  

Again, that article is about a report for one case. So what? I can show 10 cases where the shield is very important. It will always come back to Maxwell's equations and their applications! A properly done shield will reduce the electromagnetic coupling, and that is why we have shields, and use them for more than a century, in audio, video, instrumentation, telephony....

Yes, transformers have good common mode characteristics. And better yet, a shielded cable and a transformer. Better yet is double shielded and a transformer, or double shielded cable plus a shielded transformer... So don't remove the shield for transformer coupled applications either.

BR
Dan Lavry