Trim line input comprehension

GroupDIY Audio Forum

Help Support GroupDIY Audio Forum:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.

matthieu68

Active member
Joined
Jul 11, 2017
Messages
25
Hello, I have been following the Forum for some time, very interesting.

I have already done several DIY projects, and I want to learn more, I bought a few books that deal with audio electronics, and to learn you have to do as they say.

I therefore embark on a first stage, that of the line input.

I had a desym PCB made some time ago which I think (I hope) is good.

I would like to put a trim on the line input, I would like to be inspired why not by the trim of the old MCI jh600 console (trim -6db / 0db / + 6db)

But in the original diagram the trim controls the deymetrization opamp, Is it possible to integrate a similar system after my deymetrization system?


This raises another question to which I don't have a clear answer,
An input is in + 4dbu, after the desym stage my signal should be identical or 1.228v, but I read here and there that the working level in a unit or mixer should be turned to -6db or 0.388v.

To make this fall what means is it used?
An inverter opamp?

It would therefore be necessary to put after my desym an inverter opamp with switchable resistors? to obtain respectively -12db / 0db (which will be the nominal level of the circuit is -6db) and + 6db

Thanking you
 

Attachments

  • Screenshot_2020-12-09-05-55-00-649_com.miui.gallery.jpg
    Screenshot_2020-12-09-05-55-00-649_com.miui.gallery.jpg
    531.3 KB · Views: 67
It looks like that circuit is debalancing and adjusting gain at the same time so I'm not sure what you're trying to do or fix. CMRR is probably not that great but none of the simple debalancing inputs are. Good CMRR needs to op amps. Or if you want hi-perf use a THAT1206 IC.

I don't love that MCI circuit because it looks like it might have some kind of negative impedance converter thing going on which doesn't make sense to me. You could just use the switch to adjust the negative feedback resistor which would be fine for -6, 0, +6 adjustment.

How you make an input depends on what your objective is. If you want to make some old vintage circuit, then just copy it exactly. Don't try to make some kind of weird permutation of some already weird circuit. If you want to make something that works, use a THAT1206 and follow that with a simple level control or a simple inverting gain control. Use good op amps with low-offsets so that you can skip the coupling caps and keep things low impedance for good noise performance.

A good book to learn about this sort of thing is Douglas Self's Small Signal Audio Design.
 
Hello, thank you for your answer.

Small audio signal is precisely the book I just got.

I am not necessarily trying to establish the MCI circuit, I had opted for this diagram because I could not find solutions / explanations on a variable trim on a line input.

Douglas's book self explains how to insert a line input into the stage of a preamp via a PAD, but this would amount to lowering the signal to re-increase it, not necessarily the best, right?

It also deals with the "self input variable gain"

I'm a little lost, what would be the best solution?  Because I would like to be able to subtract but also add gain (+/- db)

A simple potentiometer would only reduce the volume, right?

It would therefore be necessary to use a desym stage with that 1206 (which applies -6db at the output if I remember correctly) then follow it with an inverter aop for the trim?

Thank you
 
Desymetrization  :eek: Its usually called a differential amplifier because it takes the difference between input voltages, I guess you could see it like the signal simmetry is being lost, but desymetrization stage sounds weird. Anyway, Douglas Self proposes a circuit which uses an opamp and a pot in the feedback path of the diff amp to actively control the gain, you can use that circuit and adapt it for -/+ 6dB trim. I've built this circuit and it works as advertised.

http://www.douglas-self.com/ampins/balanced/balfig11.gif
 
Yes indeed I use bad terms, I will note it to myself not to make the mistake.

So I will look at this diagram, maybe with switchable resistors instead of a potentiometer for more precision and using the chip that 1206 as differential amplifiers  ;D to directly lower my signal.

For my curiosity, (leaving that 1206 chips out), in general what process is used to lower the signal?
An inverter opamp?
But it seems to me that the output signal will be inverting?  How to restore it?


I was able to observe some diagrams (unfortunately no schematic) where the trim remains on the path of the symmetrical signal, there too I was wondering ..., is there a way to use a trim  by remaining symmetrical, or is there necessarily a differential amplifier / trim / then a re Symetrization of the signal?


Thank you
 
matthieu68 said:
But in the original diagram the trim controls the deymetrization opamp, Is it possible to integrate a similar system after my deymetrization system?
First you have to understand how this particular one works. The switching in the schemo makes it a little difficult, but after all it's a very basic differential opamp. The gain of a differential opamp is governed by the ratio of the feedback resistors to the input resistors. In the MCI, the resistors are fmodified by the switch, which puts resistors in parallels with teh fixed ones.
This is very impractical to adapt for potentiometers.


This raises another question to which I don't have a clear answer,
An input is in + 4dbu, after the desym stage my signal should be identical or 1.228v, but I read here and there that the working level in a unit or mixer should be turned to -6db or 0.388v.
1.228V is +4dBu, thus turning it down 6dB makes it -2dBu, or 0.614V.

  To make this fall what means is it used?
Too many possible answers... Simple attenuatior (voltage divider), inverting opamp with feedback resistor being half the input resistor, 2:1 transformer...


It would therefore be necessary to put after my desym an inverter opamp with switchable resistors? to obtain respectively -12db / 0db (which will be the nominal level of the circuit is -6db) and + 6db
Do you really need to make it easily switchable? Very often I've seen DIP switches for that.
A THAT 120x is a very good solution in terms of CMRR but it's fixed gain (actually attenuation). If you want to use opamps, only one is necessary to provide both debalancing and gain/attenuation.
 
This raises another question to which I don't have a clear answer,
An input is in + 4dbu, after the desym stage my signal should be identical or 1.228v, but I read here and there that the working level in a unit or mixer should be turned to -6db or 0.388v.

To make this fall what means is it used?
An inverter opamp?

It would therefore be necessary to put after my desym an inverter opamp with switchable resistors? to obtain respectively -12db / 0db (which will be the nominal level of the circuit is -6db) and + 6db

It depends, for example if you are using a very simple electronically balanced output like driving pin 3 from the output of an inverter, then, there is a 6dB increase in gain so if you want to keep unity gain you need a differential amplifier with a gain of -6dB. If you are using a transformer balanced output or an impedance balanced output, there is really no need for attenuation at the input stage and you would want a diff amp with a gain of 0dB.

If your question is about the internal level of a mixer, thats a completely different question.

matthieu68 said:
For my curiosity, (leaving that 1206 chips out), in general what process is used to lower the signal?
An inverter opamp?
But it seems to me that the output signal will be inverting?  How to restore it?

Like abbey said, there are many ways to lower a signal, the simplest is a voltage divider, if you are using opamps you can either use an inverting configuration or a non inverting one. If you use an inverting amplifier and you want to restore the signal to its original polarity, then you must use another inverting amplifier, but in general, it is not good practice to invert and then add another opamp just to re-invert, there are some exceptions in which this is acceptable.
 
Hello, yes I was wondering about the internal volume of a mixer.

My goal would be little by little to already understand and draw a channel strip (.... Then the rest) this would I think is a good exercise to understand all the stages and the impedance ...

Line input / trim / insert / eq type baxandall / fader / aux ....


I wonder ... If the level inside a mixer needs to be lowered, why not just use a potentiometer for the trim on the same principle as the faders?

For the faders if I have understood correctly, we insert an amplifier after the fader to obtain the 10db boost, then lower the fader to obtain the normal internal level.

The signal will already be amplified compared to the level we want in the mixer ...
We arrive at + 4dbu, it would be enough on the panels to indicate where is the unit gain, the potentiometer or the potentiometer delivers us the desired level which we want to use in the console?

We could therefore amplified up to our + 4dbu and attenuate?


Or my idea looks more like crappy DIY?
 
If you want to see good complete schematics of real-world modern channel / mixer designs that are solid (and very consistent with D. Self's circuits incidentally) look at the schematics for the Mackie 1402 VLZ3 series available on mackie.com (or any 12+ channel mixers of that series).
 
matthieu68 said:
I wonder ... If the level inside a mixer needs to be lowered, why not just use a potentiometer for the trim on the same principle as the faders?
It is often done that way.

For the faders if I have understood correctly, we insert an amplifier after the fader to obtain the 10db boost, then lower the fader to obtain the normal internal level.
Correct.

  The signal will already be amplified compared to the level we want in the mixer ...
Not necessarily. The input gain and/or trim are tehre to make sure the level is nominal (nominal is whatever the designer thinks is adequate).

We arrive at + 4dbu, it would be enough on the panels to indicate where is the unit gain, the potentiometer or the potentiometer delivers us the desired level which we want to use in the console?
Do you mean the signal connected at the input is always a good solid reliable +4dBu? Wake up!  ;)
 
matthieu68 said:
I wonder ... If the level inside a mixer needs to be lowered, why not just use a potentiometer for the trim on the same principle as the faders?

For the faders if I have understood correctly, we insert an amplifier after the fader to obtain the 10db boost, then lower the fader to obtain the normal internal level.

You may want to keep an eye on your gain-staging. Regarding your input: with a really loud input signal you might overload your first input stage before you pull the level down with a pot or voltage devider. The same applies to the fader: boost before the fader reduces the headroom of the signal chain.

Michael
 
Hello.  signal overload should not be a real problem since the utility of my trim would be used to lower the signal to the line level of the console (-2dbu?) my first idea of ​​wanting a differential aop with gain adjustment, n  after all, this isn't the best idea ... putting a voltage driver just after the differential AOP would allow me to lower the input to the desired level, and why not even put several switchings for several stages.  In the selfe book he mentions this, and adds if I remember correctly, that adding a buffer before the voltage drive helps keep the noise low. 
Do you think that if the voltage driver (or Switch pad) is positioned after the differential aop this would also help to have less noise?

Hey, and good christmas
 
matthieu68 said:
Hello.  signal overload should not be a real problem since the utility of my trim would be used to lower the signal to the line level of the console (-2dbu?)
  Pro audio line level is almost universally +4dBu. dB takes a capital B.

  In the selfe Self book he mentions this, and adds if I remember correctly, that adding a buffer before the voltage drive helps keep the noise low. 
I don't think that's what he says, because that would be incorrect.

  Do you think that if the voltage driver divider (or Switch pad) is positioned after the differential aop this would also help to have less noise?
Naah, don't hink so...
You have to understand how noise builds up.
The noise produced by the 1st stage is added to the 2nd stage, which in turn amplifies the resulting noise, and so on along the subsequent stages.
Usually, when the input signal is small, most of the gain comes from the 1st stage, so that the noise of the subsequent stages is almost negligible.
But when the input signal is already high, the contributions of each stage must be taken into account; then, if all the sages have similar gain, noise increases roughly according to the square root of the number of stages.
 
matthieu68 said:
Hello.  signal overload should not be a real problem since the utility of my trim would be used to lower the signal to the line level of the console (-2dbu?) my first idea of ​​wanting a differential aop with gain adjustment, n  after all, this isn't the best idea ... putting a voltage driver just after the differential AOP would allow me to lower the input to the desired level, and why not even put several switchings for several stages.  In the selfe book he mentions this, and adds if I remember correctly, that adding a buffer before the voltage drive helps keep the noise low. 
Do you think that if the voltage driver (or Switch pad) is positioned after the differential aop this would also help to have less noise?

Hey, and good christmas
I don't know if this is what you are asking (or what an AOP is).

For best S/N you want to put most of your gain in the front end. Gain in later stages also amplifies noise from the previous stages. So best practice is to have as much gain as foreseeably needed in the first stage.

I agree with -2dBu as a practical nominal internal operating level.

The next gain stage is the channel fader that routinely provides up to +10dB. The gain is mainly for use in mixing to easily balance levels between channels. It is also common to see another up to +10dB available in the master fader, mainly for managing final output levels.

Left over from the old school analog days when recording media would overload gracefully, most metering is referenced to a nominal 0VU level approximately 20db below saturation. Modern digital media is (or was) much more overload sensitive so metering is referenced from max level  (0dBFS) downward.

Hope this helps.

JR
 
 
thank you, yes I can see it more clearly.  in fact I think it is useless to be too complicated, the use of a voltage driver after the differential opamp would be enough to lower my level to -2dbu.

I suppose it would be possible to have a voltage driver with several switchable outputs.  maybe you can help me, I am looking for the calculator to calculate the resistances that it would take in the case of a voltage driver with several outputs, if I know the input voltage and the desired voltages for the different  exits?

I take this opportunity to be with you to ask you another questions, an opamp is applied after the faders to generally apply a gain, but especially to be able to control many buses and auxiliaries.  In the event that I don't want to have any gain on my fader (ie just attenuation) my opamp should be made as a buffer and should be inverted to be able to obtain a unit gain?  or is it possible to obtain a unit gain with a non-inverting opamp?
 
matthieu68 said:
I am looking for the calculator to calculate the resistances that it would take in the case of a voltage driver with several outputs, if I know the input voltage and the desired voltages for the different  exits?
This is a basic case of voltage divider. The formula is Vout/Vin=R2/(R1+R2). R1 being the series resistor and R2 being the shunt. There are many calculators for voltage dividers on the web. Just google it.

  or is it possible to obtain a unit gain with a non-inverting opamp?
That is a voltage-follower, where there is a direct connection from the output to the inverting input. It's a particular case of the non-inverting stage where Rf=0 and Rb=infinity, Rf being the resistor between output and inverting input, Rb being the resistor from inverting input to ground.
 
yes indeed I have already seen this calculates.  But hey, I remember this was what was mentioned in Self's book, adding a voltage follower after the voltage driver would help keep the noise down.

  For the voltage followers would it have a negative aspect to be used in a console, just after the fader to control the buses and auxiliaries? 
I mean, does an opamp used as a voltage follower keep the same noise properties etc as if it was used as an inverted buffer for example

thanks again
 
matthieu68 said:
But hey, I remember this was what was mentioned in Self's book, adding a voltage follower after the voltage driver would help keep the noise down.

I'm aware of Self's bit about preceding a voltage divider with a buffer stage to obtain lower noise.
In "Small Signal..." I think it's under the section - "How to attenuate quietly" or similar. In my edition at least.
But the point there is that it's applicable where you would otherwise need to use high value resistors to avoid loading, say, a high impedance source or a circuit where a lower impedance would be a problem.
So in a case where you would need, for example, a divider made up of two 100K resistors - you could buffer the signal then drive that into a divider consisting of eg two 1K resistors.
So the point is that the noise from the amplifier is less than the noise advantage gained from using lower resistor values.
Adding a buffer where you not changing resistor values just gets you more noise.
Obviously you need to calculate at what point you would gain taking into account resistor values and the specific op amp spec'.

 
matthieu68 said:
yes indeed I have already seen this calculates.  But hey, I remember this was what was mentioned in Self's book, adding a voltage follower after the voltage driver would help keep the noise down.
Obviously you haven't read everything there...

 
For the voltage followers would it have a negative aspect to be used in a console, just after the fader to control the buses and auxiliaries? 
I mean, does an opamp used as a voltage follower keep the same noise properties etc as if it was used as an inverted buffer for example
An opamp has generally better noise performance as a non-inverting stage than an inverting one.
The demonstration involves some mathematics that you are not yet in a position to grok.
The particular case of the voltage-follower is the one that has the best noise performance.
 
Back
Top