Hi, I'm looking at building a microphone preamp with balanced and unbalanced line outputs.
My question is: if I build the line output stage circuit to provide +4 dBu (1.736 V peak) signal and pass this through a unity gain balanced output opamp stage, the balanced output Hot and Cold lines will each be +-1.736 Vpeak and the receiving equipment balanced input differential voltage will be twice +4 dBu. Is this OK or should I build my output stage circuit to provide only 0.868 Vpeak to the balanced output opamp stage so that the voltage differential delivered to the receiving equipment will be +4 dBu (1.736V peak)?
Thanks in advance for any advice you can give.
My question is: if I build the line output stage circuit to provide +4 dBu (1.736 V peak) signal and pass this through a unity gain balanced output opamp stage, the balanced output Hot and Cold lines will each be +-1.736 Vpeak and the receiving equipment balanced input differential voltage will be twice +4 dBu. Is this OK or should I build my output stage circuit to provide only 0.868 Vpeak to the balanced output opamp stage so that the voltage differential delivered to the receiving equipment will be +4 dBu (1.736V peak)?
Thanks in advance for any advice you can give.
+4dBu is the nominal 0VU output level (across both outputs) and you'll want to allow for at least 20dB of headroom above this.
All good fortune,
Chris
All good fortune,
Chris
Hi Chris, thanks for your quick reply.
I'm still unsure what I should be doing. Would you mind clarifying a bit more please? When you say +4dBu is 0VU output level (across both outputs) does this mean for a balanced output the peak voltage on hot and cold should be 1.736v or 0.868v?
And then you say I need to allow another 20db of headroom. So by that do you mean I need to allow the output to be varied up to 10 time the +4dBu voltage level?
I know I'm probably not seeing the obvious here. Thanks for your help.
Sean.
I'm still unsure what I should be doing. Would you mind clarifying a bit more please? When you say +4dBu is 0VU output level (across both outputs) does this mean for a balanced output the peak voltage on hot and cold should be 1.736v or 0.868v?
And then you say I need to allow another 20db of headroom. So by that do you mean I need to allow the output to be varied up to 10 time the +4dBu voltage level?
I know I'm probably not seeing the obvious here. Thanks for your help.
Sean.
The +4dBu "standard" for prosound connections is for a VU meter reading of 0VU. These are averaging meters, with ballistics chosen to best match apparent volumes of different sources to each other, for use in a radio station, for example. The connections have been carried over into modern pro audio use because they're robust, (and because they were already there!).
This is very different from a modern digitally sourced "standard" like 2 volts output from a CD player. Digitally derived levels are already peak outputs, and no headroom need apply.
Averaged, damped numbers need to be allowed some headroom for signal peaks. It's common for professional use to allow 20dB, or a little more if possible, headroom for recording, or often a good bit less for sound reinforcement. Good homebrew design means a +24dBu output capability. This can be split between both polarities (+18dBu and +18dBu) if you know that both will be used by the receiving end, and your metering is across both polarities (or scaled as if it is).
All good fortune,
Chris
This is very different from a modern digitally sourced "standard" like 2 volts output from a CD player. Digitally derived levels are already peak outputs, and no headroom need apply.
Averaged, damped numbers need to be allowed some headroom for signal peaks. It's common for professional use to allow 20dB, or a little more if possible, headroom for recording, or often a good bit less for sound reinforcement. Good homebrew design means a +24dBu output capability. This can be split between both polarities (+18dBu and +18dBu) if you know that both will be used by the receiving end, and your metering is across both polarities (or scaled as if it is).
All good fortune,
Chris
Last edited:
Assuming both outputs are driven by opamps at the same level on each line, then for +4dBu balanced line level, the peak voltage w.r.t ground on each line will be 0.868Vpk.
Normally we don't talk about ground referenced levels on balanced lines, because they could be floating or driven single ended. Talking about peak voltage can cause confusion, just use Vrms or dBu.
You don't need variable gain to crank the normal operating level up another 20dB, if thats what you mean. Headroom is max output level above operating level before clipping. If you use the +4dBu balanced and -2dBu internal unbalanced levels you'll have enough headroom.
It may be useful to think about VU based "standards" by analogy to another averaging measurement, Sound Pressure Level. SPL is most commonly measured with an averaging meter, and we all know that peaks will be somewhere above the metered average.
Same-same for VU based measurements, and exactly as much headroom will be needed for any particular sound source. For example, a drummer hitting a snare hard but slowly will need very large headroom over measured SPL, an organ drone tone will need comparatively little.
All good fortune,
Chris
Same-same for VU based measurements, and exactly as much headroom will be needed for any particular sound source. For example, a drummer hitting a snare hard but slowly will need very large headroom over measured SPL, an organ drone tone will need comparatively little.
All good fortune,
Chris
What about for mic preamps with variable gain and/or trim?
Also, WRT to the OP's preamp, is the +4dBu output referenced to a particular standardized mic input level? (different mic's have different outputs)
Also, WRT to the OP's preamp, is the +4dBu output referenced to a particular standardized mic input level? (different mic's have different outputs)
*All* general-purpose mic preamps must have variable gain, or trim, to optimise their own internal dynamic range. A more general name for this is "gain staging". For a homemade mic preamp, and for use in a particular setting, with a particular microphone, that rule could be relaxed.
But this is a separate issue from output level and how it's specified.
All good fortune,
Chris
But this is a separate issue from output level and how it's specified.
All good fortune,
Chris
OK, but what I mean is that the gain adjustment range has to be such that +4dBu can be achieved at the output for all anticipated mic types, is that correct?
Yes, required headroom is variable with source. Trouble is VU and other metering standards (only PPM?) are well defined in terms of response time (ballistics) and we're stuck with them. I think best solution is LED clipping indicators. I took it for granted that a gain control will be used.
Hang on a minute; are VU meter ballistics well defined ?
ISTR reading specs for it years ago. I do know PPM is well defined.
I think we've got a bit OT - the OP was only asking about the line output stage of his mic pre.
Hang on a minute; are VU meter ballistics well defined ?
ISTR reading specs for it years ago. I do know PPM is well defined.
I think we've got a bit OT - the OP was only asking about the line output stage of his mic pre.
Yes, and it's pretty challenging! The possible SPL's at the microphone can cover a pretty large range, and mic sensitivities another pretty large range. You almost can't get there from here.
One elegant approach is to feed the two halves of the input differential amplifer with two identical current sources, and to put a gain control between the halves' emitters/sources/cathodes, variably degenerating input transconductance.
All good fortune,
Chris
Note however that that requires an annoying 'reverse log taper' pot if you want a gain control linear in dB.
Personally I tend to feel that switched gains in 6dB steps will usually let you get close enough, and the repeatability is useful for things like MS and A format recordings.
Don't be afraid to split the gain over two stages if it makes sense, a more or less fixed 20 or 30dB at the front end (With a switched 20dB or so pad in front of it) will deal with the noise, followed by the usual highpass filter and then the rest of the gain has something to be said for it.
One other thing to be very careful of in mic amp designs is the charge on the phantom blocking caps (And what happens to it when a cable short happens...), smoked e/b junctions are not unheard of, and if you make the series resitance as large as you really want to the thermal noise gets you.
Regards, Dan.
Personally I tend to feel that switched gains in 6dB steps will usually let you get close enough, and the repeatability is useful for things like MS and A format recordings.
Don't be afraid to split the gain over two stages if it makes sense, a more or less fixed 20 or 30dB at the front end (With a switched 20dB or so pad in front of it) will deal with the noise, followed by the usual highpass filter and then the rest of the gain has something to be said for it.
One other thing to be very careful of in mic amp designs is the charge on the phantom blocking caps (And what happens to it when a cable short happens...), smoked e/b junctions are not unheard of, and if you make the series resitance as large as you really want to the thermal noise gets you.
Regards, Dan.
Thanks EssB, your answers are what I was looking for. I fully expect to provide internal gain controls up to +4dBu on the output, I just needed to know on each line of a balanced output what the peak signal voltage would be 1.736 or 0.868, which you've answered.
I've found in my simulations that using a mic input transformer provides a really big advantage when it comes to S/N ratio.
Also, many musicians want their mic preamps to be FX boxes also, based on conversations with them. I think it's only the classical music folks who want "true to life" "zero distortion" electronics, but I don't know that for a fact. Most other musicians want "warmth" in their sound, so that means making the electronics non-linear in a pleasing way....
Also, many musicians want their mic preamps to be FX boxes also, based on conversations with them. I think it's only the classical music folks who want "true to life" "zero distortion" electronics, but I don't know that for a fact. Most other musicians want "warmth" in their sound, so that means making the electronics non-linear in a pleasing way....
That may be Chris. I was asking what the peak voltage of the signal should be on each line of a balanced line output at +4 dBu. I know +4dBu peak voltage is 1.736 v on an unbalanced line output but if this signal is used to drive a unity gain opamp balanced line output, the receiving equipment would effectively see the voltage difference between Hot and Cold which would be twice the original +4 dBu signal. Which concerned me. Hope this makes sense.
dmills - Thanks for the advice and I'm pleased you said this as I'm planning for a variable first stage giving 10 - 20 dBu gain with pad switch, followed by filter stage then final gain stage and line output. I'll also take in to account your warning on phantom power. What would you suggest?
Regards
Sean.
If an input transformer is providing a noise advantage then you have not optimised the curent noise Vs noise voltage tradeoff for the notional source impedance.
They can be the way to go, if you must have a gain stage with low current noise and lots of voltage noise combined with a low Z source, but generally a better design tradeoff in the input stage will be cheaper and just as effective.
Agree aout the confusion between gain stage and effects box, it is weird as even the 'shiny' stuff only really exhibits interesting effects when absolutely driven flat out.
Regards, Dan.
They can be the way to go, if you must have a gain stage with low current noise and lots of voltage noise combined with a low Z source, but generally a better design tradeoff in the input stage will be cheaper and just as effective.
Agree aout the confusion between gain stage and effects box, it is weird as even the 'shiny' stuff only really exhibits interesting effects when absolutely driven flat out.
Regards, Dan.
I meant an input transformer with a step up ratio, which basically provides "noise free" gain.
I'm not sure how you can optimize any circuit that's supposed to accept a fairly wide range of source impedances. I'm not skilled enough to do something like that.
I'm not sure how you can optimize any circuit that's supposed to accept a fairly wide range of source impedances. I'm not skilled enough to do something like that.
- Status
- Not open for further replies.