Print

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

[Andy Simpson]

With all the fascinating talk about negative feedback over in the other thread, I have been brought back to a question I posed several months back, which was not really answered (to my memory!).

Anyway, over the last couple of years I have really started to feel like the more pre-gain I use, the more I like the sound. To the point where I much prefer the sound of my dynamic mics over my condensers (none of which are high-end, admitedly).

Now I suspect (from a non-electronics designer position) that my gain pot is directly linked to the amount of negative feedback in my circuits? Which is why I prefer using more gain....?

Could this be a possible reason why ribbons produce such real sounding recordings - because the pre is always wide open...?

Also, does this negative feedback principle apply equally to the average hifi amp? - ie. the wider open you run it, the better it sounds? Which would be a good case for quieter recordings.....?

Anyway, if this is a particulary stupid and ill-informed question, I appologise in advance!



Andy

PS, I have also come to feel like if I run my monitor amps wide open and my levels way low, I much prefer the sound....despite the added noise levels....

[compasspnt]

Andy,

A real expert will have to answer re: the negative fedback.

However, re: your monitors, you don't want "a sound you like better," you want the most accurate sound you can obtain, so that you know where you stand.  I would be wary of that.

[Andy Simpson]

But of course!

However, my theory is that the higher negative feedback at lower gain levels is giving false information (ie. making it sound worse than it truly is).....so it's not so much that it sounds better at wide-open gain, as it sounds worse at low gain (high nfb?).
high nfb = innaccurate....?

And, my point about general level still stands - that the classic records sound better on consumer systems because they're quieter and allow the hifi amp to be openned up much more for the same percieved listening levels, with less nfb.....
Example....on my kitchen boombox, californication is 'party loud' at ~1/4 volume. (painful loud).
However, tapestry or innervisions is 'party loud' at ~1/2 volume and sounds soooo much bigger, fills the room. Lovely.

I'm assuming ya'll don't run your monitor amps really low and crank the desk/monitor controller? If you did, it might sound worse/less in control......?

Andy

....somebody is bound to pop up right about now and tell me that the volume control is in no way connected with nfb and that I am barking up the wrong tree.....

[Yannick Willox]

I always thought gain in a standard amp is fixed, the volume control only attenuating the signal.
So there would be no link between the volume control and NF.

Maybe there are some exotic designs where the pot controls overall gain (a certain mic pre comes to mind, Gordon audio maybe ?)

I think the quality loss you hear is linked to the quality of the pot - or inaccuracies in volume (more - louder - better !)

[danlavry]

Yannick Willox wrote on Fri, 17 June 2005 07:59

I always thought gain in a standard amp is fixed, the volume control only attenuating the signal. So there would be no link between the volume control and NF. Maybe there are some exotic designs where the pot controls overall gain (a certain mic pre comes to mind, Gordon audio maybe ?) I think the quality loss you hear is linked to the quality of the pot - or inaccuracies in volume (more - louder - better !)

First, I see a lot of factors being thrown in the last posts -too many variables to be able to come up with conclusive statements.  

Regarding the comment about gain being always fixed, with a pot doing attenuation, it is not always the case. But one way or the other, we can talk about negative feedback in a fixed gain stage, or in a variable gain stage. The negative feedback loop (or loops) is an arrangement were one feed a portion of the amplified signal to be subtracted from the input, and in fact, by introducing huge amount of gain to the difference between the input and feedback signals, each loop is able to better correct for all the non linearity within the loop.

So if your amp has say 100dB gain (at some frequency), and you wish to amplify the signal by 10dB, you have 90dB gain left for correcting distortions. But if you wish for 50dB signal gain, you have only 50dB gain "leftover" for the correction. The theory says that infinite gain yields perfect results from signal transparency standpoint. But we can not have stable operation with infinite gain. In fact, at 120dB stability gets to be tough, though still practical... In practice, a given circuit withe some given non linearity will yield better results (less distortions) when you raise the open loop gain... The practice does follow the theory very closely!

Regards
Dan Lavry
www.lavryengineering.com        

[Terry Demol]

Yannick Willox wrote on Fri, 17 June 2005 07:59

I always thought gain in a standard amp is fixed, the volume control only attenuating the signal. So there would be no link between the volume control and NF. Maybe there are some exotic designs where the pot controls overall gain (a certain mic pre comes to mind, Gordon audio maybe ?) I think the quality loss you hear is linked to the quality of the pot - or inaccuracies in volume (more - louder - better !)

Generally speaking, most current mic pre's use a volume control
that adjusts the amount of feedback. There are exceptions, but
most opamp (discrete or monolythic) based designs use this
approach.

The Gordon audio pre is probably different from any current
pre available from what I can gather from GC's posts on
Gearslutz. It uses NO global or interstage feedback
(as do our designs) but Grant appears to have optimised
operational parameters for each and every gain setting.

It appears this pre is quite different from most and would
really require Grant C to comment himself to prevent
misinformation. He seems like a very smart guy to me.

It also appears his design is very very good (sonically)
from users reports.

Regards,

Terry

[David Satz]

Yannick, I've tested many different mike preamps; they've all had at least one active stage before any gain control. As a result, the preamp has a limit on the highest voltage it can accept coming in, regardless of what the preamp is putting out.

I've noticed that a lot of people don't seem to be aware of this situation. Most spec sheets and equipment reviews avoid any mention of it. Also, the overload indicators on most preamps almost never respond to input overload--they're only connected to the output stage(s).

That's dangerous because different studio microphones may have a 60:1 range of sensitivity. With some preamps, the inputs can handle any signal that you're ever likely to get from the highest-output condenser microphones; also, some preamps have such high basic gain that the outputs will always go into clipping before the inputs ever do. But those are the exception rather than the rule. A preamp may have lots of "headroom" with one type of microphone, but it may suddenly run out of "headroom" completely if you switch to a more sensitive (higher-output) microphone.

--best regards

[Andy Simpson]

This is interesting.
Perhaps, I prefer my volume pots wide open. That would make sense.
Are there any volume pots which sound as good at -30dB as at 0dB?

Whatever the case, presumably, in pre-amps, running the gain very low is either applying alot of nfb or attenuating the signal.
Both are almost certain to sound worse, n'est pas?

Which, if true, would be a good reason to use ribbons & dynamics over condensers. (not counting the additional headroom consideration!).

Andy

PS. maybe I should get some lower gain tubes for my pres to run my drum overheads at full gain - at present I get them up to about 45-50dB gain with dynamic mics for overhead.

[Jim Williams]

Terry Demol wrote on Sat, 18 June 2005 03:18

Yannick Willox wrote on Fri, 17 June 2005 07:59 I always thought gain in a standard amp is fixed, the volume control only attenuating the signal. So there would be no link between the volume control and NF. Maybe there are some exotic designs where the pot controls overall gain (a certain mic pre comes to mind, Gordon audio maybe ?) I think the quality loss you hear is linked to the quality of the pot - or inaccuracies in volume (more - louder - better !) Generally speaking, most current mic pre's use a volume control that adjusts the amount of feedback. There are exceptions, but most opamp (discrete or monolythic) based designs use this approach. The Gordon audio pre is probably different from any current pre available from what I can gather from GC's posts on Gearslutz. It uses NO global or interstage feedback (as do our designs) but Grant appears to have optimised operational parameters for each and every gain setting. It appears this pre is quite different from most and would really require Grant C to comment himself to prevent misinformation. He seems like a very smart guy to me. It also appears his design is very very good (sonically) from users reports. Regards, Terry

Most modern preamps don't vary the feedback, but vary a gain shunt resistor off of the first amplifier stage. The Trident 80 preamp is an exception as it does vary feedback resistance.

I would like some input to the Gordon design as well. The web page shows specs but omits THD+noise, SMPTE IMD, CCIF IMD, DIM, FFT plots and any other distortion measuring techniques. For such an expensive piece to omit such basic measurments is disturbing, if not misleading.

Just what are we paying for here?
Inquiring minds want to know...

Jim Williams
Audio Upgrades

[David Satz]

andy_simpson, a much simpler way of trying out your idea would be to get some balanced, in-line 15 dB resistive pads such as the Shure A15A, and using them at the preamp inputs whenever you plug condenser mikes in. Then you can open up the gain controls by 15 dB to see whether you like the sound any better.

Pads like those should be part of every engineer's kit anyway, in my opinion, because they can be lifesavers when a preamp's inputs are clipping (if the microphones themselves aren't being overloaded). The pads don't interfere with phantom powering, and the Shure units that I've measured have all had extremely well matched resistors where it counts, so they don't degrade a preamp's ability to reject common-mode interference.

--best regards

[Andy Simpson]

David Satz wrote on Tue, 21 June 2005 16:33

andy_simpson, a much simpler way of trying out your idea would be to get some balanced, in-line 15 dB resistive pads such as the Shure A15A, and using them at the preamp inputs whenever you plug condenser mikes in. Then you can open up the gain controls by 15 dB to see whether you like the sound any better. Pads like those should be part of every engineer's kit anyway, in my opinion, because they can be lifesavers when a preamp's inputs are clipping (if the microphones themselves aren't being overloaded). The pads don't interfere with phantom powering, and the Shure units that I've measured have all had extremely well matched resistors where it counts, so they don't degrade a preamp's ability to reject common-mode interference. --best regards

I may well do that.....is there any degredation in signal with said line pads?

I'm inclined to try some lower gain tubes just for interests sake anyway, but I can see plenty of uses for some pads like those, if they don't kill the sound.

Cheers.

Andy

PS, if I may be so bold as to offer an opinion on the intriguing gordon pre's and what they do for the money, I understand from the docs that every single gain stage is basically hard-wired and optimized, rather than doing a one-size-fits-all variable kind of approach.
5dB gain steps, I think I read.
I don't know if it's basically 14 discrete circuits - one for each gain level or something similarly 'discrete' - but that's my guess - and what an extravagently simple concept!!!

Fuckit, I'd like a seperate custom pre for every gain level I'll ever need, at 2dB intervals, in sets of 16, in an ideal world.
That's 16x40 pre's in my rack, baby, and no gain pots at all!
Imagine the sound!

[Jim Williams]

David Satz wrote on Tue, 21 June 2005 08:33

andy_simpson, a much simpler way of trying out your idea would be to get some balanced, in-line 15 dB resistive pads such as the Shure A15A, and using them at the preamp inputs whenever you plug condenser mikes in. Then you can open up the gain controls by 15 dB to see whether you like the sound any better. Pads like those should be part of every engineer's kit anyway, in my opinion, because they can be lifesavers when a preamp's inputs are clipping (if the microphones themselves aren't being overloaded). The pads don't interfere with phantom powering, and the Shure units that I've measured have all had extremely well matched resistors where it counts, so they don't degrade a preamp's ability to reject common-mode interference. --best regards

Hello David,
 Just what value resistors are they using in those pads? I would expect something like 1k resistors in series with a smaller value to set the attenuation. Like the problem you had powering the Schoeps mics from the dbx 760x, the current drop through those resistors could lower the current to the mic, possibly to the level it would affect high draw mics like schoeps. Any insight on this?

[Terry Demol]

Jim Williams wrote on Tue, 21 June 2005 15:56

Terry Demol wrote on Sat, 18 June 2005 03:18 Yannick Willox wrote on Fri, 17 June 2005 07:59 I always thought gain in a standard amp is fixed, the volume control only attenuating the signal. So there would be no link between the volume control and NF. Maybe there are some exotic designs where the pot controls overall gain (a certain mic pre comes to mind, Gordon audio maybe ?) I think the quality loss you hear is linked to the quality of the pot - or inaccuracies in volume (more - louder - better !) Generally speaking, most current mic pre's use a volume control that adjusts the amount of feedback. There are exceptions, but most opamp (discrete or monolythic) based designs use this approach. The Gordon audio pre is probably different from any current pre available from what I can gather from GC's posts on Gearslutz. It uses NO global or interstage feedback (as do our designs) but Grant appears to have optimised operational parameters for each and every gain setting. It appears this pre is quite different from most and would really require Grant C to comment himself to prevent misinformation. He seems like a very smart guy to me. It also appears his design is very very good (sonically) from users reports. Regards, Terry Most modern preamps don't vary the feedback, but vary a gain shunt resistor off of the first amplifier stage.

The gain shunt resistor is in the FB loop, otherwise the IP
stage would overload at high gains.

Quote:

The Trident 80 preamp is an exception as it does vary feedback resistance. I would like some input to the Gordon design as well. The web page shows specs but omits THD+noise, SMPTE IMD, CCIF IMD, DIM, FFT plots and any other distortion measuring techniques.

Which will not tell a whole lot about what it will sound
like. For the record, not many mic pre manufacturers give much
of this info, usually just a THD and that is usually at a
lowish gain to make it look better.

One exception I can think of is Fred Forssell who has
quite comprehensive specs and plots on most his products.

Quote:

For such an expensive piece to omit such basic measurments is disturbing, if not misleading. Just what are we paying for here? Inquiring minds want to know...

Sound. Read some of the reviews of this box. From what I can
gather it is somewhere near if not at the top of the class.

Maybe you should try and audition one before casting judgement.

Also, FWIW, at what gain are the specs on your pre's?

Cheers,

Terry


 

[ML]

"Anyway, over the last couple of years I have really started to feel like the more pre-gain I use, the more I like the sound. To the point where I much prefer the sound of my dynamic mics over my condensers "

very few preamps exibit identical sonic qualities throughout their entire gain range. there are many reasons for this which i'm sure dan will be able to explain better than me, however, having designed and built mic pre's for my thesis on the subject, it quickly became apparent that my early attempts had wildly differing characteristics for high, mid and low gain settings (both high and low gain giving poor performance and mid gain settings producing very good performance)

although i've never used prism sound's Maselec MMA-4 mic preamps, they have been designed to give "consistently good sound over the whole range of gain settings". i think that if you tried these then the phenomenon that you have described would not be an issue.

i would hazard a guess that your mic preamplifiers (like your mics) are inexpensive and thus, most likely integrated circuit based instrumentation amplifiers. if this is the case then they have localised feedback, although i wouldn't assume that it is the feedback causing the sonic differences. it is most likely due to the altering performance characteristics of the preamp over different gain settings.

The op amps that are used in instrumentation configurations typically have an open loop gain of 10^6 and feedback is used to control the open loop gain.

hope this is (correct and...) helpful

[danlavry]

[quote title=ML wrote on Mon, 27 June 2005 14:06The op amps that are used in instrumentation configurations typically have an open loop gain of 10^6 and feedback is used to control the open loop gain.

hope this is (correct and...) helpful[/quote]


It is the other way around. The open loop gain is used to "control" the closed loop gain. "Control" means to help the output track the input with less distortions.

Regards
Dan Lavry
www.lavryengineering.com