Problem with too much Preamp gain.

jazzholiday · « **on:** 7 Sep 2005, 01:48 am »

I have an Audio Research LS2B preamp, and an Adcom 5802 power amp. My problem is too much volume gain with this combination.

If the volume is set at the 8 o`clock position, sometimes this can be too loud. I`ve heard of in-line attenuators to correct this problem, and I was wondering if they cause more harm than good. If not, then what kind would someone recommend?

BTW, I using the balanced ouputs and inputs on both my preamp and power amp. If not in-line attenuators, is there any other way to correct this problem as well?

Thanks, and any advice will be greatly appreciated!

Charles Hartmann.

warnerwh · « **Reply #1 on:** 7 Sep 2005, 02:23 am »

If you can use the single ended connections this will cut your gain in half. That will help alot.

Don Maico · « **Reply #2 on:** 8 Sep 2005, 08:44 am »

Hi Iam new to this forum. Have been experincing the same problem with my Croft modified Leak stereo 20 which is too sensitive for my Audio Innovations 800c preamp.Someone suggested buying a nogain Croft Vitale and some esle a passive preamp. Not sure whethr there is any difference???

Ric Schultz · « **Reply #3 on:** 8 Sep 2005, 07:45 pm »

The most sonically transparent way to attenuate gain is to add a resistor dividing network right before the pots in the preamp section (or on the input of the amp). Two resistors per channel are all that is needed. You could also solder them directly onto the input jacks but then the gain would be attenuated for that source only. If you wanted to attenuate 10 times (20db) you could use a 20K resistor in series with the hot and then a 2K resistor to ground per channel. You could use Nude Vishays, Caddocks, Tantalums, or whatever resistor you favor. Very inexpensive and the purest way to attenuate. Obviously, you would have to know how to solder and have a small understanding of the insides of things. Any tech or tweaker could do it for very little fee.

Uptown Audio · « **Reply #4 on:** 16 Sep 2005, 11:09 pm »

Use the RCA outputs and you should be able to get the volume control to about 12 o'clock, which is where you want to be with it. You may also find the sound to be a little sweeter than before as well. if that is not enough, there are commercially available line attenuators that you can place in-line between the preamp and power amp, ideally at the input of the amplifier.
-Bill

Bill Baker · « **Reply #5 on:** 16 Sep 2005, 11:37 pm »

HI Jazzholiday,
Actually, the best way, IMHO, is to add an attenuation circuit before the preamp circuit right at the volume control. The resistors used would depend on the impedance of the volume pot itself. Given what you said, I would think 10dB of attenuation would get you where you want to be. Do you know what pot is used?

SteveLim · « **Reply #6 on:** 17 Sep 2005, 05:07 am »

Hi,

I have the same problem with too much gain , preamp is the AudioResearch LS8 and amp is the Aragon 8008BB.

Sorry i am not so electrical savy, please instruct me exactly (before the volume pot in the preamp) where should i solder the Resistor Dividing Network (2x resistor per channel) ?

Also what value resistor is required ?
And what about the power-rating of the resistor needed ?

(Not sure if this is relevant infomation, but quote from the specifications : Input-Impedence is 50K Ohms, Volume Control is a 64-step micro-processor controlled device, Gain Main-Output - 12.2 dB unbalanced)

I would like attenation for all inputs.
I have tried a 6dB In-Line Attenuator but it was still insufficient (probably only increased the volume level from 8:30 o'clock to about 10:30 o'clock at usual listening level.

Acutally , am wondering if ARC preamps don't already cater some way to adjust the gain ? Anyone has experience getting that done to their preamps by ARC authorised dealers ?

Thanks a lot ,
SteveLim

Davey · « **Reply #7 on:** 17 Sep 2005, 02:33 pm »

Why does ARC design these line-only preamps with so much gain? 12.2db and then a user has to add a 10db attenuator. It doesn't make much sense to me.

Davey.

art · « **Reply #8 on:** 17 Sep 2005, 02:49 pm »

Sounds like a typical PGA2310 preamp........way too much gain.

I would advise leaving it balanced, but going for 20 dB attentuation. That would be a 10:1 ratio. Problem is, putting 500K in front of a 50K pot would make it noisy. Better way would be to add somthing like a 10K:1K divider in front of the pot. Most gear can handle a 10K load.

Or.........better yet.........maybe, depending on your outlook in life and audio:

Get a Jensen 4:1 stepdown transformer. That PGA will sound better if you keep the RF out of it. Solve 2 problems at once. That can be built in an external box, so you won't have to open yor preamp. Stick it in the tape loop.

Yeah.......I know........easy for us technically savvy guys.......not so for the user.

Pat

Steve · « **Reply #9 on:** 17 Sep 2005, 11:40 pm »

The problem with having so much gain in the amp, and the extra gainstage, is the frequency dependent feedback thru the power supply. The feedback varies, phase wise, from at least midband to deep bass, and from 1 degree to at least 90 degrees, to 180 degrees (and even more), depending on the power supply design.

The feedback can easily be as much as a typical "feedback loop" in a typical negative feedback amp (from the output of an amplifier to the input).

Thus advertising a "no feedback" amplifier (or even preamp if more than one stage) is not correct.

What makes it much worse is there are multiple loops of feedback, each with different phases. The more stages with a combined power supply, the bigger the problem.

Which, I guess, leads me to prefering a high gain preamp and lower gain, And fewer stages amplifier.

In otherwards, if one can get rid of a stage in the amplifier, you will be better off sonically than adding an external attenuator. But the former may not be feasible in your particular amp. Tube amps are much simpler.

Rocket · « **Reply #10 on:** 17 Sep 2005, 11:55 pm »

Hi,

My problem is that there is not enough gain between my preamp and amp combination.

Regards

Rod

art · « **Reply #11 on:** 19 Sep 2005, 08:41 pm »

Since when is modulation on the supply rails considered feedback?????

Having too much gain in the preamp......around 12 dB...... is because it has too many stages? How can it have less than one, unless it is nothing more than a volume control and a selector switch?

Pat

Steve · « **Reply #12 on:** 20 Sep 2005, 04:14 am »

Quote from: art

Since when is modulation on the supply rails considered feedback?????"

Hi Pat,

Yes, this phenomenom is prevalent in almost every design, but almost no one discusses it. It can easily be demonstrated that a signal from one stage "travels" to another stage, thru the power supply itself. Integrateds are more prone as there are more stages, and more loops.

Pat

art · « **Reply #13 on:** 20 Sep 2005, 05:52 pm »

It is not feedback! It is bias modulation. You are the only person that I have run across in almost 40 years that invents new termnology for something that everyone else calls what it really is.

Negative feedback is used to improve performance. Positive feedback is used to generate oscillations. Bias modulation does neither, because it is not feedback.

Just because a lot of us design in fixes for this problem does not mean that we do not know about it. We don't talk about it because it is easy to reduce to damn near zero.

And we don't call it feedback.

Pat

Steve · « **Reply #14 on:** 20 Sep 2005, 07:24 pm »

Quote from: art

It is not feedback! It is bias modulation. You are the only person that I have run across in almost 40 years that invents new termnology for something that everyone else calls what it really is."

The only difference between this type of feedback and conventional output to input feedback is the "port of entry" into the first stage. (But power supply feedback can occur in all stages in a component.)

Conventional output to input feedback enters either thru the grid or more likely the cathode/source/emitter of the first stage while power supply feedback enters either thru the plate/collector/drain of the first stage. Never the less, it is feedback as signal/music travels from one stage to another in both cases.

"Negative feedback is used to improve performance. Positive feedback is used to generate oscillations. Bias modulation does neither, because it is not feedback."

Here is just one reference. The Radiotron Designers Handbook (written by at least 27 engineers at RCA) discusses this over 50 years ago. This insidious feedback varies in phase to the input signal from 0 degrees to 180, to 270 or more, depending on the power supply design.

This means it could be negative feedback at some X frequency and positive feedback at Y frequency as the phase changes. Positive feedback doesn't necessarily cause an oscillation unless, when analyzing the design and plotting the results on a nyquist diagram, some portion of the response passes thru or circles the point 1, 0. This means certain criteria needs to be inplace, such as feedback strenth etc, to overcome losses and sustain an oscillation.

For example, a "motorboat" oscillation problem is a case where positive feeback is traveling from a latter stage to a previous stage, almost always thru the power supply, causing an oscillation.

"Just because a lot of us design in fixes for this problem does not mean that we do not know about it. We don't talk about it because it is easy to reduce to damn near zero."

Good, I am glad if you have accomplished it. I would love to read any comments or discussions concerning this.

"And we don't call it feedback."

But then the general public doesn't know what is actually happening in the design, that it is acutally musical feedback.

PS. A famous RCA 50 watt amp circuit contains a direct feedback loop from the plates of the output tubes to the plates of the driver tubes. This is an extreme example as it bypasses the power supply entirely. RCA refers to it as feedback, not bias modulation".

PSS. I have been looking over dozens of designs, printed out, and only one seems to address this issue. Maybe you could provide a link to one or more that does? Much appreciated.

/quote]

art · « **Reply #15 on:** 24 Sep 2005, 02:25 am »

I am not going to mention all of the amp designers that I know. You will just have to trust me. You know all of their names.

BTW.........all SS designers, if it makes any difference. We have tricks that aren't easily done in tube gear.

And yes, they have a problem with your terminology.

But:

Quote

For example, a "motorboat" oscillation problem is a case where positive feeback is traveling from a latter stage to a previous stage, almost always thru the power supply, causing an oscillation.

You have proven my point. Positive feedback causes oscillation. A signal modulates the rail such that the phase of the modulation is postive, and it pumps.

So, when the phase is not positve, it does not oscillate. It only makes things worse.

Anyway........call it what you want. Poor designs are subject to supply modualtion.

Pat

Steve · « **Reply #16 on:** 24 Sep 2005, 05:37 am »

Dear Pat,

Didn't mean to upset you, but it is very easy to prove my position.

Quote from: art

I am not going to mention all of the amp designers that I know. You will just have to trust me. You know all of their names.">

No problem, and I am sure alot of manufacturers don't have their schematics available. But what I have seen, mainly public schematics of tube designs, definitely shows feedback problems.

<"BTW.........all SS designers, if it makes any difference. We have tricks that aren't easily done in tube gear.">

I am sure Pat. I know some of them.
I never said it was impossible; just that most schematics that are available that I have seen don't address this issue.

<"And yes, they have a problem with your terminology.">

I have two texts, my 1st/2nd year College text and a major Engineering book that discusses these subjects.

So who are "they", who haven't read, or understand a 1st/2nd year text??

As I said, the only difference in intentional feedback and power supply feedback is the "port of entry". Very easy to prove.

<"You have proven my point. Positive feedback causes oscillation. A signal modulates the rail such that the phase of the modulation is postive, and it pumps.">

I don't know how you came up with that comment, but the example of motorboating was an Extreme example. Positive feedback easily occurs without oscillation, even in amps. Here are some examples from the two books I mentioned above.

First, from Radiotron Designers Handbook, page 306, in the Audio Section:
"Feedback may be applied to any amplifier, at any frequency, and may be either positive or negative."

You see, components may have voltage, current, bridge feedback, and both positive and negative feedback loops as well (see page 314).

Semiconductor and Tube Electronics, by James G. Brazee, page 380
"Toward the end of this chapter, a feedback system employing both negative and positive feedback will be discussed as an example of the possible use of positive feedback."

In a nyquist diagram (been a long time since I worked with this), the left hand quadrants indicate negative feedback. However, when the phase shift results in the plot extending into the right hand quadrants, in a circle of 1 radius, with 1,j0 being the center, we are talking positive feedback.

However Pat, there is NO oscillation unless the plot passes directly thru the coordinates 1,j0, or loops around that cooridinate.

These are pretty basic, solid concepts, in a 1st/2nd year college textbook.

One has to be careful whom one listens to.

Quote

art · « **Reply #17 on:** 24 Sep 2005, 04:38 pm »

Yes, I am sure that it has been a long time since that right half of the plane class. It was boring.

Ok....not as boring as the class on motors. Maybe you were lucky and didn't have to take one.

Pat

peakrchau · « **Reply #18 on:** 25 Sep 2005, 02:25 am »

Quote from: Davey

Why does ARC design these line-only preamps with so much gain? 12.2db and then a user has to add a 10db attenuator. It doesn't make much sense to me.

Davey.

Part of the problem is that some of these are defined for 250mV line outputs since it is the lowest common denominator (tape decks, fm tuners, tape decks) but most CD players put out a 1V (and sometimes 2V RMS) output.

Read through this post I put up earlier this year for more detail

PeAK

Steve · « **Reply #19 on:** 25 Sep 2005, 08:03 pm »

Nice point Peak..

There are wide differences in gain of amps and preamps. Some amps have sensitivities from .25 to 5 volts, and preamplifiers have gains from slightly less than unity to 26 db or more.

Then there is also the issue of matching the sonics one likes to the sensitivity one desires. Some like integrateds and preamps with little or no gain, while others like separates with external preamp gainstages.

Mismatch these and one can get either too much or too little overall gain.