[mamsterla]

Hi everyone:

I just purchased one of those handy Radio Shack sound pressure meters and was trying to calibrate my system.  It appears that I may have a 0.5db inbalance between my Aksa 100wpc amps.  Since I am planning to go in and Nirvana'ize them (when Hugh reads my difficult order request in his inbox), I wondered how to correct the imbalance.  

I am guessing this is a biasing issue and that I need to get the bias of both channels exactly right, but if not I am curious to learn.

Thanks...

-MA

[Pleb Plebian]

Hiya Mamsterla;
Just checking - you have verified your results with the power amps swapped? - ie remove both speaker and line leads from one side, put them on the other and re-measure?

I had similar issues with my system - until I discovered a resonance in my room that was corrupting the results. Obviously, in my case, it was a speaker (room-interaction) issue rather than the electronics that needed attention.

I solved it by moving house

Just food for thought, anyhow
-D

[mamsterla]

Hiya Mamsterla;
Just checking - you have verified your results with the power amps swapped? - ie remove both speaker and line leads from one side, put them on the other and re-measure?

I had similar issues with my system - until I discovered a resonance in my room that was corrupting the results. Obviously, in my case, it was a speaker (room-interaction) issue rather than the electronics that needed attention.

I solved it by moving house

Just food for thought, anyhow
-D

I was thinking that - I will try swapping the leads.  I found that I have a rather nasty room resonance at 95Hz - that subwoofer sweep tone is very interesting to watch on the meter.  In any event my path of thought led down to wondering how I would adjust channel balance if that turned out to be the case.

[OT] If you ever want your wife to really think you are crazy, have her walk into the room as you are playing test tones and holding up a sound meter.  The expression on her face was priceless - something like "I guess I knew what I was getting when I married him, but does he have to show it off like that....geek."

[Pleb Plebian]

Yeah, I now have a dedicated listening/ HT (ahem.."boys") room.
I got the same look when I screwed a 4-1/2' tall pyramidal speaker to the roof as a center-channel. She Immediately took a photo of it and sent it off to  an NZ gadget-mag with the caption 'am I the most understanding woman in the world or what??' - and won a prize!
Mind you - she doesn't come downstairs all that often now

Re channel imbalance - if 'twere me, i'd do it passively: get a 20K pot (a dirt-cheap WW will do for testing purposes, although you might want to spend a couple of extra $$ and get a cheap plastic one) and hook it across the inputs on the higher-level amp - same circuit as used on the TLP/GK volume.
Adjust until you get power levels set perfectly - swapping channels to confirm as above - then remove the pot and measure. Duplicate the circuit in good quality 1/2W MF R's (you'll only need 2) or if you want to go totally ballistic, Bulk Foils or similar.

All of this presupposes, as you've stated, that the bias is set correctly on both amps...easy to do, if a touch tedious: run a test signal thru both channels for the same period of time, then adjust - and repeat

Cheers
-Darl

[EchiDna]

Hi everyone:

I just purchased one of those handy Radio Shack sound pressure meters and was trying to calibrate my system.  It appears that I may have a 0.5db inbalance between my Aksa 100wpc amps.  Since I am planning to go in and Nirvana'ize them (when Hugh reads my difficult order request in his inbox), I wondered how to correct the imbalance.  

I am guessing this is a biasing issue and that I need to get the bias of both channels exactly right, but if not I am curious to learn.

Thanks...

-MA

just my $0.02 worth... I actually work for a company that  does noise monitoring in industrial plants and factory floors and the little cheap Radio Shack SPL meter is NOT usable in any internationally recognised tests... you know why? because it is NOT accurate enough. 1/2 a dB is nothing to your ears (obviously - because it took the meter to show you that it existed). A decent SPL meter costs over US$4k and one that can monitor multiple positions at once to give a comparison starts at about US$6k. My point is this, if you can't hear it, don't worry about it. Chances are you are more deaf in one ear than the other anyway, or as previously mentioned, essentially all of us are without perfect room dynamics which can make far greater than just 1/2 a dB difference....

GL with it, and I hope you can get what you want out of the extra effort, but sorry, I doubt you will pick the difference

Cheers

[mamsterla]

So - the difference is probably due to room resonances.  It is an odd shaped room with different height ceilings and though both speakers are away from side reflections, there is one closer to a corner than another.  I switched channels and measured similar differences when the AV receiver was leveled the same left/right.  The room resonance was about 2db which I think I can hear - it shifts the image left/right and though I would be the last to claim I have golden ears, I can hear the difference from before and I like it better.

I agree that I expect little precision from this $35 tool, but it is helping with some general settings and I am amazed at how much better the AV sound is when the levels are uniform (my test tones measure to withing the same 1 db) at my prefered sitting position.  Pretty good IMHO - nice smooth pans now.

The meter also helped me quantify some nasty room resonances with my subwoofer.  I see a pretty big hump at about 68 Hz - I need to figure out which dimension this is - probably a floor/ceiling resonance is my guess.  It is about 6db or so and is audible in music.  It would be nice to get a notch filter for this, so I might pursue building something there or buying a parametric equalizer to fix bass frequencies.  

Thanks for the info - I will definitely trim the biases on the channels when I do the Nirvana upgrade...

[AKSA]

Michael,

Interesting assessment you have done on your room.

Generally, only the best ears can pick up 1dB;  most can pick 2dB, and all can pick 3dB.  I have never been able to pick lower than 1dB, so in this sense, your sound meter is doing well..........

Room asymmetry and resonances would certainly account for the differences you have heard.  No doubt about it.

However, the gain setting components on the AKSA power amps (and the GK-1) are accurate to about 1% and set the gain to 2% - the aggregrate of the two 1% resistors used.  It is the ratio of (82+2.2)K to 2K2 in the feedback network, and when you do the math, gain can vary from 31.48dB to 31.83dB - a difference of just 0.35dB - something you could never hear.

In truth, most good loudspeaker systems will move up and down at least 2dB over their full range, normally peaking at resonance, which is suspiciously close to the 68Hz you mention.

Hope this is helpful,

Cheers,

Hugh

[fred]

I also have a channel imbalance, that I am certain is due to room acoustics (wall on left, open on right).  It's very audible, and I am correcting this with the balance adjustment on my "preamp."  When I get a GK-1, this won't be possible, so what should I do then?  Adjust bias on the amp?  adjust within the GK-1?

[eico1]

mam, you must measure electrical properties of your amp to determine any channel imbalance, not acoustical. Use a test cd and an dvm.

steve

[PSP]

Hi Fred,
I have the same problem due to the same room geometry you have.  I've just been living with the imbalance, but I may just get my butt in gear someday and fix it... A good way to correct channel imbalance might be to:

a) get a reasonably good measurement of the imbalance (no point trying to get better than +/- 0.5 - 1 db)... using an accoustical measurement made while listening to music (?) since measurements made on the amp alone will not correct for speaker/room interaction and accoustical measurements at a single frequency might be influenced by standing wave phenomena.  Could we do this with an Ratshack meter, playing only one channel of a mono recording at a time?

b) use an L-pad or T-pad with good quality fixed resistors to pad down the loud side (probably at the volume pot).  An excellent resource for L-pad and T-pads appears in VoltSecond's attenuator page (written for the Foreplay preamp, but the same physics works in other amps too   !!).
http://www.siteswithstyle.com/VoltSecond/12_posistion_shunt/12_Position_Pure_Shunt.html

Scroll down to sections 6, 7, 8, and 9; in particular, study section 9, "Padding of the output or the input of a preamp both have benefits and deficiencies."... again, this is written for the Foreplay, but the same physics applies here.  There is an Excel spreadsheet on this page that you can download that will calculate the resistor sizes for you.  I've used this with great success to adjust the output level of the Foreplay, which in stock form is usually WAY too loud when used with SS amps.

Hugh will have to comment on the suitability of this approach to the GK-1.  I am currently running my AKSAs using a passive attenuator while I re-work my TLP... now that I think about it, this is a good time to fix the channel balance problem too!!

Good luck,
Peter

[AKSA]

Fred, Peter,

I'm carefully thinking on the balance issue.  I definitely have a solution - unique of course, but hey!  -  what else can you expect??

Cheers,

Hugh

[Larry]

I believe Hugh will come up with an excellent solution for sound levels at AKSA/GK-1 for this.

As Hugh pointed out, the pre/amps just have a very limited impact on sound levels while the room and positioning profoundly affects your perception of audio delivery.

The following chart shows the filter response in frequency domain with Behringer Ultra Curve Pro DSP8024, which brings a leveled frequency response at the listening position.  It's the compound response from GK-1/AKSA/DC-X(speaker)+REL Strata II(subwoofer which is closer to left) with my assymmetric room and speeker positions. Without GK-1, the soundstage shrinks, while with 8024, the timbre is improved. I listen music not in a dedicated room but in my usual chair. (Let the system serve me; not me serve the system.) In a/b test, no one would miss such difference that the 8024 brings up.

[fred]

Fred, Peter,

I'm carefully thinking on the balance issue.  I definitely have a solution - unique of course, but hey!  -  what else can you expect??

Cheers,

Hugh

Hugh- did you ever come up with something for this?  I expect to be ordering my GK-1 soon, and this channel balance issue will be a problem for me because of my room acoustics.

[AKSA]

Fred,

Thanks for the post.  I have been moving on this, and believe I have a solution now.

This issue of channel balance is tied up intimately with level control, and in fact is a differential level controller.  Recently I have made some strides in attenuator development, and discovered some very worthwhile sonic benefits, which I'll share later, of course, and it will be a zero cost retrofit.

In essence, using a linear pot, it's possible to implement a shunt balance control together with a shunt volume control.  This gives minimal sonic impact, and as a worthwhile advantage allows you to use an inexpensive potentiometer without sacrificing sonics, which is MORE than important.

This is not to say that there won't be any sonic impact, but I'm betting that it will be imperceptible because of the topology.  Actually, I hope to do this test at the beginning of the week, so I'll know pretty soon and all my hunches tell me this will be precisely as I expect.

When the GK-1R is released, not too far now, a balance control will be incorporated, and for the GK-1E and GK-1M it will be a lowish cost option.

Cheers,

Hugh

[fred]

Hugh - As you know, I'm well on my way to completing my GK-1R (analog board complete; other modules in progress).  So I'm nearing the point at which I'll need to deal with channel imbalance.  Have you settled on a solution yet?

I'm open to hardwiring for my own imbalance, rather than a more traditional approach with variable control.  Can I just slap a resistor (value determined through trial/error) in the signal path to attenuate one side?

[AKSA]

Hi Fred,

I've solved the problem, and configured it in shunt mode so that the quality of the balance control has virtually no impact on sonics.  This means that a crummy $2 carbon pot works brilliantly as a balance control.  I've amended all subsequent instructions to reflect this, of course, but have sent you a gif of the configuration.  I won't put it here because it's a novel solution I'd like to keep out of the public domain, but, Sire,

'Check your email'!!  

Cheers,

Hugh

[fred]

RECEIVED your e-mail.  Muchas thanks!  Dang, that looks pretty easy!