Lifeless Sound at Modest Volume

[tangmonster]

so an amp plays fine when you drive it into clipping/ high volume

But sound crap at normal or low listening volumes.

and there is 1500uF capacitance on a voltage rail that can easily reach peaks of 4 amps.

I would change / add caps to the power supply first before worrying about attenuation or preamps.

 

[abraxalito]

so an amp plays fine when you drive it into clipping/ high volume

But sound crap at normal or low listening volumes.

and there is 1500uF capacitance on a voltage rail that can easily reach peaks of 4 amps.

I would change / add caps to the power supply first before worrying about attenuation or preamps.

But there's a contradiction here - maybe you didn't spot it. The current peaks of 4 amps won't be present when playing at low volumes. So I'd expect the caps to be an issue when playing at higher levels, not lower. That's why its more likely to be noise sapping the dynamics than undersized caps.

 

[tangmonster]

We can go into this scientifically. Or trial and error.

Trial and error: Get two bigger caps and up the capacitance on the power supply and give it a listen.


Scientifically: first we need to pin down what "lifeless" means. does It mean not enough bass or what? Then you need oscilloscope and connect to output and a 8 ohm load and send through certain frequencies like 50 or 80 hz or higher frequencies. Then compare input vs output and low and high output volume.

also test for high frequency oscillation , and all sorts of other tests.

 

[tinitus]

I had a USB cable connected to my PC, and the other end hanging loose
For faster download of from my camera

I found that it was enough to spoil the fun (read music)
Once the USB cable was removed completely I got the silence back again, and the music sounded good again

Probably acted as a high frequency antenna

 

[abraxalito]

We can go into this scientifically.

Your version of 'scientifically' doesn't sound too scientific to me Let's apply science to your trial and error solution - what's the hypothesis for why bigger caps will make it sound more dynamic at lower volume? Do bigger caps lower the noise floor?

My hypothesis is that the noise floor is too high, probably because of RF grunge like tinitus's example just given.

 

[jstan]

Thanks for all your ideas -- this has been an interesting conversation so far. At this point I think I am going to try adding a preamp, and will probably add capacitance to the ps while I'm at it. I'm looking at Pedja Rogic's J-fet buffer and Joe Rasmussen's tube-base preamp. I love the idea of using tubes (probably comes from being a guitarist and my diy electronics background being with guitar electronics), but it would be a more expensive option. Any opinions about these two preamps (or any other designs you would recommend)? This is a great community -- thanks for all the help.

 

[zebra100]

jstan,

I have an Gainclone with a passive attenuator based around an Audiosector LM3875 Premium kit (18-0-18v, non regulated PSU). For my passive attenuator I used an 50K alps dual log pot. Initially, I had similar problems to you with the sound quality at low level. Also, the volume got very loud, very quickly with only a small turn of the pot.

I solved the issue using a shunt resistor (metal film .25w) of 47k directly in the input line, using the Alps pot as a bypass to ground (look up volume pot shunt mod on Google).

The difference was night and day. At low level, it was like the difference between having your fingers in your ears and then taking them out.

For about 14 pence (about 10 cents), and a few minutes with a soldering iron, it is well worth doing, and much, much cheaper than an active pre.

Chris.

 

[zebra100]

^ forgot to add:-

This mod also ensures silence at zero volume.

Link to the WD forum: Shunt Pot Volume Control - World-Designs-Forum

Chris.

 

[osscar]

I use a DIY LM 3875 (non inverting minimalistic design) With passive preamp (Discrete pot 50k) - it sounds good With 89 and 91db 8 omhs speakers in any level.
PSU - MUR dieodes, 2x100VA transformers + big RIFA PEH 15,000 uf per rail (no bypass caps on chip).

 

[jez]

A good system should not (IMHO) particularly sound good at low volumes....
The human ear is very non linear in many ways and it is only at "realistic" volumes that the frequency balance can be expected to sound correct

 

[zebra100]

My system sounds fine at low levels, and just as good at moderate and high levels. It is clean and detailed all through the volume range. I mainly listen to it in a near field monitor set up, so I generally don't have it too loud anyway.

My amp has 2x 18-0-18v, 120VA transformers & 2x MUR860 diode bridges (dual mono) giving just under 25v off load (unregulated) at the rails with 1500uF caps.

Chris.

 

[tinitus]

A good system should not (IMHO) particularly sound good at low volumes....

On the contrary
but never mind

I once talked to an old speaker builder
he claimed that any speaker should be adjusted(crossover) to the expected SPL they would be used at

half right half wrong
its in the nature of a low SPL setup to be superiour at exactly that
and a high SPL setup is ofcourse the winner at high SPL

Now, what you prefer is solely up to you
its not a matter of being right or wrong
just preference

 

[AndrewT]

My ears are the same age as my brain.
I expect a car door slammed shut outside my bedroom window to sound quite different from one slammed 100metres away.
I still recognise both as real car doors being slammed. They are both lifelike.
My ears/brain tell me that. I know from experience and processing power what sounds different and yet the same.

It does not need the same SPL for both to be car doors being slammed.

 

[tinitus]

And in audio its expected that you can tell the car make and model as well, and not just some car door being slammed

 

[kevinahcc20]

I'll second AndrewT on this...a well performing system should provide the illusion that the volume control magically moves one physically closer to or more distant from the source. The brain is quite adept at interpreting the nuances of the Fletcher-Munson curves in the real world.

 

[jan.didden]

On the other hand, most seasoned sales persons in audio know that when a customer compares the sound of systems, they tend to prefer the one that's played loudest. Something that is often used to 'help' the customer decide for a specific buy.

jd

 

[maxlorenz]

Sound lifeless at low volumes--> use a buffer (cheap) or a TVC (expensive).
I wonder who has used the PowerReg from Teddy Pardo on his chipamps, here???

Good luck.
M.

 

[jstan]

I've ordered the parts to put together a tube-based buffer (a la Joe Rasmussen), and have already disassembled the amp, so I can't try the shunt resistor just now. I'll let you all know the results when I've put this together.

 

[zebra100]

When you do get round to putting it back together, just give it a try with the shunt resistor in place before you add the tube buffer. You may be plesantly surprised. A buffer should give a bit of sparkle at lower volume levels, but it will be altering the original signal, albeit only slightly in a well designed buffer. At the end of the day, it's down to personal preference.

I get the best results keeping the signal path as clean as possible. With chip amps, you need to get the basics right, and you will reap the rewards. It's just a case of careful component selection, an adequate power supply, effective grounding and careful layout (shielding) that will obtain the best performance from the chip.

Chris.

 

[jez]

I stand by my earlier comments.