Electronic Engineering and hi-fi set-up

[Blackmorec]

5 hours ago, Kal Rubinson said:

 

Removing the AC/PS reduces/removes the noise.  The battery neither adds nor removes anything (ideally).

Which is precisely but obviously not cleary what I said in the fist place. 

[marce]

11 hours ago, Ralf11 said:

this is a quantitative issue: does the battery have more noise than an AC outlet power supply?  which is easier to filter out, etc.?

 

here is an oldie but moldy for those wanting to find out...

https://tf.nist.gov/general/pdf/1133.pdf

Digital requires quite sharp current requirements when switching, it is this that can cause issues, I;ll try and find the relevant thread where it was discussed in detail and the links. I am not saying batteries are worse than any other main power solution (or better) it is all in the implementation.

 

9 hours ago, Albrecht said:

""Of course without electronic engineering and the advances made over the years we wouldn't have the kit we do today."

True that. And it is high performance audio electrical engineering that have really advanced, - guess what, - high performance equipment.

Where-as electrical engineering outside high performance audio, in different disciplines, - have actually hindered development. This is especially the case in multi-function computer design, - where certain engineers confuse design elements that make performance of many functions very good, and one function, (audio), - perform less well.

One example of this is Esoteric's VRDS NEO transport. Where-in many designers thought previously that software/firmware error correction didn't really do as good a job as super-accurate, beefy motor & magnesium disc clamping mechanism.

""Finally, having separate power supply's with umbilical's, many add on boxes with extra cabling (clocks!!!!) is not the best way for a noise free system, loops everywhere, without some serious measurements, you are only guessing whether its reducing noise""

Not really. High performance audio engineers measure and listen for that noise. The potential device purchaser doesn't need and shouldn't have to measure, - the proof will be in the comparative listening, - this will determine to them the effectiveness.

 

"far more critical is the on board supply components such as decoupling capacitors,"

No, far more critical is the HF noise leakage, and the out of phase AC current, sent back into the mains supply that also affect other sensitive digital gear in the chain. John Swenson has written about this, somewhat extensively, and provides great explanations about what is actually occurring with these low voltage digital circuits.

 

You don't work in electronic design. do you?

[esldude]

If you've ever had a good stereo in your car, one with good bass response to 50 hz or a bit lower (good response, not under-dampened booming thud) you might have experienced the following. 

 

Pull into your driveway, listening to music for a bit with car idling.  Shut off the car, but leave music playing.  And it seems like someone turned up the bass maybe 2 db, and cleaned (tightened) it up some.  I could talk about the battery vs the alternator supplied juice.  What happened was the fairly high level though subjectively just noticed low frequency rumble of the engine running was noise in way of the bass and its details.  It was partly obscured.  Turning the engine off cleared it up like a fog lifting. 

[Miska]

11 hours ago, Arpiben said:

 

Completing your reference some measurements. Not all batteries are equal vs noise...

http://www.hoffmann-hochfrequenz.de/downloads/NoiseMeasurementsOnChemicalBatteries.pdf

 

...and once you use battery, the output voltage range and stability is not likely what you need... So in most cases you have DC-DC converter connected to the battery to produce useful voltages. Which is essentially an SMPS...

 

[Miska]

7 hours ago, Kal Rubinson said:

That doesn't make any sense to me.  If the source of the noise is inside the device, neither should affect it.  If it exists with the PS, it or the AC line is the culprit.  The battery (assuming it is adequate to the load) should not introduce any.

 

Use of battery just requires different kind of surrounding electronics which is another source of problems.

 

For example many battery power banks used to provide USB charging voltage are noisy as hell thanks to cheap ass DC-DC converter / switching regulator needed for the task. (using a linear regulator with a power bank would be immensely stupid because it would waste lot of precious electricity into heat)

 

[marce]

10 hours ago, Albrecht said:

Where-as electrical engineering outside high performance audio, in different disciplines, - have actually hindered development.

Nope, blinkered view. Technology gains and lessons learned apply to audio reproduction.

 

10 hours ago, Albrecht said:

"far more critical is the on board supply components such as decoupling capacitors,"

No, far more critical is the HF noise leakage, and the out of phase AC current, sent back into the mains supply that also affect other sensitive digital gear in the chain. John Swenson has written about this, somewhat extensively, and provides great explanations about what is actually occurring with these low voltage digital circuits.

Nope, wrong. For the quietest operation of digital circuitry the Power Delivery System is paramount. I would suggest learning something regarding digital circuitry. I would also think that Mr Swenson would agree with me on how critical the power delivery system is for minimising switching noise, ground bounce and maximum signal integrity. I would suggest before you state something as being more critical to digital,

 

 do more research.

Some basics:

http://www.ewh.ieee.org/r5/denver/rockymountainemc/archive/2007/Dec_2007/Transient_Impedance_Slides.pdf

https://www.cadence.com/content/dam/cadence-www/global/en_US/documents/tools/ic-package-design-analysis/sigrity-resources/sigrity-facts-input-impedance-power-ground-planes-cp.pdf

http://www.ewh.ieee.org/r4/se_michigan/emcs/DL-ARCH-decoupling3.pdf

 

10 hours ago, Albrecht said:

"Finally, having separate power supply's with umbilical's, many add on boxes with extra cabling (clocks!!!!) is not the best way for a noise free system, loops everywhere, without some serious measurements, you are only guessing whether its reducing noise""

Not really. High performance audio engineers measure and listen for that noise. The potential device purchaser doesn't need and shouldn't have to measure, - the proof will be in the comparative listening, - this will determine to them the effectiveness.

How can they test for every different cable used, what the supply is hooked up to, what wires are lying next to each other, wire length, construction etc. Again there is a wealth of information, EMC covers a lot of it. You can only really solve the problem of noise by measurements...

Clocks down wires, seen some terrible implementations, you don't need a clock down a wire for domestic audio, you need a clock next to the device being clock for the ultimate in clock fidelity, with the clock supply next to or underneath the clock... Why do you think master clocks are slow (10MHz).

[marce]

What do you disagree with SandyK, please enlighten us, instead of just irritating us. Because for the life in me I cannot see what I have put that upsets your sensibilities... Can we have a more technical based discussion without you throwing your dummy out of the pram?

[Kal Rubinson]

5 hours ago, Miska said:

Use of battery just requires different kind of surrounding electronics which is another source of problems.

Certainly.  My response was to the situation proposed here where the battery supply was stated to be lower in noise than the AC PS.

[Speedskater]

19 hours ago, sandyk said:

I believe you, but as a technical person can you offer a plausible explanation for this ?

 Are you talking about FM stereo ?

Yes, FM stereo, WCLV Cleveland.

Why could we hear it on a clock radio and not on a good system.  But the clock radio had a frequency peak at the problem frequency.

[Ralf11]

it might be fruitful to give DIYers some guidelines on what battery type & capacity, etc. would allow them to power a DAC without any conversion or 'surrounding' electronics...

[mevdinc]

20 hours ago, Kal Rubinson said:

 

Removing the AC/PS reduces/removes the noise.  The battery neither adds nor removes anything (ideally).

This is exactly I meant.
The PS is designed to reduce the AC noise whereas the battery avoids it although it may add some noise of its own.
 

[sandyk]

3 hours ago, Ralf11 said:

it might be fruitful to give DIYers some guidelines on what battery type & capacity, etc. would allow them to power a DAC without any conversion or 'surrounding' electronics...

 Unless the battery A.H. capacity is well above that required for typical listening session lengths , and automatically recharged after a listening session this will not normally be possible, as otherwise performance is likely to suffer as the battery drops in voltage due to use, and is no longer able to follow peak demands on current without substantial variations in it's voltage.

 Given that many consumer batteries such as Li Ion also have internal electronics for overload shutdown etc. you really need to filter their output. Shunt Regulators such as the John Linsley Hood PSU add-on design work well in this area, as their voltage In can equal the voltage Out. (within a few mV due to PCB track resistance etc.) and their modest current requirements.

The JLH also has a capacitance multiplier area and an active cancellation circuit where ripple etc.to >300kHz  is amplified and applied in opposition to the original ripple, but it can't compensate for slow drops in the battery's terminal voltage due to use..

[fas42]

 

32 minutes ago, sandyk said:

 Unless the battery A.H. capacity is well above that required for typical listening session lengths , and automatically recharged after a listening session this will not normally be possible, as otherwise performance is likely to suffer as the battery drops in voltage due to use, and is no longer able to follow peak demands on current without substantial variations in it's voltage.

 

Yes, this is typically what N. up the road has found - the 'annoyances' with battery are that the performance slowly degrades with time, as the charge is consumed; and that if current spikes are required then this 'compresses' the SQ - with the latter, there was a very clear volume level, where the rapid onset of losses at higher levels made it pointless to listen there.

 

So, this is one area where 'grotesque' over-spec'ing, of the battery supply, is likely to be of real benefit - keeps the circuitry in the sweet zone over a much greater range of usage.

[fas42]

Dup.

[Miska]

4 hours ago, Ralf11 said:

it might be fruitful to give DIYers some guidelines on what battery type & capacity, etc. would allow them to power a DAC without any conversion or 'surrounding' electronics...

 

Problems are that battery cells have some inherent per-cell voltage, so battery pack voltage is some multiple of that. In addition, battery voltage is not stable but also depends on charge level and load. In addition, it is good to have some protections for things like short circuits and such so that there's less risk of fire and explosions. Car batteries don't have such, as they need to supply few hundred amps of current for short periods...

 

In addition batteries have various other properties, for example car/starter batteries don't like deep discharge and get bad very quickly if you discharge them more than about 50%. There are then auxiliary power lead-acid batteries (nominally 12V and 24V) that are designed for boats and motor homes to supply power for lights and such, these are built to persist deep discharges, but cannot supply as much current. If your battery voltage reaches 12V, it is pretty empty already and at about 11.5V it is out.

 

Li-Ion cells are more sensitive beasts than lead-acid batteries, so you need to be more careful how you deal with those...

 

But overall, batteries are one of the subjects where you could write several books worth of technical story..

 

[Ralf11]

to run scientific instruments, I've done what Sandy described above - high capacity, etc.

 

when not worried about noise I've used a higher voltage than needed and regulated it down for use

 

the former should work well to drive a DAC, etc. as not all equipment requires super-ultra-mega-low noise supplies...

[sandyk]

18 minutes ago, Ralf11 said:

to run scientific instruments, I've done what Sandy described above - high capacity, etc.

 

when not worried about noise I've used a higher voltage than needed and regulated it down for use

 

the former should work well to drive a DAC, etc. as not all equipment requires super-ultra-mega-low noise supplies...

 

That does NOT mean that many devices won't perform quite noticeably better when powered by a very low noise and low impedance Power Supply. Even replacing the supplied A.C. wall wart for my Musical Fidelity X-DAC V3 with an external Linear LOW noise + and -15.5V supply resulted in  MARKEDLY improved performance.

 The 330 ohm PW5 resistor seen in the photo was fitted to more closely match the loads on both the + and  -supply rails, as the -15V rail is only used to power the Analogue area opamps.

 This also helps to minimise " turn off" thumps.

[Summit]

On 11/6/2019 at 1:33 PM, Blackmorec said:

This morning I was looking out of the window as my wife commented on how hard it was raining. I looked but could not ‘see’ any rain and there were no puddles so I didn’t see any rings. Thinking about it I realised that I was looking out through very fine mesh mosquito netting, left over from summer, so I released the netting and let it roll up. Looking out of the window again I could now clearly see it raining.  While the mosquito netting is almost invisible and has very little if any affect on the lake view so wasn’t noticed, it completely obscured the fine detail of falling rain and its exactly these fine details that we use in hi-fi to provide sonic clues about soundstage, timbre, ambience etc.  So what’s the equivalent of the mosquito netting in hi-fi? Essentially it would be anything that obscures the system’s ability to resolve detail.  Remove them and hey-presto, it sounds like you’ve added something!  The electronics engineers who build our hi-fis are able to design, manufacture and measure equipment to perform at the very pinnacle of the art, but is that what we’re actually going to hear?  Not if there’s a step missing.....the set-up that ensures that there’s no sonic mosquito nets between you and your  music. 

 

 

I think you with the mosquito net have made a very good and “clear” analogy, bravo.

[Summit]

On 11/7/2019 at 1:45 AM, Kal Rubinson said:

 

Removing the AC/PS reduces/removes the noise.  The battery neither adds nor removes anything (ideally).

 

The battery can be disconnected from the mains and by being disconnected it will be isolated from the mains noise, ground loops and less prone to transmit leaking currants.

 

Battery has their cons as well, naturally, and I would not say it’s better than a well-made LPS or anything.  

[sandyk]

10 hours ago, Summit said:

Battery has their cons as well, naturally, and I would not say it’s better than a well-made LPS or anything

 

They can be worthwhile though with USB in particular, which doesn't like any capacitive coupling back to A/C mains earth. Even connecting the transformer shield lead to mains Earth with an  R-Core transformer resulted in a small but noticeable degradation. I was able to verify this after fitting a toggle switch at the rear of the PSU to try with, and without this connection.

[sandyk]

For those interested in a comparison of the output noise of several different types of voltage regulation, and especially the noise from batteries , this earlier series from TNT Audio may be of interest.

 However, it does not cover more recent lower noise battery types such as the LiFeP04.

https://www.tnt-audio.com/clinica/regulators_noise4_e.html

 

The article starts at https://www.tnt-audio.com/clinica/regulators_noise2_e.html

 

 

 

[fas42]

1 hour ago, sandyk said:

 

They can be worthwhile though with USB in particular, which doesn't like any capacitive coupling back to A/C mains earth. Even connecting the transformer shield lead to mains Earth with an  R-Core transformer resulted in a small but noticeable degradation. I was able to verify this after fitting a toggle switch at the rear of the PSU to try with, and without this connection.

 

I went through one "nightmare" phase of playing  with all sorts of strategies with how leads, etc, were shielded, and then connected to grounding points. This was an exercise in madness, because every change altered the sound, 😜.

 

What did this 'prove'? Well, that ordinary ways of hooking up audio gear are far too poor, in regards to being robust against shielding and grounding weaknesses - this sort of thing should have been sorted out decades ago ...

[marce]

https://www.bennettprescott.com/downloads/grounding_tutorial.pdf

https://www.rane.com/note151.html

http://www.hottconsultants.com/pdf_files/aes-2007.pdf

 

[sandyk]

6 hours ago, marce said:

https://www.bennettprescott.com/downloads/grounding_tutorial.pdf

https://www.rane.com/note151.html

http://www.hottconsultants.com/pdf_files/aes-2007.pdf

 

 A typical type response from you and several other members.

 You just can't help yourself, can you ?

( Screen grab from the 1st "Resource"  linked to by you.)

 

 Consumers shouldn't need to know these things, they should have been taken care of in the original equipment design stages if the E.E.s were competent.

[Ralf11]

It is a typical response because it seeks to counter the common mode noise that is presented at such a high frequency here and elsewhere.