@ danielwritesbac
There is a commonly overlooked setting within a lot of BIOSes. In mine it's called "detect pci" in others it's called something else. It basically checks all of the pci slots to see if any hardware is active on the slot and turns off the ones not it use. If this is not enabled you can get a ton of EMI from the PCI slots being turned on when not in use. This is different computer to computer. But I have a feeling this could give you terrible audio quality with a ton of cards and I never see people really talking about it or noticing it.
There is a commonly overlooked setting within a lot of BIOSes. In mine it's called "detect pci" in others it's called something else. It basically checks all of the pci slots to see if any hardware is active on the slot and turns off the ones not it use. If this is not enabled you can get a ton of EMI from the PCI slots being turned on when not in use. This is different computer to computer. But I have a feeling this could give you terrible audio quality with a ton of cards and I never see people really talking about it or noticing it.
Just to finish nicely, here you can se the relay, which is mandatory by law to have in every household in DK. It only allows µA´s of current going to earth, before cutting the power of.
It is mandatory to test it every 12 months by pushing the testbutton.
Googles billedresultat for http://www.elpaco.dk/uploadedImages/4976-hfi-rel%C3%A6.jpg
Oh yes, last year we had a big go-round with power topics and the discussion about the chassis of electronics which must be grounded for safety, versus the risk to U.S. residents of grounding humans which puts half the circuit into place just awaiting contact with the other half to close the circuit. Of course the grounding is the safer option, even if you are in the U.S.
However, in order to match up with all of the international safety issues all at once, we had to either use a wooden amplifier enclosure or install a GFCI.
The GFCI idea was objectionable to some, so we chose the [flammable] wooden amplifier enclosure. Upon finally insisting on a wooden enclosure, the project was finally unstuck and could then progress.
As I remember, the big problem was not after the amplifier was in service, but rather because it was a diy project for publication, and it was expected that a person's hands may be inside the enclosure in the presence of potentially charged up capacitors. In this case, the metal enclosure made it near-impossible to avoid a shock.
Because of this, replacing the largest conductor (metal enclosure) with an insulator (wooden enclosure) was the readily available solution.
Sure, there were some technical solutions available for metal enclosures, like spring loaded switches to automatically disconnect and drain the power supply; however, reliability and compatibility with various enclosures was uncertain.
Perhaps, in the future, we could arrange for a non-flammable solution.
Why was the safer (the fires caused kill a lot more than electrocution) option objectional? Is the wood enclosure surrounding a metal RFI shield? (the other reason for metal grounded chasis)
By the way, lots of detailed grounding info from Bill Whitlock of Jensen out on the net. (and if he ever lectures near you go see him, I dont care how much you think you know he will teach you something)
Well, I suppose so - as long as you are aware that's what is happening. In the case of the device Gary and I measured, it was meant to have more than one thing plugged into it. Could be rather rough on a quiet component (E.G. phono preamp) if plugged in with a noisy one. The Tyco filter posted above is meant for one device only.
What would be interesting to measure - and I'm sorry I didn't get the chance - does having a high impedance return path for the line noise cause the noise to rise in the DUT? Sort of like the old argument of whether is is better to isolate a turntable or allow its vibrations to drain away? It may well depend on the device and what else is in the system with it.
A fertile field of research, I would think. =)
It was LM1875 chip amplifier for Pioneer B20 projects. The LM1875 is small enough for preamp duties and more than big enough to rattle the whizzer on that B20, so one could say that it is more than sufficiently powerful for the task. It can make at least 20 different amplifiers. It can also make class A headphone amp, bridge pre-drive units, Class G/H amplifiers, power supply components, etc. . .
We chose LM1875 because although the distortion is low, it does favor the ear and it doesn't increase with pitch like the larger models do.
Also, wiring is quite simple.
With even the 100k value based version, at a gain of 45 for use with a battery powered source, it still has no need of rf protection. We added some anyway. Those with a 12k load (gainclone) or 10k load (for bass lovers), didn't have any need of rf protection.
The pin compatible ST version (and the fake LM1875 likewise), which can use a bit less expensive transformer, does need the RF protection.
Sufficient RF protection was provided by a metal film resistor from inverting input to output, a 1/2w carbon resistor from inverting input to small signal ground (exploiting the hf/rf failure mode of a large carbon film resistor), an RC (zobel) at the speaker jack (aimed at a much higher pitch than spec--a hi-fi adaptation) and a small value capacitor (smaller than normal because its for hi-fi) as a load at the enclosure's input jack (to "short" rf).
Varying the value of that 1/2w resistor could produce all sorts of tonalities as well as varying the gain.
Varying the value of the output RC (amp zobel) could compensate for a wide variety of cables. The cheapest possible polyester capacitor allows you the freedom to choose between either more resistance or more capacitance without undue load on the amplifier.
A frequent (and terrible) mistake is using a spendy capacitor for output RC (amp zobel); so, it took some convincing to get some folks to buy a pair of cheap caps.
lol!
The case here is either use the polyester dip caps or else spendy caps are going to mash the treble harmonics and you just know that someone will omit the output RC because of spending too much on this component.
Rf blocking is:
At entrance
During amplification
At output
At the power supply too
No need for a shield.
At the time, this was tested with a 9 meter transmitter run at point blank range to the amp. Despite the amp's filtering allowing a larger-than-spec bandwidth (typical of hifi), there was no noise.
Inductance will not reach audible values in a configuration like I mentioned anyhow. The 3D-improvement can`t be overheard on a top system.
Never heard of any problems with unstabile amps during the decades I`ve been doing this.
And yeah, this skin-effect theory is one stupid urban legend. If only SS-people had any clue about how much of their amps quality that get lost in tiny stranded threads
Btw; litz-thread is another mis-understanding. They show a flat frequense, but what does that help as long as the signal is distorted and dynamics sux
Palerider, you did not answer me on this.
If so, do you only use it on LF? Where do you buy them?
How do you prevent the speaker from falling over?
If so, do you only use it on LF? Where do you buy them?
How do you prevent the speaker from falling over?
Inductance will not reach audible values in a configuration like I mentioned anyhow. The 3D-improvement can`t be overheard on a top system.
Never heard of any problems with unstabile amps during the decades I`ve been doing this.
And yeah, this skin-effect theory is one stupid urban legend. If only SS-people had any clue about how much of their amps quality that get lost in tiny stranded threads
Btw; litz-thread is another mis-understanding. They show a flat frequense, but what does that help as long as the signal is distorted and dynamics sux
By trying to calculate with this Java calculation of Twin Feed Loudspeaker Cable.
and setteing diameter to 10mm, spacing 50 cm and length at 2 m, I came out with these values.
@200Hz
R = 0 Ohm (very low)
C = 63pF (pretty low)
L = 5198 (very high)
Gain = 0 mB
Delay 649 nSek.
@20KHz
R = 3 mOhm (as a result of skineffect and induction)
C = 63 pF (pretty low)
L = 4749 (very high)
Gain = -2mB
Delay = 592 nSek
As can be seen there is 5µH in the cables, which is very high. This will surely influence your amplifiers behaviour, and that is the main reason for trying to avoid induction.
Tss, Tss....good ol´ Schopenhauer at work?
As we are talking about perception and not deafness, yes of course the experimentator can cause listeners nondetection of differences, if he is not aware of .......
Please remember at that point my steady comment on mandatory positive controls (and negative ones as well) .
But beside that, if the inclusion of a controlled listening tests does only depend on what the experimentator think is possible/plausible/extraordinary, things will very likely draft in the wrong direction.
Regarding the unbiased test- something like that can not exist on earth; that´s why i posted the question about the bias mechanism that remain unaddressed to remember that fact.
I don´t understand that, taking your answers into account, i used an argument about a well known fact (the experimentator bias, if you have any doubt go through the literature ones more) and the conclusion is based on formal logic.
If you want me to suggest something else please be more specific, i´d be more than happy to help if i can.
Wishes
Errors that oppose the ear are easily detectable at 0.12db, whereas errors that favor the ear can be tolerated at a far higher sound pressure level. By mentioning this, I'm just hoping to illustrate that using measuring equipment to support a "this proves you can't hear it" position, may have omitted the purpose of audio reproduction equipment. Its become a popular practice.
In addition I´d say, that one should try out messing arround with single wire cables, when the music plays. Both separating them and teiing them together, the effect is easily audible. And of course so is the DC resistanc of the wire itself, especially when it gets high.
I'm looking forward to my first night ballgame with LED lighting.
Here is a link with some interesting info on the plasma lighting.
I can't see that 5 mico Henry could hurt an amp much. 5 milli, well - yes. Is that what you meant?
CODE]@20KHz
R = 3 mOhm (as a result of skineffect and induction)
C = 63 pF (pretty low)
L = 4749 (very high)
Gain = -2mB
Delay = 592 nSek
I can't see the loss of 0.02db @ 20khz as significant either. Even if resistance is down to 4 ohm ~0.04db loss. Personally I have at least -30db loss in my ears at this frequency.
Peter
I can't see the loss of 0.02db @ 20khz as significant either. Even if resistance is down to 4 ohm ~0.04db loss. Personally I have at least -30db loss in my ears at this frequency.
Peter
Has anyone examined its effect against the inductance typical of the high frequency behaviour of the vast majority of consumer loudspeakers? Or the capacitance of a stat? The truly non-reactive, 8 ohm @ 20 kHz speaker must be a rare thing.
I was never able to hear a difference in sound quality between bare, #8 gauge, insulated solid copper coil winding wire and bare 140 strands of #40 gauge insulated copper Litz winding wire.
It was only as you began to add dielectric material to the two already coated wire types that any sonic changes became obvious. The 8 gauge was basically inert to all of the tubing I had, but was affected by lying on a man made fiber carpet or hanging suspended about a foot from the carpet. The Litz (140 strands of #40) was affected by any tubing I placed it into, but was inert to the man made carpet in any position.
Bud
It was only as you began to add dielectric material to the two already coated wire types that any sonic changes became obvious. The 8 gauge was basically inert to all of the tubing I had, but was affected by lying on a man made fiber carpet or hanging suspended about a foot from the carpet. The Litz (140 strands of #40) was affected by any tubing I placed it into, but was inert to the man made carpet in any position.
Bud
- Status
- Not open for further replies.