Groundside Electrons

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
Hi Hartono,

I am willing. I am sure there will be objections raised over my comments.

Starting with the preamp. It is DIY and is a SE design using a 6G6GT as driver, a PPFX tin and poly 1 uf interstage cap and a 6v6 cathode out to a SE OPT, with a split 1 to 0.85 overall step down secondary, to a balanced output. The OPT is 48% nickel core, will tolerate 40ma DC and also has headphone output windings. This preamp does have a solid copper ground plane, but all wiring is direct to adjacent components and is held aloft by turrets and terminal strips. No actual signal ground or power ground attaches directly to this coper sheet. Power and signal are in separate, adjacent boxes for each
channel, though that is about to change with the next rev.

Power amps are also DIY and have a balanced input direct to grids of two parallel push pull 7119's with no grid blocker resistors. The 7119's drive a PP step down interstage transformer with a 1.414 to 1 ratio into 8 EL 34's in PP parallel. Each of these has a 1 k grid blocker. the El34's drive a PP OPT with a 1.56k Z ohm primary to 4, 8 & 16 ohms. There is no copper ground plane and all components are tack solder connected or mounted to terminal strips. The power portion is in a 17 X 3 X 4 chassis with all of the tubes in another identical chassis adjacent to it and the input and output transformers are in a third, adjacent, identical box.

The SACD / CD player is a SONY model with all single sided PC boards with integral ground and signal traces. the Sony is model number DVP NS755 V. There are LP and tape gear but they do not need to be included just yet.

The preamp and power amp each have three Litz wire loops per channel. One on input ground buss wire, one on output ground buss wire and one on the cathode return of the driver stage. All are left loose and lay on the steel chassis or hang in open air. The Sony has four Litz wire loops connected to it's tiny and irregular ground buss, across the various output to PCB lug pins. These also lay on the chassis, between it and the trace side, back most edge of the PCB.

So, as you can see, any help in this ground plane VS electron loitering issue will make a fair difference. The difference is not a better sound, the components pretty much sound just as they did before the Litz loops, as far as "quality" is concerned. The difference comes in retention, as information that is intelligible.

This is an across the board increase and is primarily in secondary information content. The major signal transients are basically unchanged, the information that is increased is in reflected sounds of these primary signals, the interior chord structure and decay component of the primary signals and the overall hall noise of a recording venue, as it is described by reflections of signals provided by instruments and decay reverberations of both those instruments, as in piano sounding boards and hall descriptive information.

Some times this is not a "better" sounding reproduction than without the Litz in ground positions. When it is, what does come out are subtle things, like a Red Book CD that had a fair amount of information in higher frequencies that was not getting out of the EnABL'ed speaker system as intelligible information. Just a harsh crashing and turmoil at fairly low levels. Enough so that Mercury transcriptions to Red Book were just not worth listening to, due to the things being done to those rather brightly recorded instruments.

The addition of the Litz loops to ground transformed that "hash" into intelligible string resonances, hall echoes and eliminated the thin and sharp character of those steely violins. They are now lushly busy, sweet and open and quite the best recorded violins I have. This from their Respigi recordings, which I have on tape, LP, SACD and CD. Depending upon the recording, the Red Book offering is the "best" sounding one, a fair percentage of the time......

Bud
 
Hi BudP,

"The SACD / CD player is a SONY model with all single sided PC boards with integral ground and signal traces"

so the SACD player only have single copper pour plane, please elaborate more on integral ground and signal traces.

can you identify the PCB material ? and how is it connected to chassis ground ?



I'll try to analyze this configuration first, meanwhile try to provide above information.


Hartono
 
Jeff,

If it is magnet wire Litz then I can suggest the ground, through PCB pins, on the audio output of your digital media player. This is assuming that it is a nominal commercial model, with the cost accountant and marketing driven economies of no poured ground plane between components and PCB traces. If it has a poured ground then that should provide all that the signal side needs, with respect to whatever the Litz provides. If your wire is not made up from an insulated magnet wire then all bets are off, I have zero experience..

If you play electric guitar try a 2 inch piece, with one 0.6 inch long piece of shrunk down shrink wrap tube, between each pick up coil, as the ground buss. Also to the ground lug of the first pot. Better than new strings! You do not want more wire than this and you may want less plastic. I do not know what wire surface area to dielectric material ratios might be at work here so cut and try is the rule.

With all of the Litz grounds, if the sound character becomes more vivid and detailed and then that goes away, you have too much plastic. Cut it in half and try again. No, I don't know why this occurs... maybe too many electrons in the trap just weigh too much....snigger... and they can't get out. Truly, I have no idea, that's the reason I am asking these fool questions and accepting whatever egg comes my way, as my due diligence.

Hartono.

The reason for the ground wire loops arises from the speaker and interconnect cable experiments, which arose while I was trying to find the limits of resolution of the audio reproduction out put and interstage transformers I was experimenting with, that arose from the R&D on guitar amp outputs and what it would take to make an output with a repeatable, engineered, "voice".

The guitar outputs are becoming widely accepted, Gibson has purchased a mfg license, Triad Magnetics is purchasing licenses for all of the "voices" and now it looks as though Vox may also purchase a license for a "voice".

The cables are finished and I know how to "tune" them from exceptionally clear and dull to exceptionally clear and circus band vivid, with the same music and cables, by altering the amount of dielectric material that is closely adjacent to the cables. There is a bit more to it, but that amount of information would get you to 90% of what I know here.

This just naturally led to experiments in the source and amplification boxes. Little or no effect was found on the signal side, just as you would expect, if there was no "shortage" of electrons, but on the ground side, it was quite a different story.

And, here we are.

Bud
 
Hi Jeff,

Faux Litz, with no strand insulation and a PVC jacket or a serving but still no individual strand isolation. And there is a lot of it around.

I would leave the serving on if it is Nylon. You may end up removing some of it, but Nylon is an excellent dielectric material. If it is Cotton, also leave it on and just use shrink tube over it. Works just fine. If it is an Orlon / Cotton mix, try it as is, though the Orlon has a pequliar sound to it in other cable uses.

Hortono,

The Litz pieces I use are 150mm long, with three pieces of 15mm long of shrink tubing fully shrunk down over the Litz.

The PCB in the Sony looks like regular FG material. The traces and ground lands and traces are run on the same side of the board with the components mounted in both surface mount, from the trace side and through hole mount from the opposite side.

Bud
 
Member
Joined 2006
Paid Member
Hi All,

I found an article about "Image planes on Printed Circuit Boards" at http://www.hottconsultants.com/pdf_files/image_plane.pdf

I found Figure 11 very interesting and the paper serves to confirm what BudP suspects is going on with regard to signal interference (SI) - although the writers go about things in a different manner. The information is a bit dated as it was written in 1990 but Mr. Ott is still giving lectures on the subject according to his web site.

I offer this information in an attempt to show that this is indeed a field of study by EE's in the area of designing equipment to meet requirements for Electromagnetic Compliance regulations (EMC) regarding emissions of EMI/RFI radiation. There are a number of web sites and articles that deal with the subject matter if you need more information.

Please remember that I didn't write the paper and I sure as blazes do not claim to be an expert in this field of study. If you have some questions about this paper I would suggest you contact the author about them. I'm just the messenger - and I'm wearing Kevlar!!!

Another couple of interesting things to review can be found at http://www.ce-mag.com/archive/02/Spring/kimmel.html

And http://www.seas.upenn.edu/~engheta/Exotic_Surfaces.htm

Just some thoughts for the day.

:cheers:
 
Hi BudP,

I also mostly examining the effect of your litz in high frequency, dielectric effect and radiation effect, and considering the enviroment they're being used.

I always hate PCB, yes they are wonderfully practical, but sticking to FR4(or even teflon) material and 2D layout is bound to have drawbacks.

mix and match is needed for optimum circuit peformance

Litz for reinforcing ground is a surprisingly good idea.

I'll let you know soon, after I finish my examination.


Hartono.
 
Hi,

Hartono,

one other factor with litz vs a groundplane is that the litz will have less eddy currents. I have seen a PCB grounding system implemented where the ground returns comprised multiple parallel tracks as opposed to one wide track to combat exactly this "problem" if it can be characterised as such.

This is another thought provoking thread in addition to the Enable thread, my thanks to you Bud for your generosity in sharing.

Rob.
 
"one other factor with litz vs a groundplane is that the litz will have less eddy currents"

Yes might be true, and also maybe application specific in this case. Some application benefit more than others. Actually I think it's not that Litz have less eddy currents(EC) ,if seen from magnetic point of view, the EC on one conductor of the litz cause EC in adjacent conductor thus reducing inductance / skin effect.

If seen from speed of electron point of view, the field generated by the current passing through the conductor is crossing over another conductor, and since electron speed via conductor is faster than air , hence less skin effect . Basically the same thing.

Hartono
 
Rob, Hartono,

You are certainly welcome to this foolishness. If designers were able to convince the cost accountants that a true poured ground plane with instrumentation implementation was a real benefit and even find a way to shmaltz it up for the lazier marketing groups, this Litz wire issue would never have arisen.

The only reason it affects performance is because of poor ground plane implementation, forced upon EE's, who not only know better, but would utilize a proper one in a heart beat. I have all of the proof of this I need in my headphone amp, in which all circuit boards have a poured ground plane and instrumentation grounding. The Litz wires, that are so successful in the CD player with it's cut to the bone circuit layout, provide only the slightest difference in the headphone amp. And, I cannot even claim it is a "better" sound, just slightly more low level, wide band coherence, in information content.

Bud
 
update to this thread

Thanks to Bud and Sue's hospitality, I recently got an opportunity to hear (among other wonders) the difference that of one of these little pigtails can make in an already very good sounding, but not trendy-spendy system, including speakers with which I'm intimately familiar.


I don't have the technical acumen to understand exactly what's going on here (who does, really) nor the lexicon or writing skills to poetically describe what I heard, so a few brief phrases:

refined resolution of very small signal levels ("deep down dynamic range")

coherency

image dimensionality and stability


For example, one of my reference tracks is on a piece* that includes 3 female vocals in tight harmony, and accompanied by handheld percussion as well as fiddle, guitar, a small accordion, and other instrumentation of which I lose track - the vocal tone painting is too overwhelming .
Even with Bud's mojo EnAbl treatment on the driver cones, the addition of this little piece of wire focused the articulation of the singers' individual breath intakes while others are still expressing, and the difference between the sound of the accordion when squeezed and pulled more precisely and 3 dimensionally than I'd yet heard it.

*Wailin Jennys /Firecracker ; track 5 "Swallow" (the bird) This album grows on ya even more than the earlier 40 Days


Another great track for low level resolution is Kevin Mahogany,/ Double Rainbow "Since I fell for you"- the resolution of texture, pressure and tone on the swirling brushwork is simply more convincing with the coherency of "information packets" is retained.

Of course I know none of this is not possible, so "Bud, you've got some 'splainin to do" ;)
 
Yes, this demonstration was incredible.

When Bud removed the loops, it was like hitting the "mono" button by comparison.

I can't explain it and I'm a rediculous skeptic, but this one works!

It is an excellent way to increase the downward dynamic range a driver can have.

Cheers!
 
greenvalve,

Here is a picture of a piece of true Litz wire, 140 strands of #40 AWG insulated magnet coil wire from our transformer business), tinned but without the pieces of shrink wrap tubing I use for dielectrics.

The encased item is just another piece of Litz wire, doubled back on itself, with appropriately sized pieces of tubing, The yellow end cap corresponds to a specific amount of dielectric.

This amount is what is used in the loops I have connected throughout my system grounds, to maintain small signal coherence and dynamic color. It also works perfectly for Lowther PM6A's and Fostex 127 E's, mounted on the driver ground lug or ground post on the box. I will sell you some but you have to PM me.

Bud
 

Attachments

  • electron pool.jpg
    electron pool.jpg
    52.4 KB · Views: 2,584
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.