The biggest "design flaw" is the AC safety ground, which is the single most common cause of interchassis leakage currents and the resultant noise.
Sadly, the vast majority of high end audio manufacturers haven't figured out how to design equipment chassis to meet Class II (i.e. double insulated) specs. So they burden their customers with safety grounded equipment chassis.
And even in those instances where the design might actually meet Class II specs, they still typically fit them with three pin IECs with the safety ground attached to the chassis.
Say what you will about mass market audio manufacturers, but at least they know how to design chassis to Class II specs and don't burden their customers with AC safety grounds.
It's nothing to do with cable grounds being "dirty." It's that the cable system is earth grounded at a different point than the AC mains which gives rise to the same noise-producing interchassis leakage currents that are brought about due to the AC safety ground.
Oh, and for what it's worth, a literal earth ground has absolutely no relevance whatsoever to an audio system.
se
I've been using cotton covered litz for some time now with very good results.
Are you just making your own litz by bundling up strands of enameled wire? It seems like it given the even number of strands (most all stranded wire including litz uses an odd number of strands).
Have you ever tried a proper Type 2 litz?
Type 1 uses a simple bundle or twist. And while litz wire is commonly associated with ameliorating skin effect, Type 1 litz does nothing in this regard and is fundamentally no different than uninsulated stranded wire as the individual strands tend to maintain the same radial distance from the center from end to end so you still end up with higher current densities in the outer strands compared to the inner strands as frequency increases.
Type 2 litz takes multiple bundles of twisted wires and twists those together in a "rope lay." The result of this is that on average along the length, each strand occupies every radial position from the center so that the current density is the same in each of the strands and this is what ultimately ameliorates both skin effect and proximity effect.
What I ended up choosing was a Type 2 litz wire made up of 165 strands of 46 gauge wire in a 11/5/3 configuration, meaning it starts out with 11 strands twisted together, then 5 bundles of those are twisted together, and then 3 bundles of those are twisted together.
I cover the litz with cotton braid and then braid four strands of it together and pair them up at the ends per Milloit's 1958 patent (2,958,724).
I don't know just how significant skin and proximity effect are for audio frequencies, but with this fine stranding and twisting arrangement, it's virtually non existent. It sounds wonderful and nearly as much as anything else, I love the extreme flexibility that if offers.
If you'd like some to play with, shoot me a PM. I have an order coming in next week and about a 20-30 foot remnant on the old spool I'd be happy to send you.
Works for me.
And I'd go quad braid over twist. It provides greater self-shielding over a twisted pair or twisted quad (i.e. Starquad), it's self-constraining, and has excellent inductive and capacitive properties.
Not so weird when you think about it.
The inductance is a series element so it effectively increases the source impedance as seen by the tweeter which can cause some high frequency rolloff.
se
Certainly a possibility. However your comment provides absolutely no enlightenment.
So what is it exactly that I don't understand and how is it that the AC safety ground isn't a significant problem with respect to noise?
se
I quite agree. The so called safety ground does not really ground the way it works on paper. If it did, Y caps would not need to have limited size. I've been trying to look for two prong chassis mounted power connections that can have the same current rating as the three prong connectors. Seems that there are none that are commonly used. Probably this is why high end audio manufacturers use the three prong devices.
Right. It's only a "ground" in the sense that it's a return path for fault current.
In the US, the safety ground's connected to the equipment chassis and feeds back to neutral at the service panel. In the event that the chassis goes "hot" due to a failure, the safety ground provides a path back to neutral and hopefully keep the chassis potential to non-lethal levels.
And the reason for that is precisely why I say that the safety ground is a "design flaw" when it comes to keeping noise levels low.
The safety ground is typically in rather close proximity to the AC hot lead. This results in capacitive coupling between the two and subsequently chassis leakage currents.
If the voltage drops due to the impedances of the power cords are not equal, then it results in a potential difference between equipment chassis. And if the component's signal grounds are tied to the equipment chassis you end up with current flowing through the grounds of the cables interconnecting them. Since the ground leads will have some amount of resistance, there will be a voltage drop across it and this is seen by the input of the downstream component.
Y caps are kept to a limited size to keep leakage currents low.
The IEC 320 C17 and C18 connector pair are two prong and rated for the same 10 amps of current as the more ubiquitous three prong C13 and C14 pair.
The C18 inlet is easy enough to find, but user replaceable C17's might not be so easy to find.
Of course you can simply eliminate the IEC connector entirely and use a captive cord. But that would probably cause the manufacturer to lose sales.
Perhaps. I have some other theories, but I'll keep them to myself.
se
Once you find some time and interest in doing these audition I'd be interested in your findings comparing Electron Pools with simple "termination" of speaker cables at the chassis side.
'No more than a simple connection to the top and / or bottom plate of the speaker is needed.
With respect to the zero cost and the little effort required, I found this cute little trick pretty rewarding in my SS-amp setup.
Michael
Do you know of any source that has C18 rated to 15 amps?...
The IEC 320 C17 and C18 connector pair are two prong and rated for the same 10 amps of current as the more ubiquitous three prong C13 and C14 pair.
The C18 inlet is easy enough to find, but user replaceable C17's might not be so easy to find.
...
se
This had been posted many years ago in this forum. It works like a charm to me, but it does not seem to have any audible effect to some others. I think this is related to design and performance of the driver.Once you find some time and interest in doing these audition I'd be interested in your findings comparing Electron Pools with simple "termination" of speaker cables at the chassis side.
'No more than a simple connection to the top and / or bottom plate of the speaker is needed.
With respect to the zero cost and the little effort required, I found this cute little trick pretty rewarding in my SS-amp setup.
Michael
Yea, here for example:
http://www.diyaudio.com/forums/parts/102180-groundside-electrons-5.html#post1759249
any other places?
Well, thought I bring it up again in case we possibly get some further "anecdotal reference" from Lynn or others, as it fits somehow to the discussion about grounding, shielding, potential differences, antenna effects, lines and its distributed R C L components.
Michael
'No more than a simple connection to the top and / or bottom plate of the speaker is needed.
With respect to the zero cost and the little effort required, I found this cute little trick pretty rewarding in my SS-amp setup.
Michael
I found the same thing independently a couple years ago. I was hearing a faint radio station pick up on one speaker. The system was multiamped. This speaker was powered by a SS amp. I surmised that the speaker cable was acting as an an antenna, and interacting with the amp via the feedback path. Simply touching, or even moving very close to the speaker frame or phase plug, stopped the interference. Connecting the common side to the speaker basket also solved the problem. Didn't try the other terminal.
The affected speaker was on the other side of the room from the amp - though the cable length was the same for both channels. I never tried long interconnects and short speaker cables.
Sheldon
So it seems hospital grade transformers are not favorable for audio according to you experience? Sure. Hospital grade stuff addresses the frequency above the audio spectrum, but what I gather from you long explanation, the required aspects for good audio reproduction are not considered.
My experience of noise in a listening room is that noise measured at the terminals of the driver is about 70~80 db down from 2.5V, except for 60Hz and 120Hz, which seem to be electromagnetically coupled.
Last edited:
Far as I'm aware, if it were rated to 15 amps, it wouldn't be a C18.
There's a C23 and C24 pair that's rated at 16 amps though.
se
Choke input PS filtering is not really practical with Class AB amps. The large change in current is not conducive to choke selection.
Jensen input transformers are GREAT for reducing hum-generating ground loops. However, standing DC base input currents to bipolar input amps are not desirable for transformer cores. FET input do not have this disadvantage. High beta transistors also help if bipolar. Also, the DC resistance of the input transformer's secondary shunts the amp's input resistor, which upsets the DC offset of the amp, as well as input noise optimization. Therefore, the feedback network has to be changed, and likely forces the removal of the shunt cap for DC balance. Alternatively, a series cap on the input, after the transformer, would be needed. Therefore, you can't just plop an input transformer ahead of any amp without concideration of the circuit. But when done right, NO hum at all!
Lynn, in your "solutions" to minimize stray capacitance, you failed to mention the free and very beneficial approach of simply reversing the PS leads to the transformer primary. While your amp is connected to the cheater plug, measure the AC voltage from the chassis to the ground on the socket. Reverse the cheater plug (if it is not polarized) and measure again. One direction will have less voltage measured than the other. Use that arrangement. I have found an order of magnitude difference. Reverse the primary leads in the amp, if necessary. If you don't want to go inside the amp, then reverse leads at the plug.
Last edited:
Not really. All grounds are grounded at the service entrance (same place) by code. The issue with cable grounds is that a potential ground loop is created from the cable box, back to the service entrance, and through the branch wiring to the amp(s). If the cable box is conductively coupled to the amp(S) via another lead, such as by audio out from your cable box or TV to your hifi equipment, a VERY large ground loop is created.
Last edited:
CineMags are every bit as good as the Jensens in my opinion, and don't cost as much. In a sense, the Jensens are CineMags, as it was Ed Reichenbach who designed the transformers for Deane and Tom has all of his father's engineering.
se