Never-Connected power supply

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
Does anyone have any information on how this power supply design works? I read something about it in either Stereophile or TAS, so I had a look at their web site: never-connected.com , and the sites of their dealers for upgrades to existing equipment; soundodyssey.com & trichordresearch.com .

The concept seems interesting, and the boards look simple enough. Is this something DIYers can duplicate?

--David
 
Imagine you have a bridge running straight off the mains, no transformer.

You have a filter cap.

You have a load for the filter cap (audio amplifier).

You have a DPDT switch on the filter cap that can either connect it to the mains, or the load, but not at the same time.

The load is 'never connected' to the mains.

If the switch goes FUBAR you can get fried.
 
The name "Never Connected" may be good marketing but is otherwise a tad silly. In there more common type of power supply there is no wire that connects the DC rails of an audio device to the mains simply because there is a trandformer in between. In this sense, most audio equipment is "not connected". But so what? Dependeing on how you look at it either you are already "Never Connected" or so long as you are getting power to your equipment you are "unconditionally cxonnected".

Even a regenerator is connected is a significant sense - if the mains go out, the music stops.

If one is a fanatic about the suject, onr could in principle use the mains to charge a huge bank of batteries and then run an inverter from the batteries. Then there could be endles discussion about the appropriate exotic components to use in our diy inverters!
 
Thanks Kermit. I was just going through the recent issues trying to find the reference.

Here is the interesting part:

"The Never-Connected approach takes advantage of the sometimes overlooked fact that the power supply actually takes up power from the mains over only a few degrees of the complete sinewave - the breif period in the cycle when the voltage on the secondary is higher than that on the supply rail, which is a necessary precondition for current to flow. NC's innovation is to isolate the equipment itself entirely from the wall during this part of the cycle by diverting the incoming current into a storage capacitor, then releasing it to the supply side once the AC current stops flowing."
 
I feel very sceptical about the claims Never-Connected makes. I really would like to see the patents they have on their "revolutionary" power supplies.

If I have to judge from the feeling of my guts the "Never-Connected" PSU designs are:
- An ordinary regulated PSU, since most transformers automatically create a seperation from the net, if not all transformers do that.
- A galvanically seperated AC-DC converter in the line of a FlyBack converter or Fly-Forward converter. In both cases a seperated switchmode psu
- A Power Factor Controller of some kind that features a galvanic seperation. In all cases this will also be a switchmode psu.

I regard none of these options "revolutionary", though proper installment might help reduce noise from the net..

Bouke
 
Disabled Account
Joined 2003
net-david said:
NC's innovation is to isolate the equipment itself entirely from the wall during this part of the cycle by diverting the incoming current into a storage capacitor, then releasing it to the supply side once the AC current stops flowing."

so the functioning of the psu on the equipment doesn't change, except that rather than drawing from the mains, it draws from that capacitor. How did that improve the sound?
 
millwood:

>so the functioning of the psu on the equipment doesn't change, except that rather than drawing from the mains, it draws from that capacitor<

_If_ what NC is doing is similar to what we were doing, the above is correct.

>How did that improve the sound?<

Don't know about NC, but when we were doing our experiments, the sound subjectively became less noisy and cleaner (less sense of haze and grain), but the bottom end was decidedly anemic, and the dynamics seemed weaker as well. In other words, a mixed bag. Some of our listeners considered the results to be an improvement, while others thought otherwise.

FWIW, in our case we first used the flying-cap/bucket-brigade scheme, and followed it with 3-terminal IC regulators (I think of the fixed-voltage type). May be interesting to re-trace the flying-cap/bucket-brigade path, but be more careful about the design of the regulators that follow.

hth, jonathan carr
 
The name "Never Connected" may be good marketing but is otherwise a tad silly. In there more common type of power supply there is no wire that connects the DC rails of an audio device to the mains simply because there is a trandformer in between.

But in this case you have an AC connection between the mains and load and potentially affected by any noise present on it, and the self-generated noise the narow conduction angle of the rectifier diodes creates.

The NC supply never connects load to mains, according to the blurb, and it is therefore isolated from any mains noise present.

In fact reading net-david's post makes me think it isn't a switched capacitor supply now, in the conventional sense, but simple a circuit that disconnects the load from the reservoir capacitor only during the brief conduction angle the rectifier diodes are forward biased. It may be using the rectifiers as the switching element, but one would need to be careful that increased load would not cause the circuit to drop into conduction at any point.

Need to think about that a bit more....

Andy.
 
There isn't really much info on that site, and I might have got
it all wrong. However, if the load is disconnected from the
capacitor during charging, there must be a second capacitor
to supply the load meanwhile. In the simplest case, we could
use two capacitors (per rail, that is) s.t. during every odd half
cycle one is being charged and disconnected from the load and
the other is disconnected from the rectifier but connected to the
load. During even half cycles it is the other way around. I don't
know if this is what they actually do, but anyway this is such an
obvious thing to do that even I have had that idea (why didn't
I patent it :) ). However, I also anticipated problems with it.
At some point we must switch capacitors for the load, from a
partially discharged one to a freshly charged one. If we do this
abruptly, with a relay or switching transistors we get a transient
which will also cause problems, maybe more problems than the
usual inrush current when charging the capacitors. A resistor
between cap and load switch might help, but introduces other
problems. We could also do it softly, since both caps can be
connected to the load for more than a half cycle, but then we need non-switching transistors that dissipate power during
the switchover period, giving us another headache.
Whether switching abruptly or softly, we onviously need twice
the number of capacitors compared to a normal PSU. This might still be a good idea, but it is not obviously an improvement over
a normal PSU.

This solution, wheter it is the NC one or not, should
not be a big problem to DIY, at least not if skipping the soft
switching. This is all my idea, whether someone else has come
up with it too or not. You are free to use it for DIY, but I can't guarantee you are not violating someone elses patent by doing so.
 
Bakmeel: I've been looking for those schematics for years. I suspect that they disappeared in one of the several hard-drive failures we've experienced since then. But the one linked from this post:

http://diyaudio.com/forums/showthread.php?postid=151191#post151191

is similar in concept. This particular schemartic was done by Isao Shibazaki for his "DSIX" active digital cable, which was in turn inspired by an active digital cable that we did, which was called the ADI. I discussed the design features of the ADI (and prototypes) at length with Shibazaki, and among the aspects discussed was the bucket-brigade power supply concept.

At Connoisseur we were also experimenting with multi-stage bucket-brigade designs with progressive frequency multiplying, but those designs appear to have been lost.

BTW, always make multiple backups of important design data, each in a different location (_not_ different partitions on the same hard drive). I've lost more schematics and pcb layouts than I care to admit to application crashes and drive failures. Some of those lost designs would certainly come in handy now and then.

jonathan carr (currently reinstalling Win2000 on a new HD to replace a failed Maxtor)
 
Konnichiwa,

To shed a little light - I do not know the circuit either and have been unsuccesfull finding the European Patent NC claims to have.

I can comment on the physical side of the implementation as I requested details.

For re-fit into existing gear the NC supply replaces the rectifier bridge with a small circuit board, containing a few (very few) components. I'd not venture further guesses, but the NC supply is a very simple, cheap and basic "trick"with extremely easy retrofittability and I remain convinced that other methodes of clening up the mains are as effective or more so....

Sayonara
 
Konnichiwa,

ALW said:

Well, since the vast proportion of mains noise is generated by the kit we plug into it, there's always more than one way to skin the cat!

True.

ALW said:

Anyone tried PFC in an audio supply?

What for? A sensible, valve rectified and choke input (or capacitor input with very small size input capacitor) shows very little benefit from this anyway and I think the same holds true for Batteries (the only excuse for not using batteries is the requirement for too high Voltages to run Valve Circuits).... ;-)

Sayonara

PS, I personally would be interested in getting a well filtered, PFC Input and Push-Pull Linear RF supply (meaning Class B Sinewave High AF Amplifier) UNIVERSAL BOARD to be used with an external RF Transformer and rectification/filtering for the output to make any kind of voltage one likes, clean, very low noise and with minimal impact on the mains quality. Any takers?

My experiemnts with modified inexpensive, generic switching supplies showed a lot of potential so far, but all the mods etc make an unholy mess....
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.