Hypex Ncore

Status
Not open for further replies.
Quote from Bruno regarding the LPS in the Theta Prometheus vs the Hypex SMPS1200A700:

"Regarding the supplies, the proof is printed black on white in fig 4. in the Stereophile review of the Prometheus. Those 60Hz related components are simply nonexistent in the Kalugas. The most likely cause of those in the Thetas is the big toroidal transformer combined with the big capacitor tank and the rectifiers + associated wiring putting out a stray field. It's often said that toroids don't have a stray field. They do, and it's located where the wires come out. The field you get there is very nasty, consisting of 60Hz harmonics going all the way up into the kHz region. It's so strong that it is easily picked up by the speaker wiring so the amp output could be clean and you'd still get a buzz in the speaker as a result of the few inches of wire between the amp module and the chassis.

The kalugas have no discernible hum or buzz and the construction of the power supply is part of the reason (the internal wiring explains the rest). Because of the fact that rectification happens on the primary side, currents are smaller. Also, for the same amount of energy storage the capacitors are physically much smaller so you can do a circuit board layout that has minimal current loops. The power transformer is run at 100kHz meaning it too can be smaller and its stray field (which contains no 60Hz related component!) is very easily shielded.

The designer of the SMPS has also done a custom job for Manley for microphone preamplifiers. When the people there realized that suddenly all the hum and buzz they'd been fighting for years was completely gone they went ape. It's really strange how lay people (especially those trained in the art of audiophilia by their peers who write for audio magazines) seem to think that switching power supplies are noisy. They're not. Given half a chance they're much quieter. All you need to take care about is of course the 100kHz and associated harmonics. Once I had that down the SMPS sounded much better than the big iron and copper affair that's incorrectly known as a "linear supply".


Linear versus Switching PSU in Mola-Mola Kaluga | On a Higher Note
 
Great quote there, except somewhat confusing here :

Because of the fact that rectification happens on the primary side, currents are smaller.

Which currents is Bruno referring to? Having rectification on the primary side tends to degrade the power factor - hence peak currents from the mains are likely to be bigger, not smaller. So I must be missing something.
 
Great quote there, except somewhat confusing here :

Because of the fact that rectification happens on the primary side, currents are smaller.

Which currents is Bruno referring to? Having rectification on the primary side tends to degrade the power factor - hence peak currents from the mains are likely to be bigger, not smaller. So I must be missing something.

I'm not sure, it's Bruno's quote.
 
Great quote there, except somewhat confusing here :

Because of the fact that rectification happens on the primary side, currents are smaller.

Which currents is Bruno referring to? Having rectification on the primary side tends to degrade the power factor - hence peak currents from the mains are likely to be bigger, not smaller. So I must be missing something.

That is true for a traditional diode rectifier SMPS, but as far as i know, the hypex design uses synchronous rectification.
 
Do enlighten me as to how using synchronous rectification (presumably using MOSFETs) changes the picture?

If the power supply was power-factor corrected then yes, peak currents are going to be lower. However that's not what Bruno's claiming - he's claimed that lower currents are the result of primary side rectification.
 
Its a long paper - where do they address reducing mains side currents? Care to offer a page number? I've scanned down about 16 pages and the focus is on efficiency gains compared to normal diodes.

Actually the first graph, figure 2.1, illustrates the situation pretty well. A synchronous converter avoids the non-linearity (and associated surge peaks) of a diode rectifier.
 
You mean this one (attached)? I must still be missing something - how does the FET reduce the mains current compared to the diode?
 

Attachments

  • Julf.jpg
    Julf.jpg
    31.5 KB · Views: 225
Go on, I'm still not following. Please elucidate.

<edit> I note we're assuming the use of synchronous rectifiers is on the primary (mains) side. Whereas for efficiency reasons it would make more sense to have them on the secondary where currents are higher. Put that to one side for a moment though while you explain about the 'surge'.
 
Last edited:
And how does that compare with a harmonic analysis of a synchronously switched MOSFET?

The synchronously switched mosfet doesn't have the strong non-linearity / forward voltage threshold that causes the harmonics.

But let's try to cut the chase - do you disagree with the statement that a synchronous switching power supply can have much lower harmonics and better power factor than a "traditional" SMPS (and a linear power supply)?
 
The synchronously switched mosfet doesn't have the strong non-linearity / forward voltage threshold that causes the harmonics.

Well I get that a FET is going to look resistive rather than having an exponential forward characteristic. But the FET is being switched externally, so that's a non-linearity. I don't see how its not the switching (rather than the precise forward characteristic) that causes the harmonics - what am I missing?

But let's try to cut the chase - do you disagree with the statement that a synchronous switching power supply can have much lower harmonics and better power factor than a "traditional" SMPS (and a linear power supply)?

I don't know - it certainly can have, but the better power factor would come from PFC circuitry prior to the transformer, not from replacing the diodes in the rectifier with FETs.
 
Status
Not open for further replies.