Hypex UcD400 reviewed

[peranders]

UcD400 has been reviewed in the danish/swedish "High-Fidelty" (No. 4 /2006). Good reviews was the verdict. I will translate the article later on if there is any interest for this.

[Bgt]

peranders, please do.
Thanx

[peranders]

The raw file can be located here:
http://sjostromaudio.com/ucd400_review.pdf 8 MB big but I will make it smaller + english text.

[Bgt]

just wondering why they've put the 2 modules so close to eachother with the first channel output coil radiating on the 2nd channel modulator(UCD)? No symmetry. Beats me
That must have an effect on the soundstage.
Anxious to read the story.

[peranders]

The amp is built by Hypex themselves, I'll believe.

[Eva]

Gapped coils placed like these will produce a stray magnetic field with flux lines perpendicular to the horizontal plane, thus they will only cause induction in other circuits laid across the horizontal plane. The daughter board is intentionally placed in a vertical plane to avoid induction from the coil or from the own tracks of the main PCB, that also produce stray flux lines perpendicular to the horizontal plane (not accounting for ground plane "shorted-turn" effects).

In practice it means that induction will be the minimum possible. Indeed, these "geometry" tricks are quite common in SMPS (and not so common in class D due to the rather clumsy layout abilities of most designers).

[Bgt]

Eva, my first UCD amp. I've made in the same way, the modules where a bit nearer to eachother though, and found out 1 channel was influenced by the others modulation. On a scope I could see the modulation change in the other amp.
It is not a very sophisticated way of placing switching devices like this in an audio app. My view though.

[Eva]

How did you measure that? Was there asymmetric cross-talk in the audio band?

Note that you will get induction on the own amplifier wiring and oscilloscope leads if you place them near the coils.

I agree that it's not a great idea to place a modulator just near a coil, but if they must be placed together due to space constraints, that's the best way to do it.

[Jan-Peter]

Hi,

Peranders thanks for the link! A translation in English would be wonderful....

It's unnecessary to discuss the best way of placing the UcD modules in the amplifier. The amp under review was one of our first samples for demo.

We would do it now in a totally different way (dual mono, BHC Slit foils, separate supply for the buffer op amp, etc....)

The purpose of this demo amp was to show how small you can build an UcD amplifier. And it's easier for reviewers to measure on a complete working amplifier, as on separate modules with transformers and power supplies.

Besides this we hardly find any influence when the amps are mounted very close to each other. However you WILL have problems when you are wiring the amplifiers in the WRONG way.

We have on our website soem recommondations for wiring.

Cheers,

Jan-Peter

[Eva]

Are you performing conducted and radiated EMI measurements on signal, speaker and power wires? Will you use any kind of common mode filtering to match SMPS "quietness" industry standards? That's important because any RF component radiated by wirings will be dependent on the audio signal, and it will be demodulated and turned into audio when picked up by other circuits.

By the way, one trick of which I'm taking advantage in high power class D is to use two series inductors laid in a dipole fashion. That produces inherent stray flux cancellation in the far field.