Questions For Interview With Bruno Putzeys & Lars Risbo About Purifi Audio

[maty]

57 minutes ago, mocenigo said:

Yes and no. Class D amps have a phase rotation that is a linear function of the frequency. This is just a delay. So in fact this is often much better than irregular phase rotations as in A and AB amps. Luckily, the rotation of the latter are much more contained.

 

NCore also has much smaller rotations than ICEpower and, it seems, ICEEdge - I expect the same for Eigentakt

 

Still, it would be VERY INTERESTING to hear Putzeys take on the matter.

 

 Roberto

 

 

Bambino, hai un regalo per posta.

 

Sag mir was du denkst.

[mocenigo]

I remember to have seen some examples, I do not have them handy.  I should have added “some A and AB amplifiers”... I apologise.

 

This said, a linear phase rotation is fine.

[firedog]

8 hours ago, Em2016 said:

Hi @The Computer Audiophile

 

I've seen comments across the internets (these are not my thoughts) that the switching frequency is the only 'issue' with Class D - one argument is it's not currently high enough with Hypex and ICE designs and this then needs aggressive filtering that can affect the treble performance making everything sounds smooth and warm.

 

Are there merits to pushing this switching noise higher and higher, to make filtering easier? Some say it needs to go to 5 MHz and beyond to make filtering easy without affecting the top end, again making everything sound warm and smooth... again these are not my thoughts, just what I've read over the years.

 

And what's the switching frequency with this Purifi amp project

 

Would love Bruno's thoughts on these (I've seen all the discussion on both sides so let's not get into it here).

Most seem to think at a switching frequency of 1ghz or 2ghz would solve the problems of class D and make it the superior amplifier mode. The Gallium based amps supposedly achieve this. The other Class D on the market have switching frequencies of from 300-600hz, AFAIR. So yes, it would be interesting to ask Bruno about both the switching frequency idea and the use of Gallium

[psjug]

You (Bruno) once described your pursuit of a power DAC as "folly".

https://www.soundandvision.com/content/bruno-putzeys-head-class-d

 

Do you still feel that a power DAC cannot hope to compete with your analog designs?  What about direct digital designs with feedback?

[Matias]

1 hour ago, firedog said:

The other Class D on the market have switching frequencies of from 300-600hz, AFAIR.

 

Indeed the 1ET400A module is at 500kHz according to the preliminary spec sheet.

https://6moons.com/wp-content/uploads/audioreviews/purifi/1.pdf

[mocenigo]

Do you mean 1-2Mhz? You can easily reach this with GaN or GaAs. The ICEEdge chipset can control switching up to 40Mhz so it seems ideally designed for new materials.

 

Most amps based on silicon have a switching frequency up to 500Khz, but IIRC some (Megaschino? Some Pascal modules) reach 700khz.

 

 Roberto 

 

1 hour ago, firedog said:

Most seem to think at a switching frequency of 1ghz or 2ghz would solve the problems of class D and make it the superior amplifier mode. The Gallium based amps supposedly achieve this. The other Class D on the market have switching frequencies of from 300-600hz, AFAIR. So yes, it would be interesting to ask Bruno about both the switching frequency idea and the use of Gallium

 

[firedog]

8 minutes ago, mocenigo said:

Do you mean 1-2Mhz? You can easily reach this with GaN or GaAs. The ICEEdge chipset can control switching up to 40Mhz so it seems ideally designed for new materials.

 

Most amps based on silicon have a switching frequency up to 500Khz, but IIRC some (Megaschino? Some Pascal modules) reach 700khz.

 

 Roberto 

 

 

Yes, of course kHz and MHz. Sorry for the sloppiness.

[audio.bill]

Just as another point of reference, the NuPrime Evolution One amp which uses their own Class D design (not someone else's module) has a switching frequency of 700kHz along with a 1MOhm input impedance. According to their claims: "High-speed computing simulations were performed for the Evolution One design, in order to achieve the highest input impedance and the lowest phase delay."

 

[psjug]

The latest Qualcomm DDFA can run at 800kHz.  I think that is the frequency that the Sonos Amp uses.  I have one and there is a lot of ripple at 800KHz, so I think the frequency needs to be much higher for switching noise to be tamed by a simple LC output filter.

[Shadders]

30 minutes ago, psjug said:

The latest Qualcomm DDFA can run at 800kHz.  I think that is the frequency that the Sonos Amp uses.  I have one and there is a lot of ripple at 800KHz, so I think the frequency needs to be much higher for switching noise to be tamed by a simple LC output filter.

Hi,

I examined the product brief, and the THD is about 10x higher than the Purifi preliminary module. The problem is we do not know what THD measurement filter was used for the Qualcomm unit, so the Purifi and Qualcomm may be equivalent.

 

Amplifiers were generally measured with a THD filter of 80kHz, for class A/B designs, but for class D designs many reduce this to 22kHz, due to the excessive noise above 20kHz compared to class A/B. It makes the numbers look better.

 

Then, the proponents of class D respond that no one can hear any information above 20kHz, so the 22kHz measurement filter used is valid.

 

Classic case of moving the goal posts.

 

Regards,

Shadders.

[Matias]

Will they also go after pro audio customers (supply modules and drivers for instrument amplifiers, PA speakers) or are they more focused on the high end audio market?

 

Besides amplification and drivers, possibly DACs as I have also asked before, any other product lines they have in mind?

[asdf1000]

On 5/18/2019 at 6:47 PM, mocenigo said:

Now, at a switching speed of 500khz you get about 5 bits of depth at 20Khz and 6 at 10Khz which is beyond what one can hear (even a very very young person) so the signal as perceived by the brain is already perfect. What higher switching speeds could improve is timing, and it is already quite precise.

 

Great points that Bruno will hopefully clarify but as per my earlier questions, I believe one common argument people make for higher switching speed is actually for output easier/gentler filtering? Whereas lower means more aggressive filtering and some people say this is what causes Class D amps to often sound smooth/warm (aggressive filtering affecting top end frequencies).

 

Similar to filtering quantization noise with DSD... higher DSD rates (higher than DSD64) allow for simpler analogue filtering stage...

 

[monteverdi]

I think filtering is the main issue with class D to completely surpass classical amp topologies and obviously higher frequencies are making filters easier. I am using class D to drive my woofers and therefore filter design is less of an issue but even then I can hear differences between different designs preferring Anaview AMS over Hypex. I would be very interested to try these Purifi modules!

[Matias]

Everyone is so worried about out of band noise and forget that class D has the lowest amount of noise and distortions in band compared to other classes...

[barrows]

On 5/18/2019 at 5:14 AM, mocenigo said:

Do you mean 1-2Mhz? You can easily reach this with GaN or GaAs. The ICEEdge chipset can control switching up to 40Mhz so it seems ideally designed for new materials.

 

Most amps based on silicon have a switching frequency up to 500Khz, but IIRC some (Megaschino? Some Pascal modules) reach 700khz.

 

 Roberto 

 

 

Mr. Putzeys is on record as saying that increasing the switching speed of class D amplifiers offers no advantage, so I doubt he has much interest in using Gallium based devices.  So far, the only Gallium based class D amp I am aware of is from Merrill Audio, and its specifications for noise and distortion are quite a bit worse than Ncore.

I find it quite amusing that some folks here say things "the problem with class D..."  I am unaware of ANY problem with Ncore, in my system, in direct comparison, NC-400 based amps have out performed a number of highly regarded class A/B amps of much higher cost.  I am excited to hear the next generation Purifi amps!

 

Oh yeah, a question: is there any chance for a Bruno Putzeys designed DIY DAC module using the 100 MHz 1 bit discrete FIR approach used by Mola Mola from Purifi?

[Ralf11]

Technics

[asdf1000]

2 hours ago, Matias said:

Everyone is so worried about out of band noise and forget that class D has the lowest amount of noise and distortions in band compared to other classes...

 

There's a few companies making DSP speakers that choose class D for all drivers but the tweeter... one current popular example is KEF LS50W.

 

It might be related to the switching noise and the need for aggressive filtering somehow affecting SQ (affecting the top end only).

 

I'm only guessing but would be great to know Bruno's thoughts on it.

 

[asdf1000]

1 hour ago, barrows said:

Mr. Putzeys is on record as saying that increasing the switching speed of class D amplifiers offers no advantage

 

Can you share where this record is?

[fas42]

This is pretty good ... https://www.soundstageultra.com/index.php/features-menu/general-interest-interviews-menu/455-searching-for-the-extreme-bruno-putzeys-of-mola-mola-hypex-and-grimm-audio-part-one

 

I particularly like his emphasis, here, on being extremely fussy about isolating the module from external muck - and his thoughts on how feedback as conventionally used is nowhere near good enough.

[barrows]

48 minutes ago, Em2016 said:

 

Can you share where this record is?

I cannot remember for sure where I read this, it was awhile ago, but it probably was on the massive Ncore thread at DIYaudio.com.  At the time of the introduction of the Hypex Ncore tech, Mr. Putzeys graciously answered many questions about his approach to class D.  IDK if the thread is still up, and given the rather poor search engine on the DIYaudio site I would not want to go back and try and find the exact quote, I have better things to do with my time.

[asdf1000]

29 minutes ago, barrows said:

I have better things to do with my time.

 

Fortunately we have a chance for Bruno to clarify, by @The Computer Audiophile passing on these questions.

 

Maybe Bruno's position has changed over time, with new findings...

 

[Miska]

4 hours ago, barrows said:

I find it quite amusing that some folks here say things "the problem with class D..."  I am unaware of ANY problem with Ncore, in my system, in direct comparison, NC-400 based amps have out performed a number of highly regarded class A/B amps of much higher cost.  I am excited to hear the next generation Purifi amps!

 

Traditional analog PWM class-D amps tend to have aliasing problems, when for example used with DACs that have leaky filters or images due to incomplete reconstruction.

 

[Shadders]

8 hours ago, Matias said:

Everyone is so worried about out of band noise and forget that class D has the lowest amount of noise and distortions in band compared to other classes...

Hi,

Can you provide references to support this claim ?, and which classes of amplifier topology you are referring to ?.

 

Thanks and regards,

Shadders.

[Matias]

1 hour ago, Shadders said:

Hi,

Can you provide references to support this claim ?, and which classes of amplifier topology you are referring to ?.

 

Thanks and regards,

Shadders.

 

I refer to class A and AB with linear power supply, which together should account for something like 90% of the high end market.

 

As for references, check out the THD x power graphs, or 19+20kHz x frequency graphs, that Stereophile, SoundStage and Audio Science Review forums have, and compare nCore against similar prices or even ultra expensive amps.

 

Not that I am a hardcore objectivist but still.

[Matias]

Question: would it make sense to use Eigentakt for low power amplifiers like desktop or portable headphone amps, or even embedded IC chips (DAPs, smartphones) or opamps?