Could Class D really be that good?

@eddnog  I got in and pledged for two Monos with two 48v/5a PSUs. I felt fortunate making the last-minute decision and went with two independent PSUs because later today I saw this https://www.youtube.com/watch?v=wjWVSFCR5MI where the host experienced noise/interference issues using one block.

Ric,

I really admire what you do and wish I could learn that skill set someday. It's a pretty good review, and I congratulate you on the impressive 9.5 out of 10 score you received for your mod. I agree that measurements do not imply sound quality, and I believe the listening part of the assessment is greatly lacking in the ASR review. Nevertheless, it is suggested that you send your mods to ASR for his complimentary measurements. Coupling with the excellent listening review you have received, a potential outcome from getting more comprehensive measurements done is that the forum will get to know your mods, your sales will be boosted, and the community will benefit from it. A win-win situation.

A simple upgrade is to order the Fosi with the 48 volt power brick. Go to ASR to read the recent test report.

I did try the 48-volt PSU, but the truth is, it doesn't deliver better sound than the 32-volt PSU at normal listening levels or even at 80-85 dB. Surprisingly, the 32-volt PSU offers cleaner sound, leading me to believe that it is the better and less noisy SMPS option. Moreover, in terms of power requirements, the 32-volt PSU proves to be more than sufficient for a small to medium-sized room.  The 48 volt psu also drives the amp unit noticeably warmer than the 32 volt which will potentially shorten the left span of the amplifier.  Not recommended.

Yep, I forgot to mention the quality parts it uses, which sounds too good to be true.  I think that make it sound much better than its neck-to-neck competitor Aiyima A07, which sounds grainy especially with more complicated music. 

@erik_squires I went back to you post (pls do the same) and found no mention of "sound quality" and "1990" How is one supposed to know what you were "trying" to say big shot. BTW, it is not an accurate statement that, back in 1990, the class D with PWM has already sound comparably well with sinusoidal sound wave.

@eddnog 

I was further confirmed with the 4/20 on Kickstarter which is initially informed by @jackjohnson316 and Fosi also told me that it will hit Amazon in US on June.  I am interested in SS3602 myself too but wonder why don't you consider the 2nd SS3602 instead of V7 Vivids since SS3602 is half the price and the review is good?  How does SS3602 stack up with V7 Vivids?

@erik_squires 

The basic principle of classification still defines the efficiency, cost, and form factor of a power amplifier essentially. How the audio industry implements technology on top of and around this principle to make them perform well sonically in measurements and actual listening experience is still a ongoing progress, not a cliche understatement you were making at all. See below how the forthcoming Fosi V3 Mono (priced at $280/pair) bridges the gap nicely with the noise/distortion level of the state-of-the-art Benchmark AHB2 that costs $3,500.  

(from Amirm @ASR, April 2024)

@jasonbourne71 

Thank you for pointing me to ASR, where I found these charts that Amirm nicely made. I see that at the maximum power rating of 32 volts SMPS, the 48 volts SMPS does not have a few dB noise advantage. And the 48 volts block gives more headroom/current to serve the peak loads as well when called for. However, at my usual listening level, I don't crank the volume knob to acquire that much power, and I am craving for more clean power from LPS. Also, the op-amp rolling will potentially render sonic improvement just like tube rolling on the tube amp. We will see.

@hifinut51  There is a pre-out terminal on the back designed to bypass the amplifier's built-in power amplification stage.  SE only but, if you want xlr, go with Za3 which is supposed to similar SQ based on the review.

Although Perca ($$$10k) does not publish its load-dependent distortion level, its average THD of 0.0003% (corresponding to -110 dB) puts itself closely against the Benchmark AHB2 (0.0001% or -115 dB; $$3.5k). Perca's high efficiency lies in its inherent nature of being a class D design, whereas the AHB2 is essentially a class A/B (technically class H) design but manages to achieve 92% efficiency thanks to its 'railing modulation' technology implemented on its voltage stage. In practical terms, the AHB2 generates a lot less heat compared to regular class A/B and runs as cool as a class D. Therefore, I concur with a few folks saying that the dividing lines among classifications become less and less clear; i.e., the class A/B amplifiers can be very efficient, and vice versa, the class D can be as smooth/musical as the class A/B, even though the fundamental PWM output waveform remains unchanged.

To me, if I could afford a few grands for a high-fidelity, high-end amplifier, I would jump on the AHB2 any time. Not only is it much more affordable, but the inherently smoother sinusoidal sound wave also wins me over. Despite various types of modulation, filtering, and feedback techniques implemented on modern class D amplifiers to make them sound better, I still believe in retaining the purity of the original sound wave as the ultimate avenue to achieve ultimate sound quality.

I added a THD (%) scale, which corresponds to the dB scale, into the graph for comparing more affordable NAD C268 ($1k) and the Bel Canto S300iu ($2.5k) with the Fosi V3. As shown, the NAD C268 is comparable to the Fosi V3, but the Bel Canto is far inferior. Again, measurements do not mean everything, but the THD measure does give an indication of the sound clarity. I just provided you with the data in one place for your convenience, and it is your discretion how you interpret the data.

@eddnog Yep, thank you for the reminder​​ Ed​​​​. 11 AM EDT. I had the site open. I was told biggest discount will be limited to the first 200 participants.

(disclaimer: I am not affiliated with FOSI)

@eddnog I am a bit lenient. Could always wait until June from Amazon. I plan to use a 48v/10a psu to power them with a splitter cable, primarily to save space. Do you think will that compromise SQ in any way?

@pindac Truly amazing stuff that Audiophonics has accomplished. Built on Purifi/Hypex modules/ps, the HPA S400-ET has been highly praised by ASR based on its excellent measurements. The compact size version, LPA S400-ET, has almost identical specs and performance in a half-size chassis. Similar to the Fosi Mono, the S400-ET offers multiple gain settings, dual replaceable op-amps, XLR inputs, and quality parts inside. The upgraded binding posts in the back are attractive. The sound quality is also highly regarded, surpassing class A/B or even class A based on the limited reviews I could find. I really don’t need to repeat all the positive reviews, but without exception, let me include its performance curve in the chart for the last time before wrapping up. It is almost neck-and-neck there with the Benchmark. The Fosi Mono starts losing its pace at higher loads, but that’s understandable given its price disparity.

You know what? This beast only weighs 12 lbs, so I could easily carry it back on the next trip to France. I guess that is one of the advantages of class D.

There actually was an extensive discussion on Audiogon a year ago.

https://forum.audiogon.com/discussions/new-class-d-amplifiers/post?postid=2543479

@pindac Please forward the link once you have it. I also dived into the DIY project based on Purifi 1ET400A and Hypex NC400. I am particularly interested in the hybrid one with tube buffer made by VTV. Although its price tag is relatively pricier, marrying a tube buffer might be a rather good idea. The idea is not new and, the more I think about it, the more I feel it is doable. We could biamp a tube amp (or low-wattage class A) and these powerful class D modules, where the tube amp/class A is used for the tweeter/midrange and the class D is used for the bass driver.  What do you think?

I could not agree more, @pindac. Such a ’bybrid’ amplification could be even better on a active crossover system. A good (mini)DSP could easily lead us there, which will be more convenient than the passive crossover system. A lot of class A / tube amp or these class D module amp do not even have ’pre-out’ or ’line-out’ any more for bi-amping.

@atmasphere Thank you for the very helpful information. I recalled someone asking / commenting about the musicality and I have examined all the ASR measurements but could not find one appropriate to be an indicator.  After a brief search I was only able to identify two gears (see below) that ASR give the THD vs frequency chart. I am not sure what the thought process was behind making this chart obsolete.

The THD vs frequency chart was soon replaced by the following load-dependent THD charts by frequencies.  As you may know, the load-dependent THD is usually tested using 1kHz tone in the industry but was later criticized because it ignores the adverse effect of higher frequencies over 1kHz on the overall THD measure.  So, including Erin, the reviewers start tests using multi-tones.  As shown in the following chart, the 15kHz frequency has big jumps in THD levels, while the better class D module like Purifi demonstrates a much more controlled increase in THD at these high audible frequencies.   

The phenomenon of even-order harmonic distortion (HD) contributing to a "pleasant" or "musical" sound in tube amps, while odd-order HD in transistor designs may sound "edgier," is a well-known concept among audiophiles. However, the idea that modern class D modules can also produce a "sweetened" or "musical" sound despite lowering total harmonic distortion (THD) might seem contradictory.  Can you provide some actual test done for those class D circuits demonstrating how they accomplish what you have described in the last paragraph?

@atmasphere Just look at the harmonic spectra of the Purifi module. You’ll see a dominant second.

These are freq. spretrum FFT plots for THD and IMD from the published Purifi 1ET400A data sheet. For IMD, all spikes are beyond the audible freq. range so not a concern there. But, in the THD plot, where is the ’dominant’ 2nd HD? All odd- and even- order HD are almost inaudible.

@atmasphere  I was there initially but just like asr said "... The third harmonic hovers around -130 dB which again, is almost at the limit of what we can measure..."   It is inaudible either at that noise floor.

I am sure that, with such a low noise floor the Purifi / Hypex class D modules provides, the sound from it is transparent bar none.  I also concur with you that the resultant sound traits hinges on how the modules is implemented.  But I don't think the sound will be sweetened approaching to tube amp unless specifically with a filter or other component.  Maybe that is why VTV adds the input tube buffer.  Just my two cents.

@atmasphere So if a class D has the same distortion signature as a good tube amp, it will sound like a good tube amp simply on that account.

Agreed. But I want to emphasize, with the low 2nd HD found in these Purifi / Hypex class D, you won’t be able to have distortion signature comparable to a tube amp and possibly produce the tube-like smooth sound. Let me give you an example of sound color signatures built in the Smsl Su-8. The 2nd, 3rd and other higher order spikes in the standard setting (upper left) are insignificant (-120dB, similar to Purifi / Hypex. Users could elect the "Tube 1-3" sound signatures to experience the emulated tube sound, where you could find the 2nd-order HD boosted to -90dB in Tube 1 (light) setting and -70dB in Tube 3 (strong) setting. That tells me the 2nd order HD needs to have high enough ’dominant’ spike (relative to 3rd & higher-order) to exhibit the tube-like smooth sound.

Here is the FFT plot for a real good tube amp, McIntosh 40, which confirms with the above. But, interestingly, Mc40 actually has slightly higher non-dominant 3rd hd.

@mattw73

I’m impressed with both the Nilai Mono and Apollon St in terms of their specifications. However, I lean slightly towards the Apollon for a couple of reasons: (1) it offers 25% more reserved maximum power in both 4 and 8 ohms (despite a slight sacrifice in dB noise in 4 ohms), and (2) it has the lowest measured power on/off noise, which is inaudible. Additionally, while the Nilai Mono has the advantage of being a stable 2-ohm load amplifier, the Apollon Stereo is half the cost of a pair of Nilai Monos. If you or anyone else has experience with these two amplifiers especially how they sound, I would greatly appreciate any advice or insights you could offer. Thank you!

@ricevs

Thank you for the write up / advices. I pretty much agree with what you said there regarding the sound signature heavily depending on how the modules are implemented. The reason I put so much effort there investigaing measurements is because there are still significant variations in the performance specification among the diy or commercialized products based on the same Hypex / Purifi modules. A few notes / questions here if you don’t mind:

- I am interested in VTV Purifi with tube buffer as well but I still want to see the detailed measurements there like ASR; I won’t bet / invest 2-3 grands just for experimentation. Hopefully ASR could have a chance to evaluate the product.

- Discrete op-amp like Sparko S3602 is certainly top notch but, according to FFT freq. spectrum analysis, the TI LM4562 is no slouch either and only costs 1/6. I know you might throw in the similar comment like they will sound differently even measurement is alike BUT I wish to hear your opinion in case if you had some experience with it.

I don't believe it's necessary for components to be of poor quality to show disparities in measurement. Take the Fosi V3 and Aiyima A07 as examples. Both of them use the TI TPA 3255 module, but Fosi utilizes quality passive parts like Japanese/German capacitors and inductors. Both use the same stock op-amp (active). These disparities in the quality of passive parts are sufficient to produce audible differences in THD, as shown in ASR measurements.

Alright, it is fun discussion but counter-productive (to me) talking these technical issues phyosophically.  Call me stuborn but, before making any major purchase / investment, I would like to see positive / superior objective measurements on the particular model I intend to purchase to get the maximal bang for the buck.  I will also do my due diligence to get the users' and revierers' (I thrust) feedback and get a feel how it sounds.  I won't make my decision just based on diyers claiming that all Purifi / Hypex based modules (including power supply) measured the same.  Different implementation in the input buffer makes significant difference!  And that is true!

@eddnog  I didn’t see any results from anywhere testing the V3 Mono specifically with a 5A power supply versus a 10A, making it even more difficult to say.

Erin just shows the test result today. 'Single' means single amp fed by 48v, 5A psu and 'Double' means double amps fed by 48v, 10A psu.  Having individual 48v, 5A psu feeding to the v3 mono does have 10dB advantage in THD. Interesting. 

@eddnog  Same here. I made a last-minute switch to the separates based on my instinct. I recalled that the overall package deals were identical during the pledge.

It's funny how you can find cables labeled as 6N, or even 7N/8N OCC or OFC all over Amazon and AliExpress, and they're incredibly cheap. It's amazing to see the quality control and production efficiency of the plants in China these days???  I am a bit being sarcastic.

@jl35 They are pretty much the same based on my recollection after seeing several reviews and demos. I decided on v3 primarily because of its pre-out feature, so I could do bi-amping later on with other class A amp.

@atmasphere Ralph, great post and thank you for sharing. I also enjoy your afternoon talk here, which helps me understand some of the intricacies of class D design.

@jl35 glad can of some help. enjoy listening.