LM3886 chip amp vs Purifi/Niali class D

That is sadly mostly the case. Small things are discussed for years and years, but you seldom see a subjective evaluation on how an amp sounds. And
people say "I will try this and that", but seldom comes back and report how it did work.

If you mean measurements when you say evidence, is not so hard to find. Find Amir on YT, his followers discuss SINAD to no end. But no measurements
we know so far, will show how an amp sounds.

I have often wondered how an amp will not be able show a nice square wave even at 1 KHz, when it is supposed to have only 0.000X % distortion.

I once build an SMPS powered version of a Hypex NC500 integrated stereo unit. It had so much switch noise layered on top of a 1 KHz square wave, that
I almost couldn´t recognize the square wave itself. In reality amps have distortion in the 1 figure area, or we wouldn´t be able to hear it. We just don´t know
how to measure it. If amps really had a distortion of 0.000X % they would all sound the same.
 
I never heard the sounds of the class Ds you refer but I jongled several years with LM3886. The sound of tda7498e board 2x60w 8 ohm $12 from Ali once converted into ABD mod to erase the dead time distortion, it excelled by far the LM3886 which cannot reproduce complex vocals as the requiem of Mozart or Verdi.
 
Gee is anybody reading this thread any the wiser or clear on which is best. So much opion and little evidence. Did i mention hearsay.
Great question. I think after all this... If you like class D... buy a class D. If you don't, buy something else, either an class AB chip amp or class A or valve or 1 watt early pass amp or these new schiit rekkr or whatever.

I think at the end of the day you might have to listen and find out what you like which is a bit tricky as you have to buy them or somehow try them.

I think that at some point, there is a diminishing return where it might not sound all that different (to me). I was looking to build a diy chip amp because its fun and there is a sense of achievement. Potentially with a longer lifespan as I could resolder and replace parts.

The AlphaAudio test seems to like the Niali and there are some great reviews of the purifi, but then other people also don't seem to like them; too harsh, too clean, no life or whatever you pick up when you read about them.

My original question arose because the cost of class D seems to have come down a lot and for around 1200 USD-Euro whatever you can get a purifi and niali with tons of watts. This seems to be close in cost to some diy builds (if you include bells and whistles and once you factor in the case, parts etc).

I'm leaning towards a chip amp for the satisfaction, the process, minor blagging rights and potential reparability. Chip amps seem pretty stable in terms of development but these class D amps seem to improve/change each year. Perhaps next year the purifi sounds even better and then I might buy that version. Also I'm not sure I need a bazillion watts to drive my 8 ohm speakers while I listen to music throughout the day in a small office - its just nice to have them for some reason.

Looking at Audio Science Review, the latest topping variant (which I think is a chip amp) seems to measure super clean, but then I read its doesn't make sense, lots of warranty issues, not enough watts etc. The top 20 in the ASR amp list seem to be covered by class D amps, I think the neurochrome modulus-286 is also up there.

Long story short, I think my original question was answered... it depends.. I just don't have the luxury to listen to them all in a room and maybe if I did, I might not hear the difference....
 
I have the Modulus - 68 and I simply love it! I have had lots of amps but the M86 just sounds right and I don't care about describing sound. So if you have speakers that are driven to sufficient levels and don't present a crazy load to it - that is all you need.

My experience with class D - I had the crazy Mivera IcePower 1200AS2, I came under the spell and bought the one with the silver internal wiring, fancy connectors and marvellous case, I am one of those convinced that this makes a difference. Yet the amps was simply awful - flat, two dimensional, thin, soulless, anything that Class D haters say about Class D. Dis as ter. BUT. The new Purifi is a different beast - sounds OK. If your speakers are dipping low or are low sensitivity - you will appreciate hearing them driven OK for the first time, that makes a huge difference but that is just correcting something that was wrong the whole time. The Purifi is decent enough that I could live with it. At Munich I heard the new 1ET9040 and that one, together with the SPK16 was just magical, not sure if it was the amp or the speaker or the combination but result was good, soooo good.

If I were you - I would make a safe bet and buy the Moduls 686. If that is as good as the 86 and is so powerful - you won't ever need anything else 🙂 If you are adventurous - wait for the 1ET9040 and try to build it PROPERLY - small things make a huge difference there, shown many times. It is an easy resale and you won't lose much if you don't like it.

Hope that helps.
 
If I were you - I would make a safe bet and buy the Moduls 686
Thanks for the insights' good to hear. I'm not sure I need so many watts. My speakers are 8ohms. If I need more power I think I would try a class D. A modulus 686 or 286 with all the extras gets expensive. For 8ohm I think you can use a bigger transformer at the cost of clipping of 4ohm speakers and make an 8 ohm mod-86.
 
Mmmm, to put it another way - I have been in the hobby for 35+ years. If I had started with the modulus 686 I would have saved a six figure sum, not starting with one... Sadly. Yes, the 686 is not cheap but if I could go back I would just buy it or build it and be sure that amplification is covered. Comes from a man who has 7-8 amps at the moment, gone through probably 20, looking at the Ayre MX-R monos yet knowing the 686 is what I really want 🙂 . I have had countless "this is it" attempts at amplification, all kinds of topologies, all kinds of fancy parts, 6c33c, GM70s, you name it, I just could not believe my ears when the 86 was made. Both my Mola Mola Makua and Resolution Cantata 3.0 have internal volume control which helps me a lot(a better preamps will do wonders, I know, just like the simplicity of it at the moment) and they are both balanced.

My Aequo Stillas are OK with the Modulus 86 bit that is rare, the Joachim Gerhard Surveyor sucks the power out of the 86 pretty fast even though the Surveyor was supposed to be easily driven by tubes... The 686 is one of the real bargains in audio, gives you the chance to have a really good amp and decide how much you want to spend on the enclosure but most importantly - it will help you get the best from your speakers and trust me, that alone is a fantastic money saver 🙂
 
I'm thinking of a new DIY build and like the look of the neurochrome modulus-86 and the reward of building it.

However for a similar price and minor assembly (less interesting to me) the Hypex Niali 500 as well as purifi have good reviews and performance.

Is this just a class AB vs Class D taste question, would I even hear a difference at this level and is there a shelf life/longevity benefit over AB vs D.
I've been following the thread and I'm a little surprised no one has mentioned this.
The Modulus series amplifiers aren't "just" LM3886 based, they are composite amplifiers. That means they start with a high-quality op-amp and put the output device, the LM3886, in its feedback loop so that the power amp closely follows the distortion profile of the op-amp, which lowers distortion by a few orders of magnitude. Presuming the LM3886 by itself is "audible" (at its 0.1 or 0.01 percent THD), a composite amp using it is bound to sound different (and we would hope better) with more 0's past the decimal in the THD spec.

Due to high loop gains and phase shifts of both devices, a composite amp is difficult to keep from oscillating, so it's much better to follow a known working design using the precise parts values given than try to design it yourself. I really only know of two composite designs using the LM3886, the various "myref" designs (several threads here on diyaudio), and the Neurocrome Modulus designs. Neurochrome also has an inexpensive "LM3886 Done Right" PCB design using only the LM3886, which is similar to many other LM3886 kits, and has the the higher distortion of the LM3886 alone. I've been tempted to get one each of the "done right" and Modulus 86 just to hear the difference myself.
 
Traditionally amplifier designers try to minimize non-linear distortion (usually harmonic distortion, sometimes two-tone intermodulation or multitone intermodulation distortion)
Having zero THD is equivalent to zero IMD, and that would be the holy grail of an amplifier. The problem with THD is it's a single number, an amalgam of harmonic measurements so it doesn't indicate the "character" of the sound. Especially, a Class A amp, a Class AB amp and a Class D amp could have the same THD figure, yet each sound quite different (unless the figure is zero).
without caring much about frequency response and loudspeaker designers try to optimize frequency response without caring much about non-linear distortion.
Frequency response is easy to get "optimized" in amplifiers (thus no reason to think about it), but hard to get optimized in speakers.

Loudspeaker designers also care about non-linear distortion, but it's harder to control, mainly by driver choice. You have better control over it if you're designing the drivers, but then it's still hard.
 
There are many amplifiers with frequency responses that are not flat to within +/- 0.2 dB, especially when loaded by a loudspeaker. Meanwhile, unintended frequency response or level errors of +/- 0.2 dB are enough to invalidate double-blind ABX tests that are meant to test for something else.

The distortion of just about any amplifier is so many decades below the distortion of just about any loudspeaker, that I don't believe for a second that it is of any practical significance, even if the harmonic content is different.
 
Shure V mark 3 was a revered pickup cartridge in 80's, used mainly in radio stations, recording studios etc. Audio magazine have measured it, 2.5% 2nd harmonic 0.5% 3rd harmonics and 3%IM. Only mythomen can pretend to here distortions down to 0.01%THD.
 
@Forane12

I'm also looking to build another DIY kit as my LicePower 1200as2 has bit the dust..lol Still trying to find a schematic on it, but I guess LicePower has it locked up in a vault! Appears that the amp is going into a protection mode/standby, then comes ON for a brief moment, then continues switching. I don't think it's the PS caps or thermal issues, perhaps a bad transistor? Who knows. Bad thing with these amps is that they had a 10% failure rate 🙁 Wish I would have know that before!

I have been looking at the Purifi, Hypex Nilai amps, but now maybe considering one of the Neurochrome amps. You mention the 686, I take it by your experience that this amp build would be better than the others here I mentioned? You also mention the upcoming Purifi 1ET9040 build, would this be the flagship over the 686? Seems like they were talked about being released soon-ish, and that was last year around May 2022? Thanks!
 
The cost of the DIY Niali is attractive , its 1200 Euro (excl VAT) 250 W but... class D.

For a modulus 286 with all the bells and whistles (soft start and speaker protection and v. nice case) I get to around 2000 USD (125W into 4 ohm 65 W into 8 ohm).

A modulus 86 would be cheaper around 1500 USD (60W into 8ohm with larger power supply but 55w at 4ohm) vs 250W Nilai. If you skip the soft start and speaker protection its around 1200 USD.

A 686 needs a bigger case and is around 400 USD more.

I'm not saying the neurochrome isn't worth it - its just its almost double the price. A modulus 86 (with and without soft start and speaker protection) comes close to the Niali price but for fewer watts. Do you need 250 W of Class D power - I don't know.

The build quality of the neurochrome seems very high, and probably more serviceable than the Niali. The modulus are also tried and tested for a number of years now, Niali seems to update each year. In theory, you could also build a Modulus 86 and upgrade later to a 286 later.

If you get the modulus 86 stereo kit and you really get to build it yourself, which is the point of DIY right?
 
Source quality is very important. With a good source, modern class D will sound really good. I tried a few class D amps and to me, they sound like wire with a gain. But the source would need to be really good, as I said. Otherwise, the top end could sound a bit unpleasant.

There's a Peachtree Audio GaN with a S/PDIF coax input, i.e. a true "digital" amp that I also managed to listen to. This was a very invigorating amp, an exceptional experience, especially when I used my Holo May DAC coax out. But, it did not have that spatial presentation... that I like so much. And no, it is not my speakers or the room acoustics to blame because I get a beautiful walk-through soundscape with Aleph J. I personally value that aspect of HiFi more than anything else. Of course, burning 200W at all times for a measly 20W of useful power could be deemed wrong at so many levels... but, them watts do sound very nice and are more than I'll ever need.

By the way, I also got very disappointed with SMPSs.... building a linear power supply for a class D amp is something my back and old age disagree strongly.....nevertheless, more on SMPS performance in general, in high-end audio can be found here:
https://www.diyaudio.com/community/threads/post-you-smps-noise-spectrum-measurements.394518/