Purifi Class D: Junk?

Thanks Mike. I believe the Weiss opamp is pin compatible with the Pro series opamps from Sparkos and Sonic Imagery, but not their dual discrete opamps. 
I want to give my amps a couple more weeks of burnin before I decide how to proceed with the Weiss opamps. 

I think it is perfectly reasonable to report that you weren't happy with the particular VTV model you have, but to lump all products that use the Purifi module in the same bucket is pretty rediculous. I haven't personally heard any products that use the Purifi module yet, but I know from my limited experience listening to products that use other class D modules that there is a lot of differences in the way these amps sound depending on how they are designed and built. 

I spent some time today switching between my Purifi amps (using Neurochrome input buffer), Neurochrome Mod-286 amps, and Parasound JC5.

First, I should say that my ears aren't as discriminating as they used to be. I've suffered with Tinitus for many years, and my high-frequency hearing is not what it used to be. So take these observations with that in mind.

I used the same interconnects for all three amps, but I had to use longer speaker cables with the JC-5 since it's a stereo amp. The cables were exactly the same construction, just 2.5 meters instead of 1 meter. Also, I used a different (although similar) power cord with the JC-5 since I needed a longer one.

I set the levels using a pink noise -10db recording with a sound pressure meter with C weighting measuring 64db. For most of the recordings I listened to, this would be about my average listening level. A couple of the recordings were a bit lower volume than I'd normally listen.

The songs I listened to were Sophie Zelmani - Why, Jennifer Warnes - Way Down Deep, Shirley Horn - Beautiful Love, Dominique Fils-Aime - Birds, Ray LaMontagne - This Love Is Over.

I turned off my subwoofers for these comparisons since they have their own amps and I didn't want their contribution to influence my impressions. My main speakers (GR Research NX-Oticas) don't play very low without the subs, so I can't really assess the deep bass performance of the amps.

All three amps have a fairly similar presentation, although the JC-5 was slightly more laid back than the other two. The overall differences between the amps were not real obvious.

The JC-5 is slightly warmer sounding with excellent dynamics and smooth midrange. The highs (to the extent I can still hear them) sound clean and natural. Compared to the other amps, the biggest short-coming is a somewhat narrower sound stage. I suspect this is because the other amps are monoblocks.

The Modulus-286 amps were the weakest of the three. They still sounded quite good, but had a bit of honkiness (if that's the correct term) to the upper bass and some unpleasant sibilance on female vocals. At higher volume levels, I've noticed these amps get a bit more congested sounding, probably due to their more limited power, but at the volume that I was playing today, they sounded fine.

The Purifi amps were just a tiny bit rougher sounding in the upper midrange compared to the JC-5, but overall cleaner than the Modulus-286. They also had a touch more energy in the bass compared to the other two amps, most obvious on the Jennifer Warnes song.

At this point, I'd be perfectly happy with either the JC-5 or the Purifi monoblocks, but I'm hoping I can take the Purifi amps to the next level with the new input buffers I just got. I'm waiting on a couple of connectors before I can install them (hopefully next weekend). The Purifi amps are also a lot more efficient which is particularly nice during warmer months.

To be fair, Purifi does not sell their module with this input buffer. In fact, they don't sell their products to the public at all except to the DIY community. The amp you bought includes a buffer that wasn't even designed with the Purifi module in mind. 

Purifi expects that their OEM customers will add value to their module by designing proprietary power supplies, input buffers, and other logic and circuitry. The NAD M33 or the LKV Veros are more the kind of products they envision, or at least the VTV with their tube input buffer.

Amplifiers built with the Purifi module are only going to be as good as the sum of its parts, like any other electronic product. If you build the amp with an entry-level power supply and the cheapest possible input buffer, it's not going to be an good indicator of what the module is capable of. 

@douglas_schroeder - I don't think you read this thread very carefully. This isn't about DIY (although I don't personally share your skepticism about DIY). The OP purchased a commercial product using the Purifi module, a basic SMPS and cheap input buffer, and is blaming the Purifi module for its disappointing sound quality instead of the surrounding parts and overall design. 

That's like saying all DACs that use the ESS9038 are crap because a $79 AliExpress special sounded mediocre. 

The power supply has at least as much influence on the sound as does the input buffer, so "all that is left" is not just the Purifi module.

Further, the Hypex input buffer was never intended as a finished product. It is about as simple a circuit as can be used to provide impedance matching and gain. While the op amps used (LM4562) are decent, they are designed to be operated at a higher voltage than is used on the Hypex buffer to get optimal performance. The 7812/7912 voltage regulators used on this board are decades old and far from state-of-the-art when it comes to noise and response speed, so there is a lot of room for improvement. And there is no frequency limiting and EMI/RFI filtering circuits, so there is lots of opportunity for environmental electrical noise to cause problems. 

And, of course, there are improvements that can be had from wiring, connectors, layout, etc, but these will likely be more subtle compared to the power supply and input buffer design.

Bottom line, this amp as currently built is about the worst case scenario for the Purifi module. You can only go up from here. 

That was the least of the issues I pointed out. But oh well..

I had a pair of Nord Hypex NC500 monoblocks a couple years ago and was kind of disappointed with the sound  quality. These amps came with input buffer boards that supported the Sparkos and Sonic Imagery discrete op amps, and I tried them both. I preferred the warmer, smoother sound of the Sparkos in the system I had then, but not by much. The Sonic Imagery opamps offered noticeably more detail, but were more sterile sounding.

These amps replaced a Krell FPB-300 which had died for the second time and I decided it was time to sell it for parts (which I now kind of regret), and they really weren't a good substitute. Granted they cost WAY less and were a lot easier to deal with (both weight and power consumption) but they didn't work for me. 

I ended up selling them and replacing them with a Parasound JC-5 which I was much happier with. That was close to 2 years ago. I have since completely changed my system (although I still have the JC5) and have really gotten into DIY. After building a pair of FirstWatt F5 turbo monoblocks and a pair of Neurochrome chip-amp monoblocks, I decided to give class D another try. 

I recently completed a pair of Purifi-based monoblocks. But instead of using the typical Hypex SMPS, I built them with a large linear power supply, similar to what you'd expect to find in a high-end class AB amp (1500VA shielded transformer with ~200,000uF of Nichicon filter caps). As a first round, I decided to use the Neurochrome input buffers which are a well-engineered op-amp design.

I plan to experiment with other designs including discrete op-amps and fully discrete in the future, but with this relatively simple and cost effective buffer, these amps sound quite nice. I'm sure there is some "pride of my own build" bias  influencing my perspective, but I have to say that I was not that impressed with my FirstWatt F5 Turbo build. The Purifi amps are better in just about every way (more dynamic, much more detail, quieter background, sweeter midrange, better soundstage, WAY less heat). 

You can see a couple of photos on my system page. https://systems.audiogon.com/systems/8726

I'm thinking about trying an input buffer using the Weiss OP2-BP op amps next. I've heard good things about them, and it would be interested to compare them to a high quality conventional op-amp implementation. 

@djones51 - if you don't trust your own ears and your ability to overcome sighted bias, that's fine, but telling everyone else that they are wrong is pretty arrogant. 

In my past experience with the Nord Hypex NC500 monoblocks using the Sparkos and Sonic Imagery opamps, the differences were pretty obvious. Once both were well burned in, I switched back and forth a few times and had a couple of friends over to listen as well (who didn't know one from the other). The Sparkos opamps were smoother and slightly warmer sounding, but clearly lacked detail relative to the Sonic Imagery parts. 

So far, I have only heard my Purifi amps with the Neurochrome buffer board which uses integrated op-amp chips (not discrete). There are pros and cons to both (IC vs discrete), and this is probably not the proper thread to discuss those. But I would like to hear how my amp will sound with different buffer designs and will eventually give one or more a try. 

I never said I was immune to sighted bias. But there are often cases where the sound differences easily overcome this bias. In the case I mentioned, the differences were quite easy to hear by several listeners. 

@djones51 - are you in the audio business? If not, why do you bother to listen to lots of different amps if they all sound the same to you?

@speakermaster  - what do you consider a "good high-current AB type amplifier"?

If @madavid0  is using the VTV buffer boards with Weiss opamps, these have Sparkos discrete regulators for the low voltage power, and these are independent for each channel. 

It's possible, though, that the SMPS supply is configured for regulated Vaux output which could certainly cause problems. 

They are basically compatible. The NC500 has a clip output which the Purifi board does not. The NC500 recommended op amp supplies are +/-15v where as the Purifi module recommends +/- 12V, but the Purifi max rating for this supply is +/- 20v, so you should be ok swapping in the Purifi module in an amp designed for the NC500. 

I have to say that I am enjoying the sound from my DIY Purifi amps a lot more than the Nord NC500 amps I had last year. I don't know if this is because of my use of a large linear power supply in my DIY amps, the difference between the Purifi module and the Hypex module. the difference in input buffers, or other changes in my system. But these new amps are much more dynamic and engaging. 

I don't have the experience or knowledge that atmosphere or audio2design have, but based on my 45+ years experience as an audio enthusiast and tinkerer, a good power supply makes a huge difference in the sound quality of all audio electronics. I don't doubt that it's possible to achieve this with an SMPS. But either way, I think many (most?) class D amplifier companies are not making the effort and/or are not willing to increase their parts cost enough to deliver close to the full potential of what class D can offer.

That can be caused by an offset problem at the input of the comparitor- if its something like that I would regard it as a malfunction.

Seems unlikely that this would be the case in both channels (assuming that's the case). Since this is a stereo amp, perhaps an issue with the power supply?

I am not noticing anything particularly egregious in the midrange in my amps, but I want to spend some quality listing time comparing to my other amps now that I've got a couple hundred hours on the Purifis.

It's been a few weeks since I listened to my Neurochrome Mod-286 amps, and a few months since I used my Parasound JC-5. I am hoping to get some time this weekend to do some critical listening and comparisons, and will report back after that. 

I ordered a pair of VTV's tube input buffers with Weiss op amps to compare with the Neurochrome buffers. These should arrive later this week, but will require some rewiring since the connectors are not compatible. I may have to order some connectors if I can't find any that fit in my stash, so it may be a couple weeks before I can try these out.