One of my first exposures to class D was the Tact Millennium amp released in the late 90s. This $10,000 integrated amp had technology that had never done before. It received pcm digital input signals and messed with it and then using software changed the signal to PWM (essentially class D) and its filter on the output used some seriously good foil air core coils and copper caps to ground. To control volume it would lower or raise the power supply voltage of the rails to change the volume so no loss of bits like a low resolution digital volume control. I thought this thing would change the world.....it did not. Apparently it has no feedback so was picky with what speakers it liked and had software issues, etc. But the concept was intriguing.
When I thought about class D it seemed to me a lot like class A. Class A always has enough current running through its output stages so it never turns off. A Class D amp switches at 500K or so and to me that seemed like almost always on. So I have always thought class D had lots of potential. However, when Spectron, Nuforce, Tripath, etc started to make their amps I noticed that the input stages were always op amps.....the parts not so great, the output filtering used ferrite core inductors and ordinary caps. Most of this is still true today. And of course, we now know that dead time, feedback, etc. all change the sound.
The latest Class D modules, including the Ice power modules, the Hypex Ncore modules, the Pascal modules, Anaview modules, and the Purifi modules are all designed by people who do not listen to every part and think about every single thing. They are not thinking of absolute audio purity when they designed the modules. Most audiophile companies that use these modules do not mod them in any way. They just add their own input stage (or give you an option for input stages) on the input and stick it in a box. Then they give you an option for better output connectors, wires or whatever. No thinking outside the box.
Class A and A/B designers tend to be more tweaky.....they listen to more things and usually have a long track record of constantly evolving their products. You do have some companies lke Nuprime and Cherry that make their own class D circuits but these companies are still not thinking outside the box. The Cherry circuits are very simple and pure and even though I have not heard them I bet they are really great for a stock thang.
Last year (2019) I made and sold a dozen amps based on the lastest IcePower IceEdge technology modules (1200AS1). However, to get the best out of them I had to mod the module like crazy. I eliminated several things from the signal path it did not need, changed a bunch of parts, bypassed the power supply with modded Wima caps.....undamped some parts, damped some parts, used the most tweaky input jacks, output connector system (my own invented binding post bypass system), tweaky fuse system, damped chassis, copper and thick aluminum plates underneath the modules....etc. to get them to really sing. After all the mods (which were done over several months) the amps were really really good. Because of health problems I was not able to continue to manufacture them but still wanted to make my services available.....so. modding light weight Purifi or NC1200 based amps made perfect sense. Something I can lift!
I had always wanted to try air core coils on the output of a class D amp so after I had done everything I could think of on the IceEdge module I decided to try a pair of air core foil coils on one channel of a stereo module....well, the amp did not like it and it fried it.......so that experiment was a failure.....but I knew that the output stage probably had dual feedback around both coils so it would be sensitive to whatever coils you used. Since there is no schematic there is no way to develop a way to do this. However, I knew the Ncore/Purifi amps all used single coil outputs and had the entire feedback after the coil.....so I figured that type of amp could tolerate any kind of coil. A friend was no longer using his NC 400 monos that I had modded for him years ago so he gave them to me to play with. It worked fine with every kind of coil I could try and every coil sounded quite different. The stock coil had good detail but sounded grainy and boring.
About coils.......It is a commonly known fact that you would never ever use a ferrite core coil in a crosssover on the midrange or tweeter of a speaker......cause....they don't sound good. Every class D amp (except maybe the $16K-$38K Merrill amps) made after 2000 uses ferrite core coils to filter the 500K switching noise. So why would you put a distortion producing thing in the signal path? Because ferrite core coils are smaller, cheaper, have lower resistance for their size and have less radiated RF. If you designed your class D amp using air core coils then your board would have to be bigger and you might have to deal with more rf which might not make your amp pass regulations...so you might would have to filter the output in other ways.....or add shielding. So, it looks like practically no one has ever tried air core coils and listened to them in a serious class D amp? Every coil I tried sounded quite different just like you would expect. My fav air core coil has way more transparency, speed, bigger sound, more air......and is just more real than the stock coil in the Purifi.....which is the same coil used in the NC400 and NC500 and a slighly larger version of the exact same thing is used in the NC1200 module...... so all these amps are limited by the last thing the signal goes through......all these years we have been listening to a veiled, compressed, 2 dimentional, slow, grainy part. And this is true for every class D amp YOU have ever heard ......unless you have heard the Merrill amps (if indeed they have an air core coil). All coils sound different. Even if you are using an air core coil....are you using the right one? You have to listen and tweak. This is what I have done.
One of my mantras is "Everything makes a difference". Every part, every wire, every fuse, every connection....everything distorts the sound. There is no such thing as a perfect anything.....except spirit.......So the first thing to do when trying to improve something is......remove all unnessary parts.......so purify the Purifi.....he he. After removing all the not needed parts then you replace parts with better parts, modifiy things to improve them, orient parts correctly, damp and shield to taste.
So, why mod amps using the Purufi module and not the NC500.....several reasons.
1. It is suppose to sound slightly better than the NC500 and costs about the same.
2. It has plenty of room where the output coil is, to substitute better coils.
3. It has multiple parallel caps on the output and that gives more room if you stack them, like I do.
4. It idles at less current so that makes using any kind of coil OK. The coil can cook, if the idle current is high. The coil acts like a heatsink to the output transistors....and since heat likes to rise...then the coil can get hotter than the heatsink plate underneath the board.
The NC1200 has plenty of room for better coils and caps but the idle current is very high.....but I believe I can add a heatsink directly above the output transistors so I am pretty sure I can make it work with any coil. Two sources have said the NC1200 is still better than the Purifi....but this may be execution related. You see, an NC1200 based amp has less connections to go through to get the power supply to it and the signal out. My mods to the Purifi simply the circuit path even more than the NC1200. So, if you need even more power than the Purifi.....then the modded NC1200 is the ticket.
So, why mod the Purifi over IceEdge and Pascal? The IceEdge uses all built in op amps on the input and also the output coils cannot be changed. Same with the Pascal. Plus the Purifi is easy to work on.....relatively speaking.
So, how good is this modified Purifi? Pretty dang good!!!!!!! Please click on the right link below to a review of the Purifi evaluator board amp....the reviewer said it sounded identical to $40K Constellation mono blocks ...... what I am doing is way, way better than that evaluator board amp. You will see comparisons with other amps soon. But if you think a stock VTV, Nord, Appolon, etc. based Ncore or Purifi sounds good.....well, those amps modded will sound a whole nuther world better. I have been modding and manufacturing high end audio for over 40 years. I do a lot of things that no one has ever thought of and that no one would do as it is so time consuming and tweaky (for instance, binding post bypass system, modified capacitors, hand made wires, modified regulators and modified op amps, paralleled surface mount resistors, marking caps for polarity, removing connectors and hardwiring things together, etc.).
Are the EVS modded Purifi and NC1200 amps reliable? I don't do anything to the basic circuit in the Purifi/Nc1200 modules. All the built in limiters and safety features are still intact. The Purifi modules and Hypex power supplies are incredibly reliable and will continue that way once modded. Of course, there is no manufacturers warrantee when you have a modded amp. If there is ever a problem with the amp I can service them.....for cost of parts plus small labor fee. Class D amps are getting more and more reliable. I cannot remember anyone mentioning any problem or reliability claims on class D amps in the last 5 years.
Below are some links to reviews of the stock evaluator amps that Purifi sent out. What I am doing is way, way beyond this.
I will have more info soon and also info about a tour of one of our modified amplifiers.....so more people can hear what is possible with class D.....using the very latest low distortion modules and super tweaking.
