Mintzar

Paul's new power supply will likely make a difference on the LIO-8. It all has to do with bandwidth. The LIO-8's internal regulators are pretty good, good enough to filter out the majority of the sub 1-mhz noise generated in the SMPS. But digital is actually impacted by noise of much higher frequencies than the khz range that most filters operate within. Actually upwards of the GHZ range. Paul's (and my) newer power supplies are specifically designed for digital circuits and have much greater bandwidth (PH's up to 1ghz and mine closer to 1.5ghz gain bandwidth at the error amplifier). Paul will tell you that it wasn't until his latest power supply that he engineered it specifically for digital. I don't use PH products in my designs, however. Most linear power supplies just don't have anywhere close to that kind of bandwidth and that's why they don't impact the LIO-8 -- which is already good at filtering out those frequencies. Batteries impact the sound of the LIO-8 because each battery will have a slightly different chemical reaction. That reaction has HF noise byproducts, which are beyond the bandwidth of the LIO-8's regulators. It's why each battery sounded different -- different noise would translate to amplitude distortion in different frequency regions. My experience is that the LIO-8 can be dramatically improved with a proper power supply, but when you consider the price of the LIO-8 + power supply there are other offerings that will sound better for +/- the same money IMO. Sum of the parts, sum of the goals.

Power supply benefits on the computer have very little to do with what's injected back into the wall. If that were the case you would hear the same benefit unplugging the device from your system. The power supply filters out ultra high frequency noise which causes amplitude distortion on the ones and zeros of a digital signal. The removal of amplitude distortion is the reason for the significant improvement of the sound. I do use a transformer with a faraday shield, but that has very little to do with my design. The true backbone is how I've grounded the circuit, parts selection, and the various technologies implemented (AC filtering, regulation, banks and cap values, bypass, magnetic grounding, etc). The power supply is 14x8x3h and weighs about 17lbs. So if you're expecting something dinky this ain't it.

At 44.1KHz a D/A converter reads a new digital word every 22.6 microseconds (.0000226 seconds). If you 8X over sample, the same converter reads a new digital word every 2.8 microseconds (.0000028 seconds). If you are up sampling to 192KHz and then over sampling 16X (very common), the converter reads a new digital word every 300 nanoseconds (.0000003 seconds). In addition to the faster clock speeds required for up and over sampling, a 32-bit D/A converter would have to read twice the number of bits as a 16-bit D/A converter within the same period of time. So you're talking anywhere from 70-450x the clock precision required in a Delta Sigma DAC compared to Multi-bit ladders. Multi-bit DACs have 16 latches and 16 data paths, one per bit. Delta sigma type DACs only have a single latch and a single data path. In Delta Sigma DACs there are all kinds of filters and algorithmic estimations that occur because the technology doesn't really exist to process data that quickly. There was a white paper written by Vishay and TI about high-resolution digital playback. The part datasheets SAY 24 or 32bit, but there actually isn't enough tolerance in the parts to achieve that resolution in real world scenarios. Even with .0005% tolerance resistors from Vishay I only was able to achieve 20bits accurately -- and that's in a discrete DAC. One of the reasons why a good 16/44.1 NOS DAC sounds arguably better is because it is a true 16 bit reproduction. 99% of Delta sigma DACs don't achieve more than 12-14bit resolution in real world scenarios. It's one of the reasons I always preach that high rez is impossible, it's also why a low-rez low-distortion file will always sound more organic than a high-rez file. In a high-distortion system the high rez file sounds better because the pieces of the pie that are distorted are proportionately smaller than the pieces of the pie in the low-rez. But you have to have really low distortion to begin to hear this.

Both Delta Sigma and Ladder DACs have resistors. In the case of ladder DACs you have 16 resistors that all have a voltage applied in order to determine the value of each bit. In Delta sigma you have a voltage applied to a single resistor multiple times. Timing is an issue with delta sigma because any timing error in application of voltage creates distortion. Extrapolated: Applying the wrong voltage to the wrong bit. But this still goes back to the voltage that's actually being applied. If proper voltage is applied the external master clock is less important because the problem isn't as much the timing, it's the cleanliness of the voltage being applied. Stock clocks have poor power, but their jitter is usually quite low (as I said, jitter is almost always WELL below the audible spectrum). The better clocks have a better power supply, which is why they tend to improve the sound in similar ways to power supply upgrades.

Hey George, Take this with a grain of salt since I've spent the last three years developing and testing power supplies and technologies specifically for digital audio and now sell them. But my experience with a power supply upgrade to the TCIT translated to: 1. More real resolution (not top-end tilting that gives the impression of clarity) 2. Far better harmonic content, tones and textures are more layered and intelligible 3. Better musical flow. Dynamics become both more intense, but less forward. 4. Substantially less fatiguing. The upgraded TCIT sounded better than the stock Metric Halo LIO-8. Bare in mind that I had a high end power supply on my mac mini with my own internal power filter inside the mac mini chassis. I was also not running a cheap supply on the TCIT. So YMMV depending on the path you choose. PM me and I'll give you some names of a few people who have power supplies on their TCIT or just ordered them. Don't get me wrong, my digital amp uses a high-end clock for reclocking each stage up to the output of the amp -- so I obviously think clocking is important, but 90% of the performance comes from the power supplies IME. Does that answer your question?

Clean sound is easy. Texture, organic tones, emotional content, and proper harmonic content -- or lack there of -- comes from the power supply. It has nothing to do with timing or jitter -- both of which are substantially below the audible spectrum. Master clocks do not fix timing, but allow the power supply voltage to be more accurately applied to the resistors in the DAC chip. More accurate voltage application translates to less amplitude distortion on the ones and zeros. This means more organic character, better harmonic content, and less of that hash and grain. The likely reason that the BP CD Player sounds better is because they aren't using a cheap switchmode power supply like the TCIT.

Michael, I was told over the weekend that the Kalos wasn't in the room. Unfortunate. Shoot me an email, I'll work something out with you. The only reason I shy away from demos is the only two returns I've gotten were items I sent off for free demos, which were tampered with and basically destroyed. If people don't have skin in the game then they don't take care of the equipment. Hence the reason for 30-day trial. So I'm doing my best to both help people make a safe investment and protect myself. All I know about tubes is the schematics, what I've learned from mentors, and listening to $150k tube amps from TRL, Robert Koda, and Wavac (among lower-end ones as well). So I'm not saying they can't sound great, there are many reasons to love tubes. I'm saying they can't have low-distortion. The heat alone causes huge amounts of noise and distortion in the circuits. Hot regulators increase resistance, and to shunt noise away from the circuit you need less resistance (filters work by creating a path of lower resistance above a certain frequency for noise to travel). So increased heat means less noise is carried away from the circuit. This is one of the many reasons why tubes almost never have a S/N ratio above 110dB. Before you say anything, S/N ratio is not a sign of low distortion either. I want someone to create a balanced, low-voltage power supply for a tube amp and have it run coolly. Please and thank you. You can pay me in peanut butter m&ms for the idea. Thanks Ted and Talos for your comments. I am not intentionally advertising, just answering questions and comments directed at me or something I've said. I don't want anyone to buy from me unless they are 100% comfortable doing so, I'm not trying to mislead anyone or convince them into purchasing just because it's a 'component of the day'. If I am trying to mislead and trick people out of their money rather than working to bless and serve people with the endeavor to help them have better sound, not only am I doing this for the wrong reasons, but my business will die practically over night. It absolutely has to be about serving and helping, not about making money. I think it's a good thing to be able to ask the manufacturer of a particular item some questions. That way you can make educated decisions about a particular item, whether based on opinion or based on trust for the manufacturer. If answering questions is advertising then I'll talk to Chris about buying ad space so the dogs stop biting my ankles. I have a crappy guard dog so he won't be much of a fight. A dog named Ducky can't be all that intimidating anyway. I'm not trying to be sneaky and get free advertising, I'm just trying to learn like everyone else.

People will either trust my logic or they won't. I am not trying to mislead anyone or make false claims. All of my technology is based on real, proven technology. My intention isn't to mask things in mystery or marketing in any way. I want to teach people where my philosophies come from. Low distortion that is mainly born from a great power supply, part choice, and simple signal path. If someone chooses to believe me and try something I've built then the only truly important element is whether or not they enjoy the end result. If they don't, then of course there's a generous return policy. I approach audio without the objective to cut corners in order to increase profits. If I could increase profit I would rather decrease price instead and pass the savings on. I work in television (development) and specialize in syndication for high cash flow commercial real estate conversions. Meaning I specialize in working WITH people and being honest so everyone wins, not working against them to make a quick buck. So audio isn't my only business. I do audio for the desire to create elite-level products and to thwart as many expectations as I can. When I started I examined how products are currently marketed, designed, manufactured, theorized, and developed and literally wanted to figure out how differently I could approach it. Far too many designers grab off the shelf parts based on datasheets, don't do real testing, and then try to get rich selling something they threw together. My power supplies no longer use Jensen 4-pole capacitors or cascading regulators because I found there were technologies better suited to filtration for computer-based audio. Yet most people still consider fancy regulators and a bank of Jensen 4-poles to be the end all of power supply design. If you haven't experienced the themes that I discuss it's not surprising. Most people have nowhere near the low distortion in their systems to hear it. (high resolution and low distortion are wholly different things). But it would be wonderful to hear about the tests you did to come to your conclusions. I am a student first and foremost. I am not married to my concepts or ideas and if I find something that is more right or makes more sense then I adapt. You will likely see throughout this entire thread I've changed my approach several times as I learn more. I'm sharing the ideas with you free of charge, why do you guys complain about that? So believe me or educate me. It obviously bothers you that you think I'm saying something that is false, so show me the light. I like holding hands.

It looks like a few of the comments on the blog are spam. The link to sex-dating seems to be a sure giveaway lol. I need to turn on comment monitoring. None of the comments are manufactured, however. Thanks for pointing that out. The guys at Microsoft and Facebook were also called unusual and their ideas frowned upon.

http://www.coreaudiotechnology.com/blog/?p=34

Yep, the .0005% resistors from Vishay are the very highest tolerance they can make. We spoke to some of the engineers and Vishay at length and continue to do so as we build this DAC. We actually stack the resistors above and below the board rather than side by side so that they can be lined up with virtually zero trace length. Closer than we can get SMDs. Thanks for the referral to that company! Great to hear that. I've heard more and more people say that resolution and file formats don't make a difference as their systems become lower distortion. It's usually people who have upgraded their power supplies on one component or another. The lower distortion your system becomes the less resolution, file formats, application, and tweaks will matter. Speaker cables and interconnects no longer make any difference in my system either, but you need a very low-distortion system to get to that point. I wrote an article about why if anyone cares to read it, just let me know and I'll send it to you.

Not the article I was thinking of, but it has some interesting tidbits about the TotalDAC, which is another R2R DAC. Says right at the top "The DAC is configured as a an R-2R ladder. It is a 24-bit network although 0.01% resistors can not easily give a resolution higher than 14 bits." http://www.digikey.com/Web%20Export/Supplier%20Content/VishayPrecisionGroup_804/PDF/vishay-cs-multi-channel-dac.pdf?redirected=1 I have a huge problem with the board layout and power supply on that DAC. No idea why he's using 16 ICs when he could be using the FPGA to do everything, But otherwise it's quite interesting and a step in the right direction for DAC tech IMO.

That's what I'm working on with my new Kryptos 2.0. I've already installed my iPF internally in the chassis (see above for details). The next thing I'm going to try is adding two more regulation stages and bypassing the ones on the logic board. The problem is the ever-changing apple tech and the fact that people still see it as a mac mini and not a music server. My Kryptos Blu-ray is windows based and in that unit I use my own multi-voltage supplies and regulators instead of the board's DC-DC converters individually to the PCI cards, drives, and Mobo. We've tried paralleling resistors. It's incredibly expensive and diminishing returns kicked in at the .01% mark with Caddock's TF020, Holco, and various SMD resistors we tested. The zero trace length makes a large difference in distortion, but achieving high resolution just is beyond modern technology still. Personally, I'm very happy with accurate 16 bit audio. If they can't attain accurate 24bit audio on the $2M Euphonix systems I've worked on then I'm not going to hold my breath about doing it on a much less expensive device. For those who aren't aware, Euphonix used to build the Pacific Microsonics Model 2. And yup... already using balanced outs on the FPGA. The FPGA can compensate for resistance drift very nicely and match the resistors almost perfectly even if we're not paralleling them. But there still isn't enough speed to achieve high-resolution. It's like an assembly line designed to build 10 units per second and we're trying to push 20 in the same amount of time. It can be perfect at making those 10 units, but no matter how much we want it to do more than that it just doesn't have the capacity to process it. I really should find that article...

Ones and zeros are read as a voltage range. For instance a zero is any voltage between 0 and 1.2V and a one is anything between 2.1 and 3.3V -- this is just an example and these ranges will change from device to device depending on what voltages the DAC chips run on, etc. So as long as the voltages generated are within those ranges we're fine. That isn't the problem really. Ever notice the difference between a flat square wave and a square wave with jaggies? Those jaggies are amplitude distortion on either a one or a zero. When those distortions are close enough to the middle ground between ones and zeros a one can be read as a zero or a zero a one by the device. Even if it is read correctly, these jagged lines when processed by the chips cause distortion in the data. This is what results in glare, hardness, or edginess in digital reproduction. Aftermarket clocks compensate for the swaying of voltage in the chip, but they can't entirely correct amplitude spikes and distortion. The better the power entering these chips the better the signal as a result. But you are correct, there is still distortion caused by the chip itself whether by swaying voltages, heat, multiple items in the signal path, the need for multiple chips, etc. You are also correct that this is part of the reason for a discrete R2R DAC. I've said this before, but high resolution is impossible. Data sheets can say 24bit, but there is not enough tolerance in the parts to achieve that resolution. I'll have to find it, but there is an article written by Vishay and TI about high resolution D/A technology and how 24bit audio is really closer to 8 or 12 bit. This is because the parts aren't fast enough or powerful enough to read that much data so the data gets compressed. People hear a positive difference compared to say an MP3 because one error on a smaller file vs one error on a larger file will be a relatively smaller impact on the high-res file. A smaller piece of the pie, if you will. But when there is less distortion you will start to hear the compression in high-res files and likely prefer native 16 bit files. Just my experience. I use .0005% tolerance resistors custom-made by Vishay in my R2R DAC and they still aren't high enough tolerance to achieve more than 20bit accuracy. Resistor networks have the benefit of much better timing without the need of additional buffers or clocks, but just by the way they function, slight variations in tolerance can hurt the signal dramatically. R2R and NOS DACs work (overly simplified) because they have 16 or 24 latches and data paths, one for each bit. So the whole word is loaded into a serial registry buffer and a snapshot is taken. With Delta-Sigma you have one latch and one data path so you need considerably more snapshots that are all added together by various algorithms. The problem with R2R is that every resistor must be matched for an accurate representation of the digital signal. Slight variations up or down change the frequency response dramatically. I am dabbling around with a FPGA by XILINX called the Virtex 6, which not only allows me to have zero traces between the 128 resistors per channel, but allows me to perfectly match the voltages in real time for each resistor and from channel to channel, which in turn are all equidistant from the FPGA. I've barely scratched the surface as to what the FPGA can do... it'll actually allow me to program and control all the inputs and outputs (firewire, USB, i2S, AES, SPDIF, Toslink, etc), program the microcontrollers in my digital amp, and do do various types of error correction. FPGAs allow for incredible control of the digital circuit. It's way over my head still. I'm working with people who know more about that voodoo than I do. But it's a neat concept that I'm sharing for fun, and not to advertise it... it's still a ways away until I'd be comfortable even demoing it for my Dog. The problem is that the better my source and amps get the less the DAC matters. I can put a $100 DAC or a $5000 DAC in my system and there's barely any difference at all. So most of my tests have to be done on a lower end system to make sure I'm actually able to hear the differences between different resistors, FPGAs, and power supply designs. The original design had only 55 power supplies.

It should be forwarding you to a coming soon page. But the new pages will be updated and available tonight. The Tact stuff is a great idea and part of why I went the modular digital amplifier route to bypass the A/D conversion and the need for an additional preamp and DAC. I'm a big fan of "less is more" especially in digital. Better power will improve every component in your system. Paul makes some excellent regulators. A regulator will not fix the noise that impacts digital signals, but it still will improve the sound. They are excellent as a first step and great for analog products, but digital sees a totally different spectrum of noise than analog. So while regulation circuits are very beneficial, they are not the ultimate solution for digital components. It'd be like comparing a bank of caps to a good regulator... it's not a one or the other that does the job better, but a sum of the various technologies and parts that are used.