This linear power supply has got some traction on The Audio Standard
https://www.aliexpress.com/item/4001288349011.html?spm=a2g0o.order_list.0.0.55c218024DJ8dl
33 responses Add your response
@audioman58 Is not wrong. For speakers the driver to speaker cost is at least 10:1, a major reason why some turn to in-house drivers. It's not just engineering but cost. I've seen some brands like Rega pump up the ratio of the electronics even higher, say 20 or 30 to 1. I'm thinking of a headphone amp I took apart once. Ugh. |
Pretty much rule of thumb on average only 25% of monies spent actually goes into theProduct ,this includes case and packaging the rest R&D overhead and markup. that’s why intruth you cannot get anything good for that money the best LPS and made in U.S.A is Linear Tube Audio ,in MD . for around $700. You get a well built unit using worlds best Belleson Regulators and exact voltage you need and DC connector type . The finest sounding LPS under $1k ,save your money and buy this LPS,you will be glad you did ! |
| Post removed |
| Post removed |
Funny coming from someone claiming to be an engineer. You claim you want rigor and then immediately come out with broad, general brushstrokes. A well designed SMPS can have extremely low output impedance and noise, far lower noise than a LPS, if well designed. It's a lot easier to design a low noise SMPS than a low noise Class D amplifier so a lot of the problems associated with Class D are easier to deal with in a power supply. Unlike a Class D amp which has to output AC, for instance, an SMPS only has to put out 1 constant voltage.
|
Yes, I know how switching power supplies work. They use very small capacitors and a small transformer. Then use a circuit that injects DC charges onto the post transformer caps thousands or tens of thousands of times per second. This is different than linear power supply which charges the main power caps about 60 times per second (or 120 depending on how you look at it). I have done work with both switching power supplies and linear power supplies. The switching power supplies suffer from weak bass because of the small caps and transformer. The only way to get a switching power supply to really "act" like a linear power supply is to place a very large amount of capacitance after the transformer/charging circuit. Let your ears be the judge. The lab measurement devices don't tell the whole story.
|
Actually, no, its the noise. And i'm not sure what you man by AC noise, the noise is (by definition) AC superimposed on DC and is a product of the way Switching mode power supplies work. ( i wont go into that here, look it up)
Don’t use analogies, define your terms and explain the engineering.
Switch mode power supplies create a lot of high frequency noise - its well known, its a problem too. It can interfere with the digital circuit by making tiny adjustments to the zero crossing (we call that jitter) and it can beat down into the analog region (we call that non harmonically related distortion).
I design this stuff. Its the noise. and its a problem.
|
If interested: 12V 50W LPS for sale |
I cannot overemphasize the importance of power supplies. I started with an Antelope Zodiac Gold, then added a Voltikus power supply. Upgrading to a Zodiac Platinum with an upgraded Volikus power supply brought the next quantum leap. It ll really came together when I replaced the Platinum Voltikus with a Sean Jacobs DC3/4 PSU. In all of this I cannot overemphasize the importance of power cables and DC cables. In digital music any incursion of RMI/EFI or ground level noise has immediately audible nefarious effect. Rather than nuy new streamers or dacs upgrading power supplies is the more beneficial trade.To wit: the power supply cost more than my dac. Same story on the Etherregen: adding a top of the line Zero Zone PSU brought the upstream link up to snuff, again at significant cost. |
I have not done any testing for linear power supplies on things such as network music servers or network switches, so I cannot comment on if this makes a difference. However, any digital transport will be improved by a linear power supply, regardless of how well the internal regulation and filtering is. I will say any digital transport because this includes all transport devices both audio and video. If you cannot hear the difference, it is possible that your system is not high enough resolution to reveal this or that your hearing may not be sensitive enough (no insult intended).
It is much more than just blocking A/C noise. It is how smooth and powerful the current is. One analogy I use is to use air compressors as a comparison to switching/linear power supplies. You can think of a switching power supply as a really small 5 gallon tank air compressor with a small pump motor running at 10,000 RPM. Then use a larger 60 gallon tank with a 5HP motor running at 120 RPM as the analogy for a linear power supply. In the end, what kind of air/power would you get out of the small 5 gallon air compressor? |
This got me interested, once again, in the idea of external LPS devices. And in addition @itsjustme made a very helpful post. I already have components plugged into a Shunyata Hydra 6, but was wondering whether the DSP I am using would benefit from an external LPS. So, I emailed the support department for the DSP unit and received a very prompt and helpful reply. Seems, the company had already done their own research and found that for their particular unit, that an external LPS "may" not offer any benefit. Per their response: "We have not seen any measurable difference between different powers (from laboratory power supplies to cheapest possible inverter supply). However, the improvement gained can also depend on the quality of your local power supply network." So yes, it might be construed that an LPS could be of benefit assuming the local power supply network isn't of decent quality, but this is something that I would think would be difficult to determine without the appropriate measuring equipment. At this point I have decided to place the idea of an external LPS on hold. |
For anyone thinking of buying or making an external LPS here are a few thoughts to give perspective. First you need to ask what kind of device you are powering to know your objectives. If the device is a pure music server - not connected directly to the DAC but rather to a network, it really doesn’t matter. What maybe useful here is placing a filter between the noisy PC and its noisy switching supply and your mains/AC power to keep that noise OUT of other stuff.
On the other hand, for most streamers and bridges that spit out USB to a DAC directly, blocking power noise (most likely on the ground) is the goal. On the best devices the USB is fully isolated and powered by its own LPS. I do this on my home made stuff and prototypes. Otherwise, an LPS is very helpful. Generally there will be some form of regulation inside to down-convert the input DC to whatever the various chips use (often 5 and 3.3v). But those regulators are typically pretty noisy themselves - no one’s thinking "high end" when they design a computer, a Pi etc. Any actual computer will draw a bunch of power, more at boot up. Finally, lets talk DACs. DACs are largely analog devices and demand very quiet power. Most have internal regulation and filtering to accomplish this. Even if they have a wall wart or a switching supply externally, it feeds internal regulation and filtering. So bear in mind that what you are doing is providing a quieter input, including less ground noise, you hope. DACs also draw a lot less power typically. The DAC chips themselves don’t draw much, so its any digital filtering circuitry,clocks, USB receivers, and finally the analog output stage. Its those last ones that matter most, and again, they are typically fed by their own dedicated "LDOs" (low dropout regulators). Any external LPS is unlikely to be any better or even good enough, and the distance both in length and resistance to an external supply would render it in effective. So keep in mind that you are providing filtered, quieter input power, but are almost surely not powering the low level circuitry directly with whatever you get/build/conjure up. If you read the datasheet on any regulator you will see that its noise is a function of the input noise. Maybe its 100:1 improvement, often less, sometimes more. Always specified at a big number that is never achieved across the frequency band. So the better the input, the better the output - simple as that. Its called the "PSRR" or power supply rejection ratio. G
|
most circuitry is locally regulated ( and much of it way below 12V - more like 3.3 or 5V for chips) so the additional input voltage is not a big concern. It will generate a little more heat lost inthe down-regulation, btu the performance if anything, is enhanced. You could also check the maximum input voltage by asking the manufacturer if its not specified. Most are.
|
The battery idea sounds like (almost) a perfect solution. Small devices, like an ethernet switch eg, could be run off a 12v motocycle battery, which would last for weeks. One issue may be that a 12v car/motorcycle battery is 13.8v when fully charged, so I'm not sure of the tolerance of the input circuitry, although a 12v wall wart outputs more than 12v anyway when not under load, although it will drop when the device is powered up. The battery will not have that voltage drop. Batteries are also, relatively, cheap. I will buy a battery to power my switch/NAS/Streamer and post the results here in a few weeks. |
i do like the auto battery idea. likely superior to almost any LPS. But big, ugly, and requires a charger. as to LPSs - DACs are relatively low powered in general; but if you are powering a USB bridge or Raspberry pi or similar, those require much higher current, and a surprisingly large amount during boot up. You might share more about the device, what you have now, and what issues you are trying to resolve (or improve). I do think an LPS makes a difference. I designed and built one for pretty much every digital device i have-- and an upgrade for the one DAC that kinda needed such.
G |
Most DACs or preamps or small a/v devices will use less than 1 amp, so a 2A linear power supply will be just fine. If you want to go for a 4A, it will not hurt anything. If you have something like a video processor device (such as Lumegan) or a computer type device like Roon or Intel Nuc, then you may want a power supply that can provide more amps. The Keces P8 that provides 8A is the most powerful linear power supply I know of, but it's $900 retail! lol. The Swagman is at a similar level to the Teradak and Zerozone power supplies. They are just fine, but not as good as Sbooster. I think one of the better supplies is the Farad Super 3, but that is also expensive at around $600 or so. |
I always over spec my lps, less stressed, better reliability. So, you need 2amp, I'd get lps that could supply at least 4amp. I see too many posts of poor experience with lps, always because people purchasing lps that can only maximally supply needed amps, heat and stress lead to failure. Don't cheap out on lps! |
http://www.swagmanlab.com/Product_PSU?product_id=605 I don’t have the time to learn... not yet anyway. Saw these. They seem OK enough. But has me wondering about Amps now because they have 3 different versions. 2, 4 and 6a. |
What amp rating do you need? If it's for a low power device such as a DAC, then you can get away with a 2A power supply. I think the best you are going to do is keep your eye out for a used SBooster 12-13V model. They sometimes are sold in the $300 area. Otherwise, your only alternative is a Teradak or Zerozone power supply on ebay from China. |
