Voilà ce que Superdad le concepteur des uptone audio disait en 2014 des teradak. Il est partie prenante donc propos à prendre avec des pincettes
http://www.computeraudiophile.com/f10-m ... ndex3.html
No Adam, it is a linear PS, but you get what you pay for. Build quality is very variable, they don't say what the bandwidth is for their noise spec, and one must be very careful not to bump the knobs or you will get much more voltage than you wanted.
Watch out for some of the Mastech models: this one does not have a fan, but a lot of their others do. Frankly, if I was going this cheap I'd get 3 fixed voltage LPS units, or just a good ATX SMPS.
Besides, ripple is not the only spec that matters in building a quality PS for audio.
Et ce qu'il disait des hdplex qu'il faudrait rajouter à ta liste à 385 dollars
http://www.hd-plex.com/HDPLEX-Fanless-L ... evice.html
elles sont en 19V/12V/9V/5V ( avec 12V/7A)
http://www.computeraudiophile.com/f27-u ... wow-23421/
Hi:
Sorry to be slow in reply to your question. There is no doubt that the HDPlex is a fine unit and a fantastic bargain at its price. It is astounding what they can build things for over in China. The truth is, my total parts cost (including the cables and packing) for the JS-2 is a few dollars more than the retail price of the HQPlex. That does not include any of my 5 hours labor and testing, the $50 per I give John Swenson as royalty on his design, or the 3-4% PayPal charge I absorb.
As for the performance and features of the JS-2 itself, there are some significant differences--aside from the flexibility of user-adjustable output voltages and 5A capability at all voltages. While I will avoid criticizing the bargain HDPlex, I can't say positive things about their choice to use $1 bridge rectifiers, but I applaud their use of an R-core transformer (just as we do; through we have ours shielded, and 25% of my R-core cost is in shipping the darn things over from China).
Maximum and peak current ratings are no problem, and our units are rather close in that regard. But when one talks about measurements (as I think SandyK was bugging me for over in the other JS-2 review thread), that becomes VERY problematic because a single figure for noise/ripple is a pretty meaningless measurement.
First off, unlike an SMPS, a properly designed linear power supply should have basically zero ripple--certainly not at line frequency and not at any particular frequency. All that should be measurable is residual noise. But over what bandwidth? From 0Hz to 100KHz? Or up to 500KHz or several megahertz? And is that under a continuos or varying load?
So not only is comparing a single number between manufacturers meaningless because it depends on the the range and the nature of the load, but actually performing the measurements is difficult simply because the ambient environment (fluorescent lights, other gear, etc.) will creep into sensitive test equipment. Heck, take you cheap multimeter and wave its leads around any gear and you will get a dozen millivolts reading.
So the only way to fairly compare noise measurements between two power supplies is to have them side by side and use the exact gear, the exact bandwidth, load etc. And still, I'd rather look at the noise spectrum on an analyzer than distill it down to one pointless number.
All that said, what really makes for a great LPS in audio are several things other than noise/ripple. Response to transient loads being number one; rejection of line noise and not putting crap back into the line being two next factors.
Stringerbell