DS-ATX DC-DC PSU Review
Review performed by: Tim Elmore
I am majoring in Mechanical Engineering at the University of Florida, and have taken several classes dealing with electronics and heat flow/heat transfer. I am competent in circuit design, including discrete components, operational amplifiers, power regulation, and control/instrumentation.
1993 Ford Explorer w/ high output alternator
4ga cabling to rear distribution block
US Acoustics 800w subwoofer amp, 65x4 speaker amp
Audio fed through Pioneer headunit
CDT Audio CL-61a components
AMD Duron 950 CPU @ stock
Abit KX7-33 motherboard
80GB Hard drive
Voodoo 3 PCI video card
D-Link wireless card
96kbit Envy DAC on Chaintech PCI audio card
Xenarc 7” touch screen on PSU power
Fluke 179 Multimeter w/ temperature probe
Packaging was excellent. PSU was inside electrostatic safe bag, wrapped in bubble tape, with shredded paper packaging inside a cardboard box. PSU came with the following:
ATX cable w/ 2 molex connectors and 3 floppy connectors
MB-ON signal cable
40mm fan w/3pin header
Serial port dongle
Testing was performed under the harshest of conditions. 4 foot 18ga power wires were run off the beefy distribution block. This is very undersized for a 20 amp power supply. 10 or 12 ga is recommended. A voltage drop of almost 1 volt was measured across the wiring, meaning that the power supply got 12.6v if the car’s electrical system was at 13.6v. This makes cranking even more harsh, as the car battery usually dips to 7-9v, the PSU was seeing 5-7v. Both power supplies were run without a fan, including during the temperature tests.
Test data for the M1-ATX at full load was not available because I was unable to keep the computer from rebooting long enough to take readings.
Voltage rails on the DS-ATX are very, very solid. I noticed a fluctuation of about .02vDC, which is very, very good. The rails were dead on, even under load. Generally, as a power supply becomes overloaded, its’ rails begin to sag. This is evident in the M1-ATX, where the PSU’s rails are beginning to sag appreciably. Voltage ripple is also very exceptional. ATX specifications require voltage ripple to be under 100mVAC, and the DS-ATX is well within specifications. I included the PC Power and Cooling power supply as a reference, simply to put the ripple voltage in perspective.
Temperatures were exceptional for a power supply of this physical size and power output. The temperatures were 15-20C over ambient with passive cooling, indicating the power supply is not under heavy load, and also indicating component choice is good. Comparing to the M1-ATX, we see that the components get very hot, 30-40C over ambient. The component choice on the M1-ATX is poor for a system of this size. The inductors on the DS-ATX are much larger physically than the inductors on the M1-ATX.
Shutdown controller / serial port / power and signal connections:
I found the shutdown controller to be excellent. It is easily configurable over the serial port, and worked absolutely flawlessly. Compared to the M1-ATX, I found the DS-ATX’s SDC to be much more powerful and adaptable. The serial port interface was also excellent, and easy to interface with Windows Hyperterminal. The reported voltages were very close to what was reported by my Fluke, like within .05vDC. Program mode worked excellently on the unit, as did forced shutdown. Power and signal connections were excellent, with beefy screw terminals for power, and solid plug/jumper blocks for signal and DC output. The serial port interface, however, needs revision, as it is too easy to bend/break the pins with the dongle sticking up from the main board. If a 4 pin extension cable was made, the dongle could be laid flat, and accidental tugs on the serial able would not bend the dongle off.
I am extremely impressed with this PSU. It is evident that a significant amount of design and testing went into this. It is by far the best DC-DC PSU I have used, and has the most adaptable shutdown controller known to me. I have used it about 1 hour so far, and have had absolutely no problems. I should add that Through 4 hours of testing, I have noticed no voltage rail fluctation, even with heavy bass, and cranking.
-Very high output, under rated power supply
-Robust shutdown controller
-Great voltages and temperatures
-Very good value at anything less than $200
-Serial interface is prone to physical damage
9/23/05 - fixed URL, added comment in overall analysis, added comment in testing methodology