iXBT Labs - Computer Hardware in Detail






Review of Power Supply Units from Thermaltake and FSP Group

The first series of power supply units we are going to examine is three models, available in our lab: PUREPOWER680-APD and TWV-480AD from well-known Thermaltake as well as FSP550-60PLN from FSP Group.

Running a few steps forward, I'd like to note that I'm not yet planning on publishing a lot of PSU photos in the exploded form in the nearest future — in some cases components are so tightly packed on a PCB that it's rather hard to make out the innards on a close-up, because of the heatsinks on key transistors, for example. Nevertheless, I'm not going to renounce such photos either.

Exterior, bundle

Both Thermaltake models come shipped in traditionally bright boxes, literally covered with icons, which may help a person to get a general idea of PSU parameters even without the English knowledge. Let's start with a junior model, TWV-480AD. The case of this power supply is made of matt black metal (painted), which contrasts with the orange plastic of two 80-mm fans. Besides the traditional user's manual, the bundle includes a 20-24 pin power adapter, an additional 80-mm fan, a power cord, and a fan control panel to be installed into the 5.25-inch drive bay. Besides two rpm controls, the control panel houses a three-section display that shows the current power consumption of the system.

PUREPOWER680-APD is another power supply unit in a mirror case (black color). Like the junior model, this supply unit is equipped with two 80-mm fans (a fan on the rear panel and a fan on the bottom that drives the air in). The bundle includes a traditional power cord, user's manual, and a 20-24 pin power adapter.

We got the FSP550-60PLN unit as part of a stand-alone solution, so unfortunately we cannot say anything about its bundle or package (if it had any). The unit itself has a regular case of a metal color, it's equipped with a single 80-mm fan. Judging from the type and number of connectors, this PSU does not pretend to be a server power supply — especially as such PSUs are sure to have additional requirements, like a hot swap feature.


To all appearances, the manufacturer positions the PUREPOWER680-APD as a server PSU — the increased number of power connectors corroborates to it. Note that this model has two power connectors for PCI Express video cards - so this PSU can be used in graphics workstations as well; the other two PSUs are intended rather for regular but powerful enough systems:

  Thermaltake Purepower 680-APD Thermaltake Purepower TWV480 FSP550-60PLN
Main connector 24-pin 20-pin 24-pin
SATA (15 pins) 4 2
Peripheral power connectors (4-pin) 10 9 5
+12 V (4-pin) 1
FDD (4-pin) 2 1
PCI Express +12 V (6-pin) 2  
+12 V (8-pin)   1
AUX   1

Before we analyze the stability of output voltage in the power supply units and their pulsations in the main (+12 and +5 V) power buses, we'd like to dwell on a couple of issues. Let's have a look at the oscilloscope specifications first:

Bandwidth, open input 0 - 1 50 MHz (-3 dB)
Input impedance 1 Mom 20 pF
Dynamic range (from the center of the screen) over 8 divisions per 150 MHz
Sensitivity 2 mV/div - 5 V/div (in 1-2-5 sequence, 1 2 steps), except for 0.2 5ms/div
Precision 3% (5% for 2 mV)
Input coupling AC, DC, GROUND
Max. admissible input voltage 4 00 V peak (DC + AC peak at 1 kHz)
Rise time about 2.3 ns
Signal excursion ±0.2 div at 5 mV/div - 5 V/div, ±0.5 div at 2 mV/div
Cross talk Max. 0.3 div at 50 MHz
Attenuator ±0.2 div. at 5 mV/div~5 V/div ±0.5 div. at 2 mV/div
Ray drift ±0.3 div. per hour
Resolution Approx. 80 psec
for repetitive signals 2 ns/div - 0.1 µc/div
realtime 0.2 5µc/div - 0.1 sec/div
ROLL mode 0.2 sec/div - 5 sec/div
Sync advance Maximum 10 divisions
Horizontal excursion Not less than 10 divisions
Data acquisition Maximum sample rate 200 MS/s for a single channel,
100 MS/s for 2 channels simultaneously in realtime mode
25GS/s for two channels simultaneously (repetitive signals)
ADC resolution 8 bit
Bandwidth limit 20 MHz
Peak detector of a single sequence 10 ns (5 µc/division - 5 sec/division)
Display memory 32 K of words per channel
Mean alternating, 2 - 1 28
Persistence realtime only
Trigger Oscilloscope sensitivity, CH1 and CH2
Input Frequency
5 mV - 5 V/div 2 mV/div
CH1 0 - 1 0MHz 0.5 div 0.5 div
CH2 1 0 - 8 0MHz 1.5 div 1.5 div
at 10 MHz - 40 MHz
EXT 80 MHz - 150 MHz 2.0 div
Trigger type Edge, TV
Trigger modes Auto, Normal, Single Sequence
Polarity + / -
Trigger source CH1, CH2, EXT (external), LINE (line frequency)
Input AC, DC, LF reject, HF reject
TV signal synchronization vertical, line
Trigger level control internal: ±3 div, external ±35% of the sync-signal, 4 V amplitude
Precision of the automatic calibration to 5 0% ±0.2 div
Trigger sensitivity in external mode 0.2 Vp-p (0 - 150 MHz)
Maximum voltage at the external input 400 V (constant + peak) at 1 kHz maximum
Cut-off high frequency Approx. 50 kHz (-3 dB)
Input impedance Approx. 1 MOm
Display Display 5.7-inch LCD, backlight — cold cathode fluorescent valve
Resolution 320 x 240 pixel
Controls brightness control on the front panel of the oscilloscope
Power supply Line voltage: 90 V - 250 V. Line frequency: 48 Hz - 440 Hz
Power consumption Max. 30 W
Operating temperature 10°C - 35°C (autocalibrated at 25°C ± 5°C)
Maximum range of operating temperatures 0°C - +4 0°C, 45% - 80% relative humidity
Storage temperature -10°C - +60°C, 35% - 85% relative humidity
Dimensions 3 70 mm x 1 67 mm x 3 38 mm
Weight 5.5 kg

All the measurements were taken with the standard TP6060 probes. Secondly, we should note the oscillogram display types. This device provides two modes: dots and vectors. Of course, the vector display mode seriously increases the number of displayed interferences, while they are not to be seen in the dot display mode (you can see only peak dots). We'll publish diagrams both in dots and vectors where necessary. Taking into account our readers' comments to the first articles about PSUs, in some cases we took additional oscillograms for improved accuracy to detect real non-random interferences. On all the diagrams published below CH1 - +12 V, CH2 - +5 V, the division ratio of a probe is 1:1.

The load error (in the voltage metering mode) is from 0.05 to 0.1%, the output is mean results for 200ms.

Purepower 680-APD

In this case pulsations in both buses (+5 and +12 V) turn out significant. Moreover, +5 V pulsations considerably pass the limits mentioned in PSDG - 10%. Voltage stability test results: the minimum measured value on the +12 V bus is +11.78, and the maximum value is +12.18 V. The minimum value on the +5 V bus is +4.78, maximum — +5.16 V, +3.3 V bus — +3.16 and +3.30 V correspondingly.

Purepower TWV-480AD

+12 V bus pulsations are about 20.8 mV, +5 V bus - max. 16.8 mV. Voltage stability test results: the minimum measured value on the +12 V bus is +11.45, and the maximum value is +12.20 V. The minimum value on the +5 V bus is +4.80, maximum — +5.14 V, +3.3 V bus — +3.18 and +3.40 V correspondingly.


+12 V bus pulsations are about 46.4 mV, +5 V bus - max. 24.8 mV. Voltage stability test results: the minimum measured value on the +12 V bus is +11.58, and the maximum value is +12.15 V. The minimum value on the +5 V bus is +4.79, maximum — +5.15 V, +3.3 V bus — +3.22 and +3.44 V correspondingly.


The reviewed power supply units seem rather stable, none of them violated the voltage tolerances recommended by PSDG. But Purepower 680-APD perplexed us with its pulsations, even additional readouts didn't improve the situation, +5 V bus pulsations always exceeded the 10% limit, specified in PSDG. Probably, these results (PSUs in general) may be governed by the load applied during the test (1) being much lower than the officially claimed one, and (2) corresponding to the relatively abstract values, suggested by the Power Supply Design Guide (we used the loads, recommended for measuring the efficiency of 400W PSUs according to the PSDG table). As this document lacks any rough values for PSUs above 400 W, we probably should review the power supply load options in order to obtain more or less trustworthy data, which would allow to evaluate the PSU quality compared to other models. The current method initially puts PSUs of different power capacity into unequal conditions - say, 460 and 550 W PSUs are loaded as an abstract 400W model.

To be continued...

Alexei Kuznetsov (axbat@ixbt.com)
April 15, 2005.

Write a comment below. No registration needed!

Article navigation:

blog comments powered by Disqus

  Most Popular Reviews More    RSS  

AMD Phenom II X4 955, Phenom II X4 960T, Phenom II X6 1075T, and Intel Pentium G2120, Core i3-3220, Core i5-3330 Processors

Comparing old, cheap solutions from AMD with new, budget offerings from Intel.
February 1, 2013 · Processor Roundups

Inno3D GeForce GTX 670 iChill, Inno3D GeForce GTX 660 Ti Graphics Cards

A couple of mid-range adapters with original cooling systems.
January 30, 2013 · Video cards: NVIDIA GPUs

Creative Sound Blaster X-Fi Surround 5.1

An external X-Fi solution in tests.
September 9, 2008 · Sound Cards

AMD FX-8350 Processor

The first worthwhile Piledriver CPU.
September 11, 2012 · Processors: AMD

Consumed Power, Energy Consumption: Ivy Bridge vs. Sandy Bridge

Trying out the new method.
September 18, 2012 · Processors: Intel
  Latest Reviews More    RSS  

i3DSpeed, September 2013

Retested all graphics cards with the new drivers.
Oct 18, 2013 · 3Digests

i3DSpeed, August 2013

Added new benchmarks: BioShock Infinite and Metro: Last Light.
Sep 06, 2013 · 3Digests

i3DSpeed, July 2013

Added the test results of NVIDIA GeForce GTX 760 and AMD Radeon HD 7730.
Aug 05, 2013 · 3Digests

Gainward GeForce GTX 650 Ti BOOST 2GB Golden Sample Graphics Card

An excellent hybrid of GeForce GTX 650 Ti and GeForce GTX 660.
Jun 24, 2013 · Video cards: NVIDIA GPUs

i3DSpeed, May 2013

Added the test results of NVIDIA GeForce GTX 770/780.
Jun 03, 2013 · 3Digests
  Latest News More    RSS  

Platform  ·  Video  ·  Multimedia  ·  Mobile  ·  Other  ||  About us & Privacy policy  ·  Twitter  ·  Facebook

Copyright © Byrds Research & Publishing, Ltd., 1997–2011. All rights reserved.