Testing
Before I get the limited testing results out of the way, I'd like to refer back to The Tech Report's graphs on PSU ripple voltage on the AC rails, +3.3V, +5V, +12V. Like recent Antec supplies, although not adjustable, the Tagan TG480-U01 has separate rails for all three output voltages. That means if the +3.3V rail takes a heavy load, since the +5V rail isn't derived from it, the +5V AC supply won't dip at the same time.Geoff measures AC ripple voltage on all three output voltages, over a 10ms period under load, recording the sway from the target output. Taking the +3.3V results linked above, we can see that a few PSUs do very well, with an average of around 50mV p-p, swaying less than 2% either side of the +3.3V target. Tagan state in the product literature that, although I'm unable to test it without a 'scope, the TG480-U01 has a similar +3.3V ripple sway of < 50mV p-p. Comparable to the results Geoff obtained and Tagan explicitly state it's a 100% load value, not 'idle'. You get a similar class of stated values for all the rails, Tagan are proud of the clean supply they guarantee the TG480 to have.
Over-voltage, over-current and short-circuit ratings are all high. For example, +4.2V for the +3.3V rail, +6.3V for the +5V rail and +17V for the +12V rail. They are peak burst protection ratings, but they should keep the PSU ticking over nicely, even in the event of a high voltage burst on any of the rails.
The stated current load limits are high for the class of supply, 28A on +3.3V, 48A (54A peak) on +5V and 28A (30A peak) on +12V. It's a supply you can happily use with high current draw devices such as Asetek Vapochill.
Finally, Tagan state the PSU will supply voltage to all rails, to at least 90% of the rated output voltage, within 100ms of power on. Full supply voltage at 100% comes less than 25ms after that with less than 500ms needed for complete power on, with full load on all rails, at maximum output voltage. Basically, it's ready to rock in under half a second.
Load Test Results
In the absence of any explicit PSU testing equipment, testing is limited to running an overclocked Athlon 64 testbed, with as many fans and disk drives as I could find, under 3D and CPU intensive loads, monitoring the voltages with Motherboard Monitor. It's barely scientific, but it allows some kind of qualitative assessment and comparison against my long-term supply, an Enermax 431W. Here's the setup.- Tagan TG480-U01 ATX 2.03 PSU
- Enermax 431 ATX 2.01 PSU
- AMD Athlon 64 Model 3200+, ASUS K8V Deluxe motherboard
- Corsair XMS3200LLPT TwinX 512MB
- ASUS Radeon 9800XT
- Seagate Barracuda ATA IV 80GB PATA HDD
- Western Digital Raptor 36GB SATA HDD
- 3 x 120mm YS-Tech 135cfm fans @ 12V
- Pioneer DVD-106 IDE DVD-ROM
- TX Jupiter 40x IDE CDRW
The disk data copy was scripted to repeat 3 times, deleting the intermediate data when finished. The voltages were measured over that time period, an approximate 35 minute period in which 78GB of data was copied and 3DMark03 and SETI were left running constantly.
So, constant fan, disk, GPU and CPU loads, all pulling on the PSU rails in some capacity and to some extent. It's not a maximum current draw test like Tom's Hardware like to do in their PSU reviews, but it's a somewhat real-world scenario that power users might come across every now and again, with power users the target market for the Tagan supply.
+12V variance results
| PSU | Maximum | Minimum | Adjusted Average |
| Tagan | 12.14 | 11.99 | 12.01 |
| Enermax | 12.33 | 11.83 | 12.02 |
+5V variance results
| PSU | Maximum | Minimum | Adjusted Average |
| Tagan | 5.04 | 4.94 | 4.99 |
| Enermax | 5.05 | 4.94 | 5.00 |
+3.3V variance results
| PSU | Maximum | Minimum | Adjusted Average |
| Tagan | 3.33 | 3.31 | 3.33 |
| Enermax | 3.42 | 3.29 | 3.30 |
The results speak for themselves. Both PSU's were equally able to cope with the testing, each holding close to the target output voltage. The +3.3V and +5V results are the most interesting. The Tagan has discrete supply for each, while the Enermax derives each output voltage from a shared supply. Under load, heavy load on one rail may pull the other rail down too on the Enermax, while theoretically the Tagan should be able to do much better.
The adjusted average figure is derived with peak and trough results removed, stopping them from skewing the results, however it's that peaking of output voltage that can cause a PSU the most problems in supplying clean output. The Tagan, with its return loop, also has an added advantage in high load situations, reducing ripple noise.
While both PSUs seem very competent, indeed I'm proud my long term Enermax did so well, a PSU that takes a newer, refined approach to supplying clean output voltage, like the Tagan, should always win outright.
While it's unscientific at best, the Tagan shines, holding steadfast to the desired voltages on each rail, regardless of load.