The 2014 El-Cheapo Power Supply Roundup

Thermal Master TM-420-PMSR

First Look

This is a power supply which I am not at all new to. They come bundled with low end Cooler Master cases, and were very popular in new PC builds around here a few years ago. I have load tested a few of them, and like the Solytech-built TM-420-PSAR-I3, they have always made it very close to the finish line and exploded only at 100% load. Let’s see if that happens today or if the near 40°C weather today causes it to go kaboom sooner.

The label is very similar to the PSAR-I3 variant, with the difference being that this unit is made out to be a single rail, 5V heavy design, while the other unit was made out to be a dual rail 12V heavy design (which was actually false). This unit’s exterior is generally very similar to that of the other variant.

Test Results

Test 1 (74.3W Load)

Rail	Load	Voltage	Ripple
12V	2.38A	12.02V	23.8mV
5V	4.99A	4.99V	10.0mV
3.3V	5A	3.4V	19.2mV
−12V	0.09A	−11.35V	18.8mV
5Vsb	0.52A	5.21V	4.2mV
AC Power	98.0W
Efficiency	75.81%
Power Factor	0.62
Intake Temp	36°C
Exhaust Temp	40°C

.
Test 2 (101.85W Load)

Rail	Load	Voltage	Ripple
12V	4.69A	11.93V	27.2mV
5V	5.01A	5.01V	11.6mV
3.3V	5A	3.4V	23.2mV
−12V	0.1A	−11.43V	21.2mV
5Vsb	0.52A	5.21V	4.8mV
AC Power	130.2W
Efficiency	78.23%
Power Factor	0.65
Intake Temp	37°C
Exhaust Temp	41°C

.
Test 3 (154.48W Load)

Rail	Load	Voltage	Ripple
12V	9.19A	11.8V	36.0mV
5V	5.03A	5.03V	12.8mV
3.3V	4.99A	3.39V	23.8mV
−12V	0.1A	−11.59V	28.4mV
5Vsb	0.52A	5.21V	5.8mV
AC Power	195.2W
Efficiency	79.14%
Power Factor	0.64
Intake Temp	37°C
Exhaust Temp	43°C

.
Test 4 (200.18W Load)

Rail	Load	Voltage	Ripple
12V	9.19A	11.99V	41.0mV
5V	9.9A	4.95V	13.4mV
3.3V	10.21A	3.37V	28.2mV
−12V	0.1A	−11.9V	32.0mV
5Vsb	1.04A	5.19V	7.0mV
AC Power	249.0W
Efficiency	80.39%
Power Factor	0.65
Intake Temp	37°C
Exhaust Temp	47°C

.
Test 5 (252.42W Load)

Rail	Load	Voltage	Ripple
12V	13.69A	11.85V	40.8mV
5V	9.94A	4.97V	16.4mV
3.3V	10.18A	3.36V	26.8mV
−12V	0.1A	−12.08V	38.8mV
5Vsb	1.04A	5.18V	10.0mV
AC Power	318.1W
Efficiency	79.35%
Power Factor	0.64
Intake Temp	38°C
Exhaust Temp	52°C

.
Test 6 (307.08W Load)

Rail	Load	Voltage	Ripple
12V	18.32A	11.78V	43.2mV
5V	10.02A	5.01V	17.0mV
3.3V	10.21A	3.37V	24.4mV
−12V	0.1A	−12.33V	42.8mV
5Vsb	1.04A	5.19V	11.0mV
AC Power	408.3W
Efficiency	75.21%
Power Factor	0.62
Intake Temp	39°C
Exhaust Temp	59°C

.
The 12V rail started at 12.02V and dropped to 11.78V, which equates to 0.22V (1.83%) regulation and a drop of 0.24V, or 2%. The 5V rail had maximum and minimum values of 5.01V and 4.95V respectively, which equates to 0.05V (1%) regulation and 0.06V, or 1.2% variation. The 3.3V rail was always a bit on the high side – it started off at 3.4V, and dropping to 3.37V, giving us 0.1V (3.03%) regulation and a drop of 0.03V, or 0.91%. This is a perfectly acceptable result, although it would have been nice if the 3.3V rail had stayed closer to its nominal value.

The efficiency just cleared the 80% mark in Test 4, but was below 80% for the other tests. This power supply was another hot running unit, with a 20°C Delta between the intake and exhaust temperatures at 300W load. As I suspected, the high intake temperatures did have an impact on the capacity of the unit. This time around, both switching transistors blew at 350W load. Now, I know some of you may be thinking that I was going a bit hard on the unit (and the Numan) by running it this hot, but guess what? 40°C summer weather is common in this part of Australia and a product that can’t survive it should not be sold here.

Rail	Test 5 (252.42W)	Test 6 (307.08W)
12V
5V
3.3V
−12V
5Vsb

.
The ripple was fairly well suppressed, and was generally below half the maximum allowed, although the 3.3V rail was a bit over at times. It’s just good enough that I won’t score against it.

Disassembly

The input filtering consists of two X capacitors, two coils, two ceramic capacitors and three MOVs. Unfortunately, the ceramic capacitors used are not safety rated for use on the mains. If they are overstressed, they will short, which will cause the case to become live with mains power if the earth connection is lost. The other key components are a 6A bridge rectifier, 470µF input capacitors from Fuhjyyu, and Jilin Sino Microelectronics 3DD13009K switching transistors. The 5vsb circuit uses a 2-transistor design with a 2SC3150 MOSFET rated at 3A as the main switching transistor. The critical capacitor is made by Sapcon – not one of the high quality manufacturers.

The 12V rail uses two F1620CT Fast Recovery Rectifiers, which are rated at 16A each, so the 12V rail would theoretically be capable of up to 32A, if the other components are up to the job. The 5V rail uses a Lite-On SBL3045PT Schottky rectifier, which is rated at 30A. The 3.3V rail uses an SBL2045CT Rectifier, which is rated at 20A. Some of the capacitors on the secondary side are Sapcon brand, while others are Canicon. Neither of them are high quality brands.

Compared to the Numan, the destruction and charring on the primary side was minor. Both of the switchers do have their washers melted, but they aren’t really badly charred. The two resistors underneath (labelled R12 and R18 in the above right picture), though, are burned to a crisp.

As is usually the case on Sun-Pro power supplies, the fan is made by Te Bao Metallic Plastic. I have seen these fans fail often. This fan had thick grease used as the lubricant, which isn’t as effective as oil. It is temperature controlled, but interestingly, the thermistor was placed in the smaller 3.3V toroid coil. This is a very bad idea, as it means that the fan speed is only affected by the load on the 3.3V rail, which is not that heavily used on most PCs. It was audible throughout the testing, and it became much more noticeable from Test 4 onwards (when the 3.3V rail’s load was increased to 10A). The heatsinks are quite thick with plenty of fins. This power supply has some of the chunkiest and most effective heatsinks of all the units in this roundup.

Specifications and Conclusions

Real Wattage	300W
OEM	Sun Pro
PFC	None
Price	Unknown
ATX Connector type	20+4 pin
Worst-case voltage regulation (12v, 5v, 3.3v)	1.8%, 1%, 3%
Worst-case ripple (12v, 5v, 3.3v)	43.2mV, 17.0mV, 24.4mV
Worst-case efficiency	75.21%
Input filtering	Adequate
CPU Connector	ATX12V (4 pin)
PCI-E Connectors	None
Molex (Peripheral) Connectors	4
FDD Power connectors	1
SATA Power connectors	2

.

Pros: Reasonable ripple suppression

Cons: Can’t deliver labelled rating (−2), Low quality capacitors (−2), Non-safety-rated ceramic capacitors where Y-capacitors should be (−1), Low quality fan (−1), Fan only controlled by 3.3V rail load (−1), Mediocre 3.3V voltage regulation (−0.5)

Score: 2.5/10

< Previous Page | Next Page >

Pages: 1 2 3 4 5 6 7 8 9 10 11