Review of the Chieftec Value APB-700B8 700 W power supply unit / Overclockers.ua

Today we are testing the Chieftec APB-700B8 power supply unit, the oldest model from the new budget VALUE series, with an 80 Plus certificate for 230V EU networks. Judging by the manufacturer’s information on its characteristics, it is not much different from other budget power supplies of this company, of which there are already quite a few. Let’s see what new Chieftec was able to offer in the budget segment.

Chieftec Value APB-700B8

The power supply unit is supplied without a box and power cable – to reduce the cost, there is no configuration.

Power supply unit with soldered cables, their number and length are as follows:

• one for powering the motherboard (55 cm);
• one with another 8-pin (4+4) processor power connector (55 cm);
• one with two 8-pin (6+2) connectors for powering a PCI-E video card (55+15 cm);
• two with two power connectors for SATA devices (50+15 cm);
• one with two power connectors for IDE devices and one FDD connector (50+15+15 cm).

All cables are made of multi-colored wires of medium length, and in large cases there may be difficulties with their neat arrangement.

The body of the block is made of steel with galvanic coating. On the side face there is a black sticker with the name of the series, on the top face there is a sticker with the technical characteristics of the unit.

The unit is built on the CWT platform, with an active power factor corrector (APFC) with an input voltage range of 200–240 V. The power converter is made according to the oblique bridge scheme, diode rectifiers with group stabilization of +5 V and +12 V lines, +3 line, 3 V is made on a separate stabilizer based on a magnetic amplifier. The circuit diagram of the block is not new and has been used since the first models of the ATX 12V 2.3 standard. A similar charge applies to other budget units from Chieftec.

A full-fledged impulse interference filter is unsoldered on the board in the input circuit, some of its elements are located on the network connector. The input rectifier GBU806 (8 A, 600 V) is installed without additional cooling. The power part of the block is controlled by the combined controller CM6805, soldered on the bottom side of the board. A pair of field-effect transistors marked CPT13N60 and a diode marked 8R06 are installed in the power part of the corrector. A pair of ITA20N50R transistors is installed in the power converter. All these elements are cooled by a common radiator. The high-voltage filter is made of an electrolytic capacitor with a capacity of 390 μF and a voltage of 400 V with an operating temperature of 85 °C, manufactured by Teapo.

Schottky diodes marked 30L60CT (30 A, 60 V) are installed in the output part of the power converter along the +12 V line, four of them are connected in pairs in parallel. The rectifiers on the +5 V and +3.3 V lines also have Schottky diodes, it was not possible to examine their marking. The output voltage on the +12 V line is filtered by a pair of Nichicon 2200 μF 16 V 105 °C electrolytic capacitors, which was unexpected in such a power supply. On the +5 V and +3.3 V lines, the filters consist of 2200 µF 10 V 105 °C capacitors from Elite and 1000 µF 16 V 105 °C capacitors from ChengX connected in parallel.

The power converter of the standby mode is made on the TNY177PN PWM controller. At its output, electrolytic Low ESR capacitors with a capacity of 2200 μF with an operating voltage of 10 V and a temperature of 105 °C manufactured by Elite and a capacitor of 1000 μF 16 V 105 °C from ChengX are installed. All other small capacitors in the standby power supply harness and the power converter are from ChengX and CapXon. The output voltage of the unit is monitored by the ST9S313-DAG supervisor from Sitronix.

Cooling of the unit’s components is ensured by a 120x120x25 mm fan marked HA1225H12S-Z (12 V, 0.58 A) and a two-pin connection. The fan speed is controlled automatically, depending on the temperature of the power components of the unit.

Installation and soldering are of high quality, the board is normally washed from flux.

Testing methodology

The power supply unit was tested using a linear electronic load with the following parameters: current adjustment ranges on the +3.3 V line — 0~–16 A, on the +5 V line — 0–22 A, on the +12 V line — 0–100 A. All contacts for connecting power supply cables with the same voltage are connected in parallel and loaded with the corresponding channel. The current on each channel is smoothly regulated, and it is stable regardless of the output voltage of the unit. A Zotek ZT102 True RMS multimeter was used to accurately measure voltage and temperature. For each power line, the required current was set and the voltage on the load contacts was measured to account for losses on the wires.

Test results

Since this block has group stabilization of the +12 V and +5 V lines, the load on one of them affects the voltage of the other. And if one of the lines is not loaded, then the voltage on it will rise significantly, and on the loaded one, on the contrary, it will fall, and it is impossible to get maximum power from the +12 V channel without a strong voltage drop, while not loading the +5 V line. In modern home computers, the consumption on the +5 V line is usually about 3–6 A, depending on the number of drives and peripherals, while the block is designed for 22 A on this channel, due to which the drawdown on the +12 V line will be noticeable. That is why the industry began to switch to separate stabilization with DC/DC converters in power supplies. Considering the above, the first part of testing the load capacity of the +12 V main line was done at a current of 10 A on the +3.3 V and +5 V lines, then a test was made at the maximum load on +12 V and a current of 20 A on the +5 line In order to check how the voltage balance between these lines will change. The second part of the testing was done at currents closer to a real home PC: 6 A on the +5 V line and 5 A on the +3.3 V line.

According to the test results, we have no better stabilization on the +12 V and +5 V lines. Without load, the voltage on the +12 V channel is overestimated by almost +5%, with an increase in the load, it sags and at 54 A we have 11.62 V at the output, i.e. -3.2%, meets the ATX standard, but on the border. At maximum load, the current was increased to 20 A on the +5 V line, while we can see that the voltage on the +12 V line increased to 11.82 (-1.5%). Also, at the same time, a measurement was made on an unloaded SATA connector to check the voltage at the output of the power supply without taking into account the drop on the wires, and we see quite large voltage losses, which is typical for budget power supplies – with high-quality wires, the stability would be significantly better. According to tests with a load of 6 A on the +5 V line, we can already see what the voltages will be in a system close to reality, and we have 11.54 V at the maximum load, which is -3.8%, which is included in the tolerances of the ATX standard, but it does not look good against the background of modern blocks. This is not a problem of this power supply unit, but of the circuitry with group stabilization itself — they all behave the same way, and for good stability in such power supplies, the load must increase proportionally in all channels. If you have a power supply unit with similar circuitry and there is a large sag on the +12 V line, then you need to add something to the +5 V line and the sag will decrease.

The block efficiency test was carried out at a network voltage of about 230 V.

The efficiency of this unit is on the border of the 80 Plus standard, probably because we have high losses on the wires, and in our test only the main power cables were used. If you also connect cables with SATA and Molex, then the total losses on the wires would become a little lower, and the efficiency of the BJ would be a little higher, or perhaps the manufacturer simply declares the efficiency of the block without taking into account the wires.

The heat test of the unit components was carried out at an indoor air temperature of 18 °C, using a Scythe Kaze Master Pro panel with sensors attached to the main components of the unit, the unit was loaded at maximum power and operated until the temperatures stabilized. At the end of the test, the temperature readings were fixed, after that the cover of the block was removed, and the temperatures of other components were measured using a pyrometer. The test results are shown in the following photo of the block board:

At maximum load, the temperature of the components was not as high as for a budget unit, while the fan was not very noisy, at the level of fans in the stand. In a PC case, the temperature and noise of the unit will be higher, it will all depend on the location of the unit, the ventilation of the case and the temperature in the room.

Conclusions

The tested Chieftec APB-700B8 delivers the claimed power, but with stability there are nuances arising from not the most modern circuitry. Among the advantages, we can note a relatively low price and not very high temperatures with noise. Among the minuses are budget components, quality of wires, outdated circuitry and the input range of the power supply network is only 200-240 V. In general, the unit is almost no different from other budget Chieftec devices, price and availability will be decisive when choosing. You can also take old models, because their components and platform are approximately the same. The unit will be fine for low-cost systems with low power consumption, temperatures and noise will be quite low, voltage sags and skews will be relatively small, and the APFC will be able to operate normally at much lower input voltages than at maximum load.