by Oleg Artamonov
04/25/2005 | 03:45 PM
Like the previous PSU-related test session on our site, this one is about high-wattage power supplies, typically selling independently of system cases (the only exception is the Macropower MP360AR ver.2; you can meet it outside the system case, but it will be an OEM version, without the colorful box, installation guide and other nice trifles). Two out the tested seven products present the highest interest: the models from A.C. Ryan and OCZ Technology allow removing or installing their output connectors as necessary, thus freeing your computer from unneeded cables which can be a big nuisance in small system cases.
But before getting to the tests, I want to clarify a few theoretical points that provoked questions from the readers of my previous articles. They concern the new 24-pin power connector on mainboards, PSUs with two +12V outputs, and the efficiency factor of power supplies.
As you know, the latest generations of graphics cards exhibit a very healthy appetite for power. They easily consume up to 60-70 watts, while the AGP slot cannot yield more than 40 watts, as stated in its specification. So, they started to put additional power connectors on top graphics card models; the connector was of the Molex type at first and then of a special 6-pin variety. The PCI Express slot that is now steadily replacing AGP permits consumption of up to 75 watts, so many PCI Express-compliant graphics cards don’t need an additional power connector on board. But when you use such a card, the mainboard’s power connector – already heavily loaded – becomes the weak spot.
Moreover, modern graphics cards consume ever more power from the +12V rail, and there’s only one pin in the mainboard’s power connector for this voltage. As a result, this pin sometimes overheats and gets burned out with the ensuing consequences for the whole system. As a way to solve this problem, mainboards with PCI Express now come with a new 24-pin power connector (incidentally, it is exactly the same as the EPS connector employed in power supplies for servers) instead of the older 20-pin one. The four added pins are “ground”, +5V, +12V and +3.3V and this allows to power up PCI Express graphics cards without any problems. You can regard the new connector as the old one with four new pins attached to its side, that’s why new mainboards are compatible with older power supplies and vice versa: the new pins of the connector are just left open in this case.
It’s clear that an adapter from the 20-pin plug of the power supply to the mainboard’s 24-pin connector cannot solve the problem of burned-out pins: just different pins are going to overheat. The only proper solution is the use of a power supply with a native 24-pin connector. It’s of course possible to make an adapter that would attach to two PSU connectors – to the main ATX connector and to a Molex connector where the adapter could take two of the missing lines, that is +5V and +12V (the +3.3V line wouldn’t be available even here, though). This would settle the matter, though partly, but I haven’t yet met such adapters. All solutions I know don’t use any additional connectors.
Besides graphics cards, the power consumption of central processors is also growing ever higher. And CPUs load the +12V power rail, too. It seems they can just increase the wattage of the power supply together with the maximum allowable current on the +12V rail, and that’s all. It’s even simpler than increasing the current on the +5V rail (and allowable currents on this rail exceed 40 amperes on some PSU models), but the developers confront the safety regulation EN 60950 which says that the maximum power on user-accessible contacts (the outputs of a power supply belong here, of course) cannot be higher than 240 volt-amperes. Thus, it’s impossible to provide a current above 20 amperes on the +12V output without violating this regulation.
How did the engineers solve this problem? They wind the secondary +12V winding of the transformer depending on the required load current; let’s suppose it is 30 amperes. The rectifier’s diodes and the throttles are installed depending on this total current, too. But after the rectifier the +12V rail is split into two wires, each of which has independent protection against current overload with the threshold set to, say, 15 amperes. This slight redesign of the power supply (the protection circuitry is quite simple) allows achieving a higher output capacity on the +12V rail (for example, one wire feeds up to 15 amperes to the CPU, while the other powers up the graphics card, hard drives, etc that can consume up to 15 amperes in total, too; the total consumption of all devices on the +12V rail can reach 30 amperes in this case), meanwhile complying with the safety regulations because the protection works as soon as there are more than 12 amperes on any of these two PSU outputs.
Summarizing this briefly, I can put it like that: new power supply models have one internal high-current +12V source that looks like two sources to the user. The load characteristic of the PSU depends on the capabilities of that internal source rather than on the artificial limitations on the outputs the user has access to. You will see below, in the description of the power supply from CoolerMaster, why I put so much emphasis on this fact.
The last thing to be discussed in this introduction is the efficiency factor of power supplies. Among other parameters we do measure the efficiency in our tests, so our readers ask what exactly efficiency value should be considered good. Besides the obvious rule “higher efficiency is better”, there are requirements set down by the industry standards. According to the latest specification, ATX12V 2.0, the power supply is required to have an efficiency of no less than 70% at full and half load and no less than 60% at a load that equals 20% of the PSU’s capacity (these requirements had been somewhat less strict in the previous version of the standard). The recommended efficiency values are 75% at full load, 80% at half load, and 68% at 20-percent load. All manufactured power supplies meet the requirements, but not all of them meet the recommendations. Not all units can boast an efficiency of 80% (but some models can exceed this value considerably).
I would also like to note that I perform my tests at 220 volts on the PSU’s input and this voltage is maintained quite accurately. If the input voltage goes down or if you switch to an 110V power grid, the PSU’s efficiency factor will decrease due to the higher loss in its high-voltage circuitry. But the difference is going to be small, a few percent at most, because the main loss falls on the PSU’s low-voltage circuitry (more exactly, on the rectifier’s diode assemblages) that don’t depend on the PSU’s input voltage at all.
The power supply from A.C. Ryan has a dark – but not exactly black – steel case. The gloss on the surface is due to a layer of scratch-sensitive lacquer, so you should be careful during the installation if you want to preserve the original appearance of the PSU. This unit is equipped with two fans: an 80mm fan on the rear panel and a 90mm fan on the bottom panel. Somewhat against the current fashion, there’s no highlighting of the fans.
The manufacturer positions this PSU as “the open platform for PSU customization and modding” but doesn’t explicitly say where this openness shows up. The PSU differs from the competitor products only in giving you the opportunity of detaching and replacing all the power cables which are plugged into the appropriate connectors on the PSU’s rear panel:
The four rightmost connectors will feed juice to your hard, optical and floppy drives. The two small 3-pin connectors in the top left corner are going to give power to additional fans (it’s always +12V on these connectors, so the fans will work at their full speed unless you use some third-party speed regulator). The remaining two connectors are for cables with a mainboard connector and a 4-pin ATX12V connector. For some reason there are no +3.3V connectors for SATA drives here – the four available connectors offer only +5V and +12V. Yes, there are no +3.3V drives as yet, but there’s no doubt they will appear sooner or later.
Otherwise there’s nothing extraordinary in the design of this PSU, so I have to come to the conclusion that this “open platform” should only be considered different because it permits to use cables of any color. Yes, A.C. Ryan offers kits consisting of fan grids, screws and other accessories of various colors, but you can easily install them into any other PSU because they are perfectly standard.
The internal design of the PSU bears a strong resemblance to the Antec TruePower unit I discussed in my previous roundup (for details see our article called Powerful Power Supply Units Roundup), save for another design of the active PFC unit (here, it is placed on a vertical card near the right wall of the case). In any other respect this PSU is identical to the Antec, including the independent magamp-based regulation of all the output voltages.
Besides that, there is now a card with the above-mentioned output connectors. It is a piece of handicraft in the bad sense of the word – the card was evidently soldered manually and not very neatly.
Thus PSU comes with a basic selection of cables:
Take note that the PSU doesn’t have a separate power connector for the graphics card, just like it doesn’t have 3-volt connectors for SATA drives, and you have to use the adapter if your graphics card has a 6-pin power connector. Then, the enclosed power cable supports mainboards with a 20-pin connector, while modern PCI Express mainboards have a 24-pin connector (and the PSU itself has a 24-pin connector). So, if you’re going to use a power-hungry graphics card without an additional power connector (such devices are already selling), it would be wise to purchase another cable from A.C. Ryan – the one with 24-pin plugs on both of its ends. If your card consumes little power or is equipped with its own power connector, you can use the cable you find enclosed with the PSU – it is compatible with the connector on new mainboards both mechanically and electrically, but four contacts in the mainboard’s connector are going to remain open with it.
Generally speaking, the detachable cables are an indisputable advantage of this PSU, especially for the owners of small system cases where it is sometimes difficult to place all the PSU cables without hindering airflows (and half of these cables are never even used at all). This problem does not exist with the Ryanpower2 – you just don’t attach the unneeded cables.
Now let’s test this unit. The table above sets the maximum load currents of the Ryanpower2 against the typical 400W PSU as described by the ATX12V 2.0 standard (by the way, this is the highest-wattage unit described there). As you see, the PSU from A.C. Ryan has a lower load current on the +12V rail despite its higher total wattage.
The diagrams of the cross-load characteristics of this PSU look perfect, but it’s natural considering the dedicated regulation of the output voltages employed in it.
The oscillogram of the output voltages at max load looks perfect, too. The peaks of the pulsations are much below the acceptable limits (which are 50 millivolts for the +5V rail and 120 millivolts for the +12V rail). I didn’t notice any low-frequency pulsation at all (particularly, at 100Hz) in the oscillograms.
The PSU flexibly controls the speed of its fans in a wide range depending on the load. The acoustic parameters of the device are rather average – the fans aren’t practically audible at small loads but speed up to hearable rotations-per-minute as the PSU is warming up.
The power factor of the PSU is very close to 1, thanks to the active PFC. Its efficiency is average, no more than 80% at best, but complying with the requirements. By the way, I measured the efficiency at 220V input voltage. If the voltage in the power grid is lower (the PSU supports input voltages from 90 to 240V without manual switching), the efficiency is going to be lower due to a higher loss in the PFC unit.
So, the A.C. Ryan Ryanpower2 ACR-PS2094 model is a good power supply on the whole. It boasts an excellent stability of its voltages, a minimal level of the pulsations, and quiet operation, while its detachable cables are very handy. Alas, but it is not free from certain defects. The quality of soldering isn’t high in some spots. The unit doesn’t have normal power outputs for SATA drives and a separate 6-pin connector for the graphics card, and the enclosed mainboard cable has only a 20-pin plug. Yet anyway, the advantages of this PSU outweigh its defects for me.
The power supply from CoolerMaster is made in a black case with a single 12cm fan (unlike the Ryanpower2, this device is not lacquered but actually painted black). The fan is highlighted with blue LEDs at work.
High build quality is evident inside. Somewhat unusual is the use of an ML4800CP chip from Fairchild in the regulator. This chip combines both an active PFC controller and a PWM controller of the PSU’s main regulator, but many manufacturers prefer to use two different chips instead.
The design of this PSU is quite traditional otherwise, save for a special tool, an indicator of the consumed power, you receive as an accessory. This tool is in fact an ordinary micro-ammeter and can work only with the given PSU that has an appropriate connector (the PSU has a special independent circuit with a current transformer). The tool measures the power consumed by the whole system from the power grid. It is the sum of the power consumed by the system components from the PSU and of the loss in the PSU itself (in other words, this tool would measure more than 500 watts when this 450W unit was under full load).
Well, this power-meter is more like a mere decoration. I checked its showings to find them contradicting the reality: the higher the consumed power, the bigger the discrepancy is. The tool shows 50-70 watts less than it should when the PSU is under full load. So, even though it is going to look most strikingly on the system case, you shouldn’t rely on its showings too much.
Unlike the Ryanpower2, the unit from CoolerMaster is equipped with ordinary, non-detachable cables, but its Molex power plugs (for hard disk drives) are shaped in such a way as to make it easier to extract them from the devices (it’s sometimes difficult to take an ordinary smooth plug out of a drive because you have to apply some force). Unfortunately, these plugs are not compatible with graphics cards that receive additional power through a Molex connector unless this connector is located at the very edge of the card – the “ears” of the connector will hinder its plugging in, so you’ll have to use either an adapter or a sharp knife.
This power supply offers you seven Molex connectors for PATA hard drives and CD drives (these cables are 55cm long to the first plug and then 15cm more to each next plug), two mini-plugs for floppy drives, two power connectors for SATA drives (these cables are 57cm long to the first plug and 20cm to the second one; unlike with the Ryanpower2, the connectors have all the necessary voltages, including +3.3V), a 24-pin mainboard connector (almost 60cm long), an ATX12V connector (on a 75cm cable), and a connector for the above-mentioned power indicator (on a 75cm cable). Thus, these cables should suffice even for a very serious computer assembled in a full-size system case (for example, in a CoolerMaster Stacker). The wires of the mainboard’s power cable are packed into a mesh pipe; the other cables are tied up with nylon straps.
This PSU is declared to comply with the ATX12V 2.0 standard, but you can see from the table below that it only corresponds to 300W units as concerns the load current on the +12V rail (that is, 22 amperes in total) whereas a 350W PSU should provide a combined current of 25 amperes.
The label on the PSU also declares the peak load currents, particularly 18amp and 16amp for the two +12V outputs (denoted as 12V1 and 12V2). This seems to surpass the capabilities of all typical power supplies described in the ATX12V 2.0 standard (the most powerful of them is a 400W unit), but it’s not quite so. The typical ATX12V 2.0 power supply must be capable of maintaining the maximum current on both 12V outputs simultaneously (in other words, the 400W unit should be ready to yield up to 14amp+15amp=29amp on its +12V rail), but the RS-450-ACLY limits the combined load current on the +12V at 22amp.
I already said at the beginning of this article that the splitting of the +12V rail in two was only required to comply with the safety regulations that limited the maximum power on user-accessible rails. And it was implemented by means of two independent current sensors. But there is actually only one +12V rail inside the power supply and it is its load capacity that limits the combined load current on the +12V1 and +12V2 outputs. Thus, the typical 400W ATX12V 2.0 power supply has a load capacity of 29 amperes on this rail whereas the CoolerMaster RS-450-ACLY – only 22 amperes. The declaration of the peak load currents of 18amp and 16amp only means that the over-current protection of this PSU is set to these values. They could have declared them as sustained rather than peak currents with the reservation that their sum should always be lower than 22 amperes.
The cross-load characteristics of the PSU indicate that it oriented towards modern systems that are mostly powered by the +12V rail, notwithstanding my complaints about the load capacity of this rail expressed above. Unlike the previous model, this PSU is not equipped with independent regulation of the output voltages. You could have guessed that by the lack of magnet amplifier coils on its output.
The output ripple is rather strong here, being very close to the allowable limit of 120 millivolts on the +12V rail and being even higher than the norm (the norm is 50 millivolts or one grade of the oscillogram) on the +5V rail. When the load power was reduced to 350 watts, the pulsation diminished to 40 millivolts on the +5V rail and to 60 millivolts on the +12V rail, which is acceptable.
The fan rotation speed is quite effectively controlled, so the PSU is almost perfectly silent under small loads. When the load is higher, the fan speeds up to 2000rpm, and the noise from the air stream is perceptible. By the way, the speed of the fan is adjusted depending on the load on the PSU rather than on its temperature.
The efficiency of the PSU is high – over 80% in all the range of loads almost. The power factor is 2-3% lower than that of the competitors, but this small difference isn’t very important in practice. Like the Ryanpower2, the unit uses its active PFC to support all input voltages from 90 to 240 volts without manual switching.
So, the CoolerMaster RealPower RS-450-ACLY is a quality product. Its advantages include active PFC, stable voltages (despite its classic design without auxiliary regulators), a quiet 12cm fan with highlighting, and long cables with numerous plugs suitable for any system case. But the manufacturer is not quite honest talking about the compliance of this PSU with the ATX12V 2.0 standard. The load characteristic of this PSU on the +12V rail is only comparable with 300W models as described by this standard. The RS-450-ACLY will suit perfectly for modern midrange systems, but you may want to consider other PSUs, with higher currents on the +12V rail (they don’t necessarily have to have a higher wattage), if you’re building up some extreme configuration.
Power supplies manufactured by HEC (Hirolchi), and the Macropower MP360AR is among them, have already been tested on our site, but they have all been compliant with the ATX12V 1.2 standard, while the MP360AR is a typical representative of the new version of the standard, ATX12V 2.0. This PSU can ship on its own or in system cases manufactured by HEC/Compucase, for example in the Ascot 6AR2.
This PSU is designed in an absolutely standard way. It doesn’t differ from any other midrange ($40-50) power supply: a gray steel case and one 8cm fan without highlighting. It is also a typical solution internally – without power factor correction, without independent regulators on the output, without detachable cables…
The assembly quality is high, as with all products from HEC. The PSU has six power connectors for PATA drives and two mini-plugs for floppy drives and only one SATA power connector; the mainboard power connector is 24-pin, like the standard demands (an adapter for the 20-pin connector is enclosed with the PSU).
As you see, the power supply exactly conforms to the requirements of the ATX12V 2.0 standard to 350W units (the full wattage of the MP360AR is a little higher, 360W, as its name suggests). The combined load power on the +5V and +3.3V rails has diminished greatly since the previous versions of the standard. It is now only 130W, so this PSU won’t suit for powerful last-generation systems that heavily load the +5V rail (for example, a senior Socket A processor on a mainboard with a 5-volt VRM plus a RADEON 9800 PRO graphics card). I should note that our sample of the power supply (it was an off-the-shelf product rather than an engineering sample) didn’t have independent current limitations on the +12V outputs – the small additional card with the appropriate circuit had not been soldered in. It does not compromise the PSU’s functionality, so this solution is quite logical unless the manufacturer wants the particular product batch to comply with the EN 60950 safety standard.
The cross-load characteristic of the PSU confirms its optimization towards 12-volt-oriented systems. It maintains the load on this rail just admirably, up to the maximum current of 25 amperes. The reaction of the PSU to a high load on the +5V rail is quite typical – other midrange power supplies behave perfectly like that.
The pulsation on the +5V rail is practically zero. The pulsation on the +12V rail seems big in comparison, but it doesn’t exceed 70 millivolts even when the rail is under full load, while the acceptable value is 120 millivolts.
The rotational speed of the fan varies depending on the temperature (you can see a copper buckle with a thermo-resistor on the butt end of the left heatsink in the snapshot above), but the min and max fan speeds are both quite high, so this is not a noiseless PSU. The fan is rather quiet at small loads, but can become the loudest in your computer as the power supply warms up.
Thus, the Macropower MP360AR Ver.2 is a typical ATX12V 2.0 power supply. On the one hand, it has a small load capacity on the +5V rail, a relatively noisy fan and an unassuming exterior (if you set it against the units from A.C. Ryan and CoolerMaster). But on the other hand, it can boast a high quality of manufacture and an excellent load capacity on the +12V rail. Considering that it costs much less than the above-described models, the PSU from Macropower is going to be a good choice for a modern midrange computer system, but more suitable for office than for home applications – because of its noisiness.
Some people confuse MGE with the well-known manufacturer of uninterruptible power supplies MGE UPS Systems but the two companies have nothing in common save for the abbreviation in their names. Here, MGE is expanded into the not-quite-modest “Manufacturer of Great Electronics”. MGE is mostly known as a maker of various modder-friendly kits: system cases, transparent windows, highlight lamps and so on. We’ve got two PSUs selling under the MGE brand for our today’s tests.
The declared wattage of the first unit is 500 watts; it has a blue aluminum case with the connectors and even the braiding of the cables to match. Of course, both fans – an 80mm fan on the rear panel and a 120mm fan on the bottom panel – are highlighted with blue LEDs at work. Aluminum is a popular material among some manufacturers and many purchasers of expensive system cases and power supplies, but its functionality is questionable. Here, aluminum doesn’t help to cool the PSU (none of the heating components touch the aluminum cover, so it is not used as a heat-spreader), but it does worsen the electromagnetic compatibility of the PSU because aluminum is worse than steel in terms of screening low-frequency fields.
The fine black grids on the fans with the large chrome-plated letters “Vortec” look superb, and the transparent window on the side panel of the case allows you to watch the highlighting of the fans at work.
The second PSU, with a declared wattage of 600W, uses a black-and-red color scheme (if you don’t look at the power-on switch and the connector to the power grid). This case is made of aluminum, like the junior model. The fans are highlighted with red, and this PSU is a spectacular sight when working – the traditional blue highlighting has already become somewhat hackneyed.
The two PSUs from MGE are designed identically inside, so I only offer you a snapshot of the 600W model below.
The design is quite ordinary, without PFC (although the dimensions of the PSU and its PCB seem to provide for an installation of a passive PFC device at the rear part of the PSU) and without independent voltage regulation or any other remarkable features. The heatsinks are small, with moderate ribbing, and colored green – not quite typical for heatsinks. Two 820µF capacitors are placed on the PSU’s input; two more 3300µF capacitors are placed on its output +3.3V rail and one such capacitor is placed on each +12V and +5V rails. The capacitances are the same in the 500W and 600W models.
Despite its very high declared current on the +12V rail, the PSU is only equipped with a 20-pin mainboard connector and comes enclosed with an adapter for the 24-pin one. I can’t really explain this decision of the manufacturer. As I said above, this adapter is not just useless, but contradicts the very idea of transitioning to 24-pin connectors. So, it is very strange for me to see such cheap economy on connectors in a power supply the manufacturer positions as a top-end solution. The length of the ATX cables is 45cm.
The PSU offers two connectors for SATA drives (the cable is 45cm long to the first connector and 12cm more to the second one) and as many as eight Molex connectors but they are all put on two cables (45cm to the first connector and 15cm more to the next one). The connectors of the Vortec units are the same as the above-described CoolerMaster’s.
The 18AWG-section wires are hidden in braided pipes (including the cables for hard drives) of red and blue color, depending on the model.
The cross-load characteristics of the PSUs look identically, save for the higher wattage of the PSVO-600 (but its protection would come into play somewhat below the load power of 600W). The stability of the output voltages of these PSUs is rather average. It is high enough for the computer to work normally, but it is worse than the above-described competing models ensure. Well, if these PSUs had a slightly lower voltage on their +5V rail, then it would be better with high loads on the +12V rail (when the +5V voltage goes out of its acceptable range). But it’s a fact that the +5V voltage is set rather high on both of the tested PSUs.
The high-frequency ripple of the output voltages lies within the acceptable range. Its amplitude at full load isn’t higher than 30 millivolts for the +5V rail and 80 millivolts for the +12V rail. There’s also low-frequency pulsation at a frequency of 100Hz and with an amplitude of about 15mV on the +5V rail and about 40mV on the +12V rail. So, the PSU comes close to the allowable limit as concerns the sum of the pulsations at full load, but never exceeds it.
Besides automatic adjustment of the fan speed, the PSU also offers you a complementary manual control. The auto-adjustment sets the maximum speed, while you can only set the minimal one (that is, if you set this control to its lowest position, the fans will anyway work at their full speed at full load; but if you set the control to its maximum, the fans will be working at their full all the time). The two curves in the diagram below show the speed-load relation for the min and max positions of the manual control (I measured the speed of the bottom, 12cm fan, only).
Despite the low speeds of the fans at loads below 300W and the lowest position of the regulator, the power supply is not silent. The problem is with the nice-looking protective grids. Even a weak stream of air produces a distinct hissing sound when passing through them. Then, about 40% of the area of the 12cm fan is covered with celluloid film at the bottom, probably to optimize airflows inside the PSU. Besides increasing the resistance to the air stream and thus worsening the efficiency of ventilation, this film was not firmly fixed in our sample of the PSU and would sometimes rattle a little. My removing the protective grids and fastening the film rendered the PSU much quieter.
Generally speaking, I don’t quite understand the purpose of two fans. PSUs from other manufacturers suggest that one big fan is quite enough for cooling.
The efficiency factor of this PSU is average. It is close to 80% but never reaches it. The power factor is quite typical for a PFC-free device.
Overall, the power supplies from MGE proved to be quite good, but they have a number of small and easily correctable defects (that’s why I don’t understand why the manufacturer didn’t correct them). First of all, they lack a 24-pin mainboard connector, while the grids on the fans make the PSU much noisier. Without these annoying trifles I wouldn’t have any complaints about the Vortec series units whatsoever.
I already reviewed power supplies from OCZ Technology’s PowerStream series in my previous article (for details please see our article called Powerful Power Supply Units Roundup), but this model belongs to the ModStream series. Such units differ from PowerStream ones in having a different distribution of loads (the ModStream products offer higher currents on the +5V rail but lower on the +12V rail). They also lack the control over the output voltages but are equipped with a 12cm fan and – the most important feature – have removable cables, like the above-described Ryanpower2 has.
Unlike with the PSU from A.C. Ryan, the mainboard and ATX12V connectors are not detachable here for quite an obvious reason – the user is unlikely to build a computer without a mainboard, while any extra connector in a circuit increases the loss because of its non-zero resistance and, accordingly, worsens the stability of the output voltages. OCZ solves the problem of compatibility with 20- and 24-pin mainboard connectors in an elegant way:
You just detach the four extra pins when you connect the PSU to a mainboard with a 20-pin power connector. So, you don’t have to use any adapters or, like with the model from A.C. Ryan, to purchase a new cable.
They also solved the problem about the connectors for the graphics card and SATA drives that I mentioned above when talking about the Ryanpower2. The PSU from OCZ has two 6-pin connectors one of which connects to a power cable for SATA drives and another to the graphics card’s additional power connector. The disadvantage of this solution is that the connectors only differ in color, so you have to be cautious when connecting everything because a wrong connection may damage your drives. Of course, it would be better for the connectors to differ in shape or in the position of the keys (that is, the splayed angles) so that the user didn’t have a chance to confuse them.
The following is supplied with the PSU:
The cables are hidden into a flexible transparent pipe with a braided screen (but the effect of this screening is close to zero because the screen is not even connected to the ground wire) and plastic braiding that shines in ultraviolet inside. Modders are likely to appreciate such shining cables, but on the other hand, these multi-layer cables turn to be stiff and unhandy.
The power cable for the graphics card carries a simple LC filter made out of a ferrite bead and four capacitors. This filter is going to suppress high-frequency interference for a better operation of the graphics card.
Internally this power supply resembles the P4-450W from be quiet! that I also reviewed in my previous article (for details please see our article called Powerful Power Supply Units Roundup). To tell you the truth, it is one and the same company, ToPower, that manufactures PSUs for both OCZ and be quiet!, so there’s nothing miraculous in these two models from two different brands being so much like each other.
The build quality of this PSU is high. I would like to draw your attention to the panel with the connectors on the front panel of the PSU which gave me cause for complaints in the Ryanpower2 section of this review. This panel is as neat as everything else in the OCZ ModStream: the wires are not soldered to this panel, but are attached with neat clips, sleeved in transparent insulating pipes (the clip of the +5V rail is visible in the snapshot above).
The load capacity of the +12V rail is somewhat lower than that of the typical 400W ATX12V 2.0 PSU, but the allowable load on the +5V rail is impressive. The purpose of such a high load capacity remains obscure to me because modern systems rely mostly on the +12V rail and the capabilities of this PSU will most likely remain underused.
Alas, the kinship with the PSU from be quiet! shows up not only in the design, but also in the characteristics. The BQT P4-450W couldn’t boast stable output voltages, and the ModStream OCZ-520 cannot, either. I think it’s already clear that this PSU doesn’t have any auxiliary regulators, like the models of the PowerStream series.
The pulsation of the output voltages might have been better, too. The amplitude is about 70 millivolts on the +12V rail, which is lower than the acceptable limit. But on the +5V rail the pulsation approaches the limit: the high-frequency pulsation of 50 millivolts combines with the low-frequency 100Hz pulsation with an amplitude of about 10 millivolts. This is observed only at full load (the oscillogram is taken at 500W load). When the load is lower, the low-frequency pulsation vanishes, while the level of the high-frequency pulsation diminishes (for example, to 25 millivolts on both rails at 300W load).
The PSU works very quietly, and its fans don’t practically change their speed. The speed is only increased at highest loads, but only to 1250rpm at maximum. And there are no overheat-related problems even at full load.
It is the excellent efficiency of 89% at full load that helps the PSU to keep the fan speed low. The power factor is about 0.7 because this PSU has no power factor correction circuitry (we tested the American version of the PSU; the European version is equipped with PFC).
So, the power supply from OCZ leaves a most pleasant impression with its highest build quality, attention to various details, detachable cables and quiet operation. I think that the OCZ ModStream deserves to be called an “open platform for modding” more than the Ryanpower2 from A.C. Ryan (these two PSUs can be considered competing products, at least in terms of functionality and target user groups). The latter company should follow the example and improve the quality of manufacture of its PSUs. On the other hand, the OCZ ModStream is inferior to the Ryanpower2 in the actual parameters, first of all in the stability of the output voltages. So, OCZ should also make some improvements – to the electronics of the PSU. I think that a PSU of the PowerStream class, even without the manual voltage control but with the functionality of the ModStream series, would be a highly successful solution. So, if you’re going to buy a power supply with detachable wires – and this is really a helpful feature – then you’re facing a difficult choice: the Ryanpower2 is more stable, while the OCZ ModStream is better in terms of build quality and user-friendly design. If you’re going for a PSU from OCZ Technology, I would recommend to you to pay attention to models from the PowerStream series instead.
This power supply differs strikingly from any other model I have reviewed so far on our site through the total lack of any fans. This unit is cooled by means of external heatsinks that consist of thin copper plates; heat is being transferred to these heatsinks via heat pipes.
The second heatsink is located on the front panel of the PSU (that is, it is going to be inside the system case in an assembled computer) and is much smaller. It cools the switching transistors of the inverter, while the external heatsink cools the output diode assemblages that heat up badly. The number of heat pipes differs, too. There are two pipes on the internal heatsink and three on the external one.
I also found rather large T-shaped heatsinks inside the unit, but I doubt their efficiency. They just warm up the air inside the PSU. It’s also not quite clear why the manufacturer didn’t establish heat contact between these heatsinks and the cover of the PSU – the aluminum cover might have added well to the total dissipation area. The power transformer and the group regulation throttle are not cooled at all.
The text on the PCB says who makes these units for Thermaltake. It is the same Sirtec that also manufactures Thermaltake’s traditional fan-cooled PSUs.
Apart from cooling, the unit differs but slightly from the more traditional models. It has a Power-On switch (the switch is highlighted with red when turned on – should we considered it a “modding feature”?), doesn’t have any power-factor-correction circuits, and offers you a switch to choose the input voltage (110/220 volts). The PSU has two connectors for SATA drives (the cable is 50cm long to the first connector and 15cm more to the second one) and as many as nine connectors for PATA devices (three cables, 50+15+15cm each), plus an ATX12V and a 20-pin ATX connector (both on 50cm-long cables).
The load characteristics of this model are rather modest. Despite the total wattage of 300 watts it is closer to 250W units in the declared allowable currents.
The unit sustains the full load well, if not ideally. But this load isn’t very high in comparison with other units of the same wattage. Note also that the +5V voltage is rather high, being on the limit of the acceptable range.
There is no high-frequency pulsation of the output voltages under 300W load. You can only note minor spikes on the +12V rail with an amplitude of 20 millivolts at the moments when the inverter’s transistors are switched over. Low-frequency ripple can also be observed at 100Hz – its amplitude is 20mV for the +5V rail and 30mV for the +12V rail, and that is far below the acceptable limit.
The efficiency factor of this PSU is 85%. This is high, as is necessary for a fanless unit. The power factor is rather low, less than 0.7.
Thus, the Thermaltake PurePower W0029 may be an interesting option for people who are trying to build an absolutely noiseless, although not very powerful computer. The PSU’s load capacity is only comparable to 250W classic models as concerns the allowable currents (the PSU works fine, without overheating, at full load, but in order to create this full load in a real computer it is necessary that the load be evenly distributed among all of the PSU’s rails). Then, this power supply is rather expensive. It costs more than many higher-wattage units from Antec, OCZ, Enermax and others which are practically noiseless, too, at small loads thanks to their quiet fans and control over the fan speed. All in all, the Thermaltake W0029 is a very special product, and I’m inclined to regard it as a fashionable device rather than a necessity.
So, you have seen seven power supply units in this roundup that differ in their wattage and functional features. I want to single out the A.C. Ryan Ryanpower2 and OCZ ModStream units first. They are curious for their detachable cables. That is, you can use only those cables that your devices are actually attached to. Alas, both these units have defects, although different ones. The Ryanpower2 has excellent parameters but its mounting quality could have been better, there is a limited number of cables supplied with it (short IDE cables, and the 24-pin connector for the mainboard is missing), and it cannot power up 3-volt SATA drives. The ModStream, on the contrary, boasts highest manufacturing quality and meticulous design, but its actual parameters are rather mediocre. One might almost wish there were a hybrid of these two units, with the electronics from the Ryanpower2 and the exterior from the ModStream…
The power supply from CoolerMaster doesn’t try to surprise the user with any outstanding innovations save for the measurer of the consumed power. But this is more a decorative element rather than a really useful tool. Otherwise this is a quality, beautiful, handy and quiet power supply that’s going to suit for almost any midrange computer. If your computer has a high-power configuration, take note that CoolerMaster is a bit cunning with respect to the declared load on the +12v outputs – this parameter of the PSU is only comparable to 300W ATX12V 2.0 models.
HEC’s Macropower MP360AR Ver.2 unit is a typical workhorse. It is a rather inexpensive PSU without a shiny case or highlighted fans and wires or any other beautiful but usually useless embellishments, but it proved to have excellent load capacity and good build quality. This PSU will suit well for a powerful office computer or other such applications. If you’re planning to buy it for your home PC, make sure beforehand that its fan isn’t too noisy for you. Also among the PSU’s drawbacks is the availability of only one power connector for SATA drives.
The two units from MGE performed well enough in my tests. They have a very high load capacity and ensure an acceptable stability of the output voltages. Alas, they are not free from blame, either. The fine protective grid on the fans increases the noise from the power supply considerably. The characteristics of the units are close to the ATX12V 2.0 standard, but they are equipped with a 20-pin connector for some reason. Then, I don’t also understand the purpose of the second 80mm fan which adds a lot to the noise. If these defects don’t frighten you, then the PSUs from MGE may be your choice.
The last was the fanless PSU model from Thermaltake. It is a special model for a specific user group: the price of the PSU is high, but its load characteristics are unpretentious. The noise from a high-quality PSU with an ordinary fan is usually not too irritating at small loads, especially considering that the PSU is not the main source of noise in a modern computer. But anyway, if you do want to buy a fanless PSU for a not-very-advanced computer, do consider the PurePower W0029 from Thermaltake.