550W Roundup: Three PSUs at Different Prices
by Martin Kaffei on October 28, 2010 4:00 PM ESTYou Get what You Pay for!
Perhaps our conclusion should also include the warning: caveat emptor (let the buyer beware), because it's clear that some PSUs should never see retail shelves. The Techsolo Black Mamba 550W belongs in this classification if you want a 550W PSU; in fact, the Techsolo is a ~250 watts power supply without an acurate working power factor correction—not to mention the fact that the efficiency is always below 75%.
If you just need a cheap power supply that will function at loads under 250W, perhaps budget products will fit the bill, but we'd suggest looking elsewhere as long-term power costs (and the risk of damaged components) makes this a gamble at best. In our case, the PSU died at 50% load, and there are no real safety functions—not even an MOV in the entrance. Both main switches are undersized and exploded. The protective ground wire is connected to a mounting screw of the main PCB. Again: Bad idea!
While the product looks decent, Techsolo offers nothing more than a few cables ties, one PCIe connector, and very short cables (40-45cm for the main cables). The cable sleeving is a nice extra, but we'd rather have the cost put elsewhere. Nothing else even suggests that this is a quality power supply. The noise for example is always over 22dBA and reached up to 28dBA at "50%" load. Another problem is the noisy PFC-choke, which is really loud and emits a high-pitched squeal. Passive PFC is always a noisemaker. In any case, Antec and OCZ are quieter and better, so just give this one a pass if you happen to be in Europe.
The OCZ Fatal1ty 550W is no superhero on the market, failing to stand out from the crowd. Your system won't run faster, and we have to wonder how much Fatal1ty gets paid for the use of his name—and if it even helps with increased sales, considering he's been quiet of late. This is a minor upgrade of the older OCZ ModXStream Pro 500W, with slightly improved internal devices, but that's about it. The EMI filtering has most necessary components, excluding an MOV. OCZ has switched to Teapo (LH, SC), after the Chinese caps weren't helpful for their public image. We can't say that they have the best characteristics, but they do a good job in power supplies. Now there are more protective foils against short circuits as well. This progress is absolutely welcome.
The downside of the PSU is that the ripple and noise results are very high for an "actual" product. More than 70mV on +12V and about 41mV on +5V can be measured. The efficieny is low at 10% load and not to remarkable at 50% load. Otherwise the voltage regulation is satisfying (never below -2% on all rails) and the power factor is high (above 0.840). The 45cm 24-pin cable should be long enough for small and mid-size cases, while the 4-pin and 4+4-pin ATX12V cables for the CPU could be much longer; we wouldn't recommend this PSU for use in cases where the PSU is at the bottom. Six SATA and HDD connectors are provided and there is a floppy connector at 90cm. The red cable sleeving is cheap but at least it's something.
What about the noise level? It's not bad at all. At low load there is no more than 18-19dBA. Even at full load the fan stays below 30dBA. Overall it is as quiet as the Antec TruePower New, but takes a small lead at high loads. This isn't a bad power supply, and it's reasonably efficient, but the high amount of ripple on the various rails is a concern for long-term stability, particularly if you're actually running this on a higher performance gaming system. Outside of noise levels the Antec is a far better PSU; you just have to decide if it's worth the additional cost.
When we look at the Antec TruePower New 550W, we can see good capacitors (Nippon Chemi-Con), many components for filtering in the entrance including an MOV, and well dimensioned +12V rails. The TruePower New has 55cm connectors for graphics cards and a 50cm 24-pin cable. The cable sleeving is again done rather cheaply, but the number of peripheral connectors gives no occasion for criticism. There is only one real problem: with a limited number of plugs you can't connect all the modular cables—even though this 550W unit should easily be able to handle such a configuration.
As for performance and quality, all the results from our tests are very impressive. The voltage is always over or near the ideal value and there is hardly any ripple and noise on any of the rails: less than 35mV on +12V, and less than 20mV on +3.3V and +5V. Up to 87% efficiency and a high power factor (above 0.850) is the primary reason for the cost; such results can only be reached with high quality devices, and those devices cost more than lesser offerings. In every case, Antec performs better than the OCZ and comes out the clear leader of this trio. And if you've ever wondered what the difference is between "budget" power supplies like the Techsolo and quality offerings like the Antec, hopefully we've managed to shed some light on the discussion.
So what's the overall conclusion? The gap between OCZ and Antec is far shorter than the gulf between OCZ and Techsolo, but you can definitely see why the Antec costs more. Long-term, the increased efficiency of the OCZ will easily pay for itself relative to a budget PSU, though we can't say the same for the Antec vs. OCZ. However, the Antec has much better characteristics elsewhere, and high ripple can put extra stress on your PC's components making the Antec a better choice for enthusiasts and anyone looking for maximum reliability. Is it worth the added cost? That's up to you, but we'd recommend spending as much on your PSU as you can afford; there's no reason to pair quality components (i.e. your motherboard, CPU, RAM, GPU, and SSD) with substandard power supplies. As our results here show, you clearly get what you pay for—though we'd recommend reading reviews to help choose between similarly priced, similarly specced PSUs.
Facebook
Twitter
RSS
82 Comments
View All Comments
wintermute000 - Sunday, October 31, 2010 - link
after a few psu-go-boom incidents I went seasonic or thermaltake, no exceptionsnot one problem since.
those stats look shocking compared to your mid-range seasonic/thermaltake product, but like the article says you get what you pay for. Never cheap out on your PSU if you run 24/7 or gaming cards
jazzisjazz - Sunday, October 31, 2010 - link
Techsolo Black Mamba STP-550, representative of many budget power supplies1 Why am I reading about a psu that I have to be in Italy, Germany, or Poland to buy?
2 Does sells mean retails for or commonly able to purchase at online or in a local store?
3 Nothing said about types of certification that would have some useful value to the shopper.
4 Time spent looking for things to poke fun at would be better spent informing the reader.
5 Why not tell us what makes it low end and outdated ?
6 No information on value of 80 plus certification.
7 Red mist of power supplies ??? meaning?
8 few changes in the details - are they unimportant?
9 modular cables, which is a nice feature for the price - would the money spent bringing us
this feature be better spent on higher quality components
10 The most expensive but potentially best - let the results bear it out.
11 another $24 premium over the OCZ and over twice the cost of the Techsolo - fluff
12 No mention as to why Japanese capacitors,DC to DC converters, and PWM fans might be better.
13 The best-case hope for the Techsolo Black Mamba is that we'll get an "ultra silent" snake that can deliver the full rated power. I wouldn't bet on it, but more than one user has been tempted into going the cheap route on a PSU. At worst, the Black Mamba is a noisy blindworm that dies at 50% load, but most likely it will land somewhere in between those extremes. - more fluff.
14 (already mentioned and super-duper - more useless poking fun
15 the PSU actually looks decent - you do realize you are talking about a budget peice where hopefully the manufacturer has concentrated the most effort on quality design and components right?
16 even cheap cable sleeving is a bump up over unsleeved cables, after all - does it come at the expense of better quality components?
17 Shenzhen Xin Wang Electronics Co., LTD. makes the fan, one of those nameless (despite the length of their name) Chinese companies.- you put it across like you know it is poor quality, do you know that?
18 Will it be silent, and can it keep the PSU cool? We'll find out in a minute - fluff
19 protective ground wire is connected to a mounting screw of the main PCB—bad idea - how should it be different?
20 missing a MOV - for what?
21 Can you see the wire cross-section of phase and neutral conductor? It doesn't look like 550W - because?
22 so stay away from this product.- any attempt to find out if the unit tested was faulty in manufacturing?
I give up. I keep hoping that some reviewer somewhere is going to take the budget consumer seriously - - but yet again no
Phaedrus2129 - Sunday, October 31, 2010 - link
The answer is: Yes. We know it's crap. We know it's crap from the minute we pick it up, and we know even more the second it's opened up. Was this the best review ever? Hell no. But your questions belie an ignorance of low-end power supplies, what makes them shit, and assumes stupidity on the part of the author.jazzisjazz - Tuesday, November 2, 2010 - link
Isn't it the purpose of this site to provide useful information to the reader (ignorant or not). If I were ignorant about low end power supplies and..... but smart enough to come to AT for good info I would hope to come away well informed. No assumption about the author or offense intended just frustrated that my little bit of money to spend doesn't seem as important as I hope it would. Not everybody can spend $600 and up, even if we get it done for <$250 after rebates, refurbs, and finding ways to engineer better for $free we still want as much useful info as we can get here at AT.mattll - Wednesday, November 3, 2010 - link
No, i'll agree with the reader jazzisjazz...the author assumes an electrical engineering background on the part of the reader. I think his point is that he does lack knowledge, and the author's approach is not helpful for improvement.And the answer, for a review, is no, you don't know it's crap from the outset. It's a court case, you have to prove the point, not just wink at the reader and share inside jokes. That it could not handle full load is useful information; snide comments on the cable wrappings, or fan grill design, or name of the fan manufacturer, is irrelevant unless and until you can prove that it makes a difference.
How do you know it's crap from the minute you pick it up? You have to detail out that knowledge, because that's the information you are trying to impart to the reader.
I thought the reviewer justified objection #1 pretty well. It happened to be a cheapy found in his area; for obvious reasons, ultra-cheap parts are not likely to be internationally available, widely known and long-lasting brandnames.
The reviewer might have connected #3 to #6, but didn't.
I agree with #12--the reviewer assumes reader familiarity with these, and even given familiarity, there's a glib assumption of obvious benefit which may not be the case. I think you mentioned Antec's run-in with production problems, and how you've given them credit for getting past them. Why would anyone assume all the manufacturers from a given country are permanently suspect based on one (admittedly widespread) incident?
#10 speaks to the bias of the article: let the results speak for themselves, or at least try to maintain the appearance that you didn't just say "I like this Antec because I like Antec. And this other crap is crap because we just KNOW it's crap."
#19 I had to roll my eyes at that. Is there really a shock hazard here? Did it pass UL/CE listing? Sure, it may be good advice to a power supply designer not to do it that way, but is it actually a problem to the consumer? Do you have any proof of that?
#20 LOL. This is an example of where the author was really speaking to other electrical engineers. You can look up Metal-Oxide Varistors on wikipedia, but that won't help a non-expert much either. More importantly, it's an implementation detail, and as such, it doesn't matter. The (lazy) assumption is that it necessarily affects performance: focus on the performance. Will it be less likely to withstand a transient? Then say or measure that. Will it adversely affect the noise profile? Then say or measure that.
#21 Again, a glib assumption of insider knowledge. Question: is it really necessary for a computer builder, or even someone choosing a power supply, to be able to identify the phase and neutral conductors on a picture of the power supply, much less be able to approximate their gauge and estimate their current handling capability? No.
Look. A rule of thumb like this, to say those conductors look pretty slim given the power they are supposed to be able to handle, is mildly useful (about like weighing the power supply). But that's no substitute for actual performance, actual measurement. Prove that it actually results in an inferior product, rather than just insinuating.
Beenthere - Sunday, October 31, 2010 - link
Proper testing of PSUs eliminates the fanboism and shows which PSUs perform well. If you don't understand the test results, design differences or details, then have someone with proper technical expertise explain them to you so you can make an intelligent decision instead of declaring one PSU a clone of another when in fact they are not.The technical ignornace that passes for knowledge on the Internet is appalling.
HangFire - Monday, November 1, 2010 - link
Yes, I enjoy proper testing of P/S such as found on jonnyguru, hardocp and lately Hardware Secrets.I own a P3, an oscilloscope, and a 500A battery tester. I don't do proper testing, but I do RMA any and all cheap power supplies that don't make PFC, efficiency, cannot deliver rated 12V power, or do not invoke OCP or OPP as spec'ed. I haven't found a Ripple failure that didn't also blow out before their reaching non-existent OCP levels. That is pretty much the level of AT testing as well.
AT might get there some day. They should read a few dozen jonnyguru reviews to see how its really done. I do like the overcapacity tests AT does, jonnyguru refuses to do that but should, if for nothing else to verify OPP/OCP. Usually they test good power supplies that don't like about having OPP/OCP, but I find the cheap ones often lie about the presence of OPP/OCP, such as the CoolMax that took out a motherboard... the beginning of my journey into power supply testing.
MeanBruce - Sunday, October 31, 2010 - link
Selling horror should be illegal, instead it simply hurts the ignorant and uninformed. This is a slime market that I hope to never see again on Anandtech!HangFire - Monday, November 1, 2010 - link
I disagree. I think the major tech sites should blow up an under-spec P/S once a week until they have all been reviewed, sales drop, and they are all taken off the market. There is already plenty of competition in the good-enough-or-better space to take up the slack.Beenthere - Tuesday, November 2, 2010 - link
There is a lot of crap products in the marketplace. The only means to determine if a PSU can deliver proper electrical power and is of a quality design and construction is with a proper physical inspection and objective, scientific load test conducted by a very knowledgeable tech. Anything less is a waste of time including fanbois opinions. Opinions are worth what you pay for them.