Well that's funny.....the link at the end of this article took me to the discussion about the earlier Sparkle.

I'm sold on the other Sparkle..... guess I'll have to try this new model. Sure looks plain though.....

With regard to efficient power supplies I was very surprised by how efficient and quiet the unit is in the Dell Vostro 400 that I bought for my parents. I reported the data previously here but I think it’s worth repeating for anyone looking for a cheap mainstream pre-built PC.

Dell Vostro 400 – Intel E4500 2.2GHz, 2x1GB DDR2-667, Samsung HD250HJ 250GB SATA hard drive, Intel G33 chipset & IGP, stock Dell Power Supply:

Powered off – 1.5W
Standby – 2W
Idle – 46W
Orthos (Large, in-place FFTs) – 77.5W (VCore set to lowest value with RMClock)

Just about every other desktop I’ve tested consumes 4W when powered off and 4 or 5W at idle which is one reason why the Dell surprised me. Did you test how much a system with the Sparkle consumed when off or on standby?

Saving 2 to 3W in these situations may not sound much but for systems that are powered off 80 to 90% of the time (i.e. used 4.8 to 2.4 hours per day) this is equivalent to saving 10 to 22.5W whilst using the PC which doesn’t seem so insignificant.

Ideally of course you turn off all power to the PC when it’s not in use but that takes more effort.

Interestingly, the Dell Vostro 400 is NOT on the Energy Star list (XLS file download)... although this could be because the model is already discontinued? Dell currently has 7 models rated to idle <50W.

Overstock.com has the unit for sale at US$40. It doesn't appear all that hard to find in the US, at least.

I was seriously considering the FSP Zen for my HTPC, but this looks like a much better fit for the NSK2480.

It’s a current model but it’s generally only listed on the Dell Business sites so easily over looked – Vostro 400.
They have a Dell Precision 690 Workstation on that list which is a Dual Processor Xeon that uses FB-DIMMs so I’m not quite sure what that’s about.

I did read that but I wasn’t sure what you were measuring; was this test with the power supply installed in the test PC? I’ve never seen such a low standby figure for a desktop system and the consumption with the system off was strangely high. Maybe it’s a quirk of using a 110W power supply!

Standby simply means the PSU is plugged into the wall, with the "power switch" (if it has one) on. In this PSU, since there is no "power switch", it's just the power draw with the PSU plugged into the wall. It makes no difference whether the outputs are connected to a motherboard and components -- except for 5V standby, which powers USB devices.

I agree that this would be an excellent replacement PSU for the NSK series (namely, the NSK2480 and NSK3480), especially with a fan swap to a Nexus 80mm. It's quite a fitting, inexpensive upgrade to these low-cost cases for those who cannot be satisfied with the stock noise floor.

I agree, it would be a great match. Maybe the Antec P180-182-190 shouldn't have been mentioned in the review because the cables of this PSU are too short for those cases. They may even be too short for the Mini P180. Let's hope someone doesn't go out and buy this PSU for their P18X case after reading that paragraph.

I’m still not clear here as you say it makes no difference and then say ‘except ..’!
Did you actually test the power supply when connected to the PC when it was in standby mode (S3) and when it was shutdown? I ask because the figures below seem anomalous:

A 6.5W load for a PC that is shutdown is horribly inefficient whilst the 0.3W at standby is laptop territory. This is the opposite of what I’ve generally experienced and I imagine many people would shutdown a PC thinking it will consume less power than on standby which is the opposite of what you’ve experienced with this unit. I’ve found that the difference between the two states is generally 0.5 – 1W so I usually keep my system in standby.

I just looked at two other recent power supply reviews and they follow the same pattern of a very low standby reading and a surprisingly high “No Loadâ€

I was under the impression that the "No Load, PSU power On" figure was with the PSU plugged into the wall and actually on, for example through the "paperclip in the ATX plug" method. No components are plugged into the PSU, though.

Yes, that's correct, and there's a second measurement we make with the PSU plugged into ac and the power switch turned on, but the PSU not triggered on to run. The latter is PSU "standby" (basically your PC is off but still plugged into the wall) -- which is not the same as standby in the OS.

Thanks for the review.

Anyone know what the equivalant FSP model might be, in the UK? I had quick look on their website, but couldn't find a match.

Thanks for clarifying that.
Is there a close correlation between the ‘laboratory’ figures that you record for those two power states and the ‘real-world’ measurements of a PC in either Standby or Shutdown?
I don’t mean that the figures will be similar across both types of tests but that power supplies that show good results in your laboratory testing also show good results when connected to a PC.

I was recently very pleasantly surprised by how efficient the Dell Vostro 400 is at idle/Standby/Shutdown (see post above) which started me wondering about this. It’s good to see a company of Dell’s size using power efficient components in their basic desktop systems.

Generally, yes.

Our vidcard test platform uses an S12-600. On PSU "standby", it pulls 3W and 8VA. When the power switch on the PSU is OFF, then it pulls 0W and 4VA.

Our HS test platform uses a FSP Zen fanless PSU. On PSU "standby", it pulls 2W and 10VA. When the power switch on the PSU is OFF, then it pulls 0W and 7VA.

Obviously the PF at these levels do have some impact.

One thing this PSU has all over the various Enermax 82+ is that it is a lot cheaper! Something tells me the ~ 400W Enermaxs are probably more efficient at every wattage, but the additional cost is far more than you will make up in electricity savings in any reasonable amount of time. I like this one better than the SPI220LE purely because of the ATX form factor. The short cables are a bit of an Achilles' heel, though. Once you have to start buying extension cables just to use it, it starts to cut into the price advantage and convenience of the form factor. I swear Newegg used to have this model, or a very close relative, but they only seem to have its 350W cousin at the moment.

Its the old Prius vs Corola argument. You'd have to drive a prius more than 10 years or so for it to end up costing less because of current gas prices.

Its not just about how much you save though. Its also about making a statement, and supporting products being made that push the market in a positive direction.

The enermax is a better designed PSU, and is a much bigger step in the right direction than the sparkle.

I agree the Enermax is one generation of SMPS technology ahead of this Sparkle. But from an environmental point of view, there's always some advantage in being cheap: Fewer & shorter cables, sleeves, connectors, packaging, fancy bags for cables, etc.... these factors also count, don't you think?

But doesn't buying the Sparkle also send a useful message -- that the general public doesn't need >> 300W PSU? Just imagine how efficient an Enermax 82+ 250W PSU would be at 50W and lower. Hell, the same materials that make an 82+ 625W might produce 85% or better in a 250W model.

Thanks Mike. Are you referring to PC standby (Suspend to RAM) here?
I wonder if Standby (S3) and Hibernate differ in power draw?

The one thing missing here for me is how much the systems consume when powered down but with power still available to the system.

Exactly. Going back to the Prius vs Corolla argument, you can't simply only factor in the price difference of the vehicles when comparing the two, you also must compare everything related - environmental cost to manufacture (are you buying a used one or new one - in general, the lower the cost the lower the environmental cost when comparing similar products), lifetime emissions (the Prius' exhaust emissions are cleaner than the Corolla, so will produce less smog), etc.

Getting the big picture when comparing two products isn't simple, but in this case when comparing the basic Sparkle to the high end Enermax, if both function basically the same (and they do if you don't need more than 250w and all the fancy cables, though it is a bit noiser tha the Enermax), then the Sparkle will save you money up front and will also be better for the environment because it is also takes fewer resources to manufacture.

No, I tried to make that clear by stating "PSU standby" -- this term standby refers to when the system is ostensibly OFF -- but the power sw on the PSU is ON and it's plugged into the wall.

But power draw on standby from the OS is very similar, irrc.

I don't know what you mean by this. Seems to me it's all covered:

System powered off but sw on and ac plugged in -- "PSU standby"
System powered off and sw OFF and ac plugged in
System on standby from within OS

Ugh, ugly JPEG instead of nice PNG on first page

Would you please indulge me MikeC and confirm what the power ratings are for these two states, both of which have the power on at the wall and the switch on at the power supply:

1. System put into Standby.
2. System Shutdown.

If you're asking about this for when this PSU is actually inside a system, I have no data.

However, I need to correct previous info I posted above.

The AC power data from the PSU reviews is all measured with a $600 instrument that has accuracy down close to 0W, to one decimal place. That info is accurate.

The "real system AC power at system off" data I posted above were made with KillaWatt and PowerAngel meters -- which turn out to be not so accurate at very low readings. When I substituted the high accuracy AC meter to these systems, I obtained the same readings on system off as with "PSU standby" on the PSU test bench.

In other words, the AC power in Standby information in the reviews is equal to System Power Off with a real system (as long as devices are not remaining powered up via the USB ports).

System Standby draws very little more power than System Power Off -- maybe a watt. This may be depending on whether your systems goes to S1 or S3 standby. I will check this later when I have more time.

Qestion for experts: is a power supply which has high efficiency at 50W hard to design / expensive to make? Or is it just the law of volume / sale advertisement more watt = better?

I am not an expert but I believe it is similar to cars: most cars cost roughly the same amount to develop, but the profit margin on larger cars/SUVs is many, many times higher. most people would not buy a psu rated at 50, 100 or even 200W. so the market is very small, mainly related to mini-ITX. small volumes = no economies of scale.

however the picoPSU + AC brick combo can be highly efficient depending on the brick, so that shows it can be done.

Getting a bit OT here, but I really wish that there was more data available regarding the efficiency of 12v AC bricks - I'd like to get a picoPSU, but am worried that I'll get a brick with low efficiency since efficiency rating information on them is basically non-existant.

There IS indeed a database of AC/DC converters.... at Energy Star. The link on the right column of that page to... For Consumers : Product Lists : AC-DC Adapters PDF will pull up (or let you download) a 66-page list of adapters with measured efficiency. It's difficult to search, however; you might do better with the full adapter list in XLS format, which makes it easier to order the way you want.