PSU tier list 2.0

[Karadjgne]

Not the same circuits. PWR_OK covers the output stages, as in power has gone through the psu and is determined to be more than 10.8v. UVP is on the line side, so shuts down the psu if the line input is less than 112v or 215v, for example.

 

[turkey3_scratch]

Hmm that's not what I learned. Gabriel Torres says UVP and OVP are directly from monitoring the voltage outputs, not inputs, and provides a table for the usual trigger point values they are set at. http://www.hardwaresecrets.com/everything-you-need-to-know-about-power-supply-protections/3/

 

[turkey3_scratch]

I think this may be my new recommended cheaper budget unit:

PCPartPicker part list / Price breakdown by merchant

Power Supply: Fractal Design Tesla R2 500W 80+ Gold Certified ATX Power Supply ($43.98 @ Newegg)
Total With 10% Tax: $43.98
Total Without Tax: $39.99
Prices include shipping, taxes, and discounts when available
Generated by PCPartPicker 2016-04-27 22:58 EDT-0400

The 650W version performed quite well: http://www.hardwaresecrets.com/fractal-design-tesla-r2-650-w-power-supply-review/

I would just be hesitant to recommend the untested 500W one, but Jonnyguru tested the 1000W one and Hardwaresecrets the 650W one, and they both perform well. For $39.99, getting a PSU like this seems like a better option than the Antec VP-450. If I'm correct this even has a ball bearing fan.

 

[okcnaline]

CapXon and Teapo... Eh...

 

[Nuckles_56]

I'm looking to buy some really shady power supplies for dirt cheap. Anyone have any ideas of units I could get?

I've found a few 'quality' ones for you
http://www.amazon.com/Coolmax-CA-400-Silent-Cooling-Supply/dp/B0012E6XF4/ref=sr_1_6?s=pc&ie=UTF8&qid=1461814206&sr=1-6&refinements=p_36%3A1000-5000
http://www.amazon.com/iMicro-IM400W-ATX12V-Power-Supply/dp/B002HREGRM/ref=pd_sim_sbs_147_3?ie=UTF8&dpID=41ejQuRc9XL&dpSrc=sims&preST=_AC_UL160_SR160%2C160_&refRID=1RH9FZ020H7JJJMRB1Z8
http://www.amazon.com/Suntec-Desktop-Computer-Power-Supply/dp/B012JSDM1E/ref=pd_sim_sbs_147_5?ie=UTF8&dpID=510TCy-Dp7L&dpSrc=sims&preST=_AC_UL160_SR160%2C160_&refRID=1RH9FZ020H7JJJMRB1Z8
http://www.amazon.com/Logisys-Corp-240-Pin-Supply-PS480D2/dp/B008MUD9IK/ref=pd_sim_sbs_147_1?ie=UTF8&dpID=415DOMpB7dL&dpSrc=sims&preST=_AC_UL160_SR160%2C160_&refRID=1RH9FZ020H7JJJMRB1Z8
http://www.amazon.com/KENTEK-120mm-Black-Power-Supply/dp/B005GHMRUI/ref=pd_sim_sbs_147_8?ie=UTF8&dpID=41AuhbD4r3L&dpSrc=sims&preST=_AC_UL160_SR160%2C160_&refRID=0WHJGNP45PG1421BFEDB
http://www.amazon.com/Apevia-ATX-XG450W-450W-Power-Supply/dp/B00YNXJMI6/ref=pd_sim_sbs_147_2?ie=UTF8&dpID=41mog8aAvJL&dpSrc=sims&preST=_AC_UL160_SR160%2C160_&refRID=0WHJGNP45PG1421BFEDB
http://www.amazon.com/IN-WIN-Development-IP-S350T1-0-ATX12V-Supply/dp/B007I55JVK/ref=pd_sim_147_8?ie=UTF8&dpID=419yzYqvcoL&dpSrc=sims&preST=_AC_UL160_SR160%2C160_&refRID=0QC3NTQJ1XHM31G53QVT
http://www.amazon.com/Genuine-ATX0180P5WB-Bestec-Computer-Compatible/dp/B00YHNMC0C/ref=sr_1_5?s=pc&ie=UTF8&qid=1461814600&sr=1-5&refinements=p_36%3A1000-5000%2Cp_n_feature_keywords_two_browse-bin%3A6906982011|6906983011|6906984011|6906985011
I hope that is enough Turkey

 

[Darkbreeze]

I think this may be my new recommended cheaper budget unit:

PCPartPicker part list / Price breakdown by merchant

Power Supply: Fractal Design Tesla R2 500W 80+ Gold Certified ATX Power Supply ($43.98 @ Newegg)
Total With 10% Tax: $43.98
Total Without Tax: $39.99
Prices include shipping, taxes, and discounts when available
Generated by PCPartPicker 2016-04-27 22:58 EDT-0400

The 650W version performed quite well: http://www.hardwaresecrets.com/fractal-design-tesla-r2-650-w-power-supply-review/

I would just be hesitant to recommend the untested 500W one, but Jonnyguru tested the 1000W one and Hardwaresecrets the 650W one, and they both perform well. For $39.99, getting a PSU like this seems like a better option than the Antec VP-450. If I'm correct this even has a ball bearing fan.

Review, 500w version, translated, so just deal with the poor syntax. Google hasn't learned how to do that yet.

Conclusion

Our first PSU Fractal Design leaves a mixed impression in the test. Sure, the external processing and design are at the high level of the Fractal case and the cable lengths qualify the power adapter for mounting the housing base . In the electronics used but there is light and shadow.

While efficiency and voltage regulation convince fully , the PSU messes in the Restwelligkeitsmessung and returns a result minimally beyond the allowable range. Our comparison sample from the commercial delivers slightly better values ​​, the ripple on the +5 volt line , however, is only slightly below the limit . Of course, such a small overshoot at risk as at our first copy of stability or life of the computer , however , the question arises why one should buy a power supply that can only at best retract sufficient results in this important point . Finally, there is a price of just under 90 euros numerous alternatives .
lateral view Ansichtseitliche

Also off the ripple & noise values ​​, there are other points of criticism : The fitting of the electronics with Taiwanese middle-class capacitors is only average considering the price range. Many competitors soothe critical customers with five-year warranty - the Tesla R2, the warranty period is only three years , and there is no way to direct processing. Heavy but weighs Fractal Designs dealing with the security features : First, the protection circuits are completely undocumented, on the other hand lacks both the over-current protection OCP and the overheating protection OTP .

Even with the volume remains untapped potential : While the volume confident at high load , the fan speeds could well be lower at lower loads . The ball bearing fans is also a compromise solution and more upper middle class than perfect.

 

[Someone Somewhere]

So according to my math, pretty much what that standard is saying is that any power interruption of less than 16ms should not adversely affect the psu whatsoever, it would continue normal operation. Now I may be wrong, but in all the years I've been cognizant of electricity, I can honestly say I don't think I've ever come across a power interruption of such limited time. When the power goes out at my house, its for hours, not ms.

If you are going to drop $1000+ on a good system, spend the extra $100, buy a decent UPS and that'll give you minutes to shutdown. It's not a difficult concept to grasp.

Ever seen your lights flicker? That's a momentary power interruption. Most loads don't really care; computers can.

And yes, most small UPSs are standby units. Efficiency would be *horrid* on ones that step down to 12V, then invert back again - like 70% or worse. You'd need tons of active cooling, and twice as much power electronics.

UVP is definitely on the output.

If you want a really crap PSU, ebay is probably an option.

 

[turkey3_scratch]

Well, the 12V rail is still the most important, as it's not too realistic to see the 5V and 3.3V rails get loaded and have high ripple (though sometimes when they are too low it creates a crossload situation and still they have high ripple). The 650W Tesla R2 also had somewhat high minor rail ripple, and Teapo caps (not Capxon okcnaline) but remember this unit is $40. You can't expect Japanese caps. And as I said, voltage regulation is very good.

Where was that review from darkbreeze? I want to read it.

 

[Onus]

The InWin unit that Nuckles_56 linked may actually not suck. It won't be great, but it would not surprise me if it is at least as solid as some of the low-end FSP units were back eight or ten years ago; solid tier-3 stuff. C.Hegge at Hardware Insights reviewed one of their TFX units for me, and it was good enough to actually use (it does use Capxon, but under light loads shouldn't be a problem).

 

[turkey3_scratch]

What I never get is that most reviewers just say what brand the capacitors are, whereas in reality the series is more important.

 

[Onus]

True; I'm pretty sure this was a "GF," one of the questionable ones.
I'm using this PSU in Igor now, and the only reason I might replace it (low priority) is to get the idle power usage down (30W now) with one of the Seasonic Gold TFX units.

 

[turkey3_scratch]

The reviewers, though, should still be telling the readers more stuff. Saying the series name is not good enough; saying how good that series is is important, because it's assumed the reader may not know what series is good or bad, and telling them that means a reduce in the amount of research required.

Another thing reviewers should mention are things like the trigger points for UVP and OVP. They usually say what ICs are in the power supplies with a link to the specification page for that IC, but realistically what normal person reading a review like this is going to read that gigantic PDF document on the specifications of that IC? Matter of fact reviewers should mention more than just saying, "This is a CM6100. Moving on..." they should be more descriptive about the exact components they are telling you are present.

 

[Karadjgne]

That is assuming that the reader is literate. Most, I'd have to say, are only basically aware of what's going on in a psu and probably skip over the technical jargon and go straight to test results and summary. The more of that technical jargon there is, the more they'd be inclined to skip to the end. So what you end up with is a very good, precise, review that took way too long to complete, that very few will fully read and even less who understand.

 

[turkey3_scratch]

That is assuming that the reader is literate. Most, I'd have to say, are only basically aware of what's going on in a psu and probably skip over the technical jargon and go straight to test results and summary. The more of that technical jargon there is, the more they'd be inclined to skip to the end. So what you end up with is a very good, precise, review that took way too long to complete, that very few will fully read and even less who understand.

I guess so 🙁 makes me sad when some people just look at the last page, because I can always find a lot more from reading the actual review.

Then again, a PSU review isn't like a review of any other hardware. Anybody can read a CPU or GPU review. A PSU is supposed to be more technical.

 

[Karadjgne]

I read a review of some motherboards. It was 2x msi vrs 2x Asus vrs 2x Gigabyte vrs 2x asrock. 1 each of mid grade boards and 1 each of highest grade. According to the review, there was a definite difference in ability between the board grades and some clear winners on things like lan (asrock) temps (asus) power efficiency (msi) and pcie/USB throughput (gigabyte). There were lots of graphs, many facts and figures, loads of details explaining why one board or brand was better. Skip to the end... Basically the boards show no performance improvement. The results were all in nanoseconds. Asrock shoving a byte through in 5 nanoseconds vrs gigabytes 12 nanoseconds is meaningless. Shaving 1/2 a second off a large download proved nothing. An interesting read, sure, but the summary said everything of any real importance. The mid grade boards had a realistic performance no worse than the high grade boards, the high grade boards just had more gimmicks, more fan headers, would take more OC abuse etc.

A psu review isn't any different. Ppl just want to know if it's built well, will perform well, is going to last and has enough protections to safeguard the other expensive components. Knowing what the vendor specified exact UVP setpoint is set to is meaningless to the vast majority since almost nobody monitors working voltages with a multimeter for extended periods of time just to see if the setpoint works as advertised.

For those who have a genuine need to know, that's what the pdf is for.

 

[turkey3_scratch]

Well for PSUs the summaries even are somewhat technical, in the fact that the reader has to know what the summary is talking about. That's why many people like the "points system" because it requires nothing other than knowing how to count 1-10, but I don't like the points system at all, it's shown to not be very good.

 

[Someone Somewhere]

Anybody can read a CPU or GPU review.

BS. Go look at some of the reviews on e.g. Haswell's architecture.

That's why many people like the "points system" because it requires nothing other than knowing how to count 1-10, but I don't like the points system at all, it's shown to not be very good.

No, the problem there is no-one can agree on what makes a good PSU.

 

[turkey3_scratch]

Anybody can read a CPU or GPU review.

BS. Go look at some of the reviews on e.g. Haswell's architecture.

That's a review on a CPU architecture, not a specific CPU. The important part of it are the graphs and comparison charts, power consumption graphs, and temperature graphs. Most people can read those.

 

[Someone Somewhere]

There's almost no difference. You basically won't find anything over "it's clocked at X GHz, has Y cores, and costs $Z" for individual CPUs.

Every major CPU review is a review of the architecture the CPU is built on, because all the CPUs built on that die are released together.

So PCIe lanes, changes to the memory controller etc. don't matter? Saying "28GB/s is better than 24GB/s" is going to get you in a world of trouble when you buy DDR3-2133 instead of DDR4-2133.

 

[Onus]

The issue is that too many people want to put their personal stamp on things. It's an ego thing, and everybody does it. I'd like to believe it is one of the things I have most improved about my own posting; I no longer feel inclined to niggle people's builds, but I do remain extremely opinionated about other stuff.
Pick your battles. For all practical intents and purposes, some stuff simply doesn't matter, like that last .1V of ripple, whether or not there's a Berg connector, etc. Let. It. Go.

 

[turkey3_scratch]

There's almost no difference. You basically won't find anything over "it's clocked at X GHz, has Y cores, and costs $Z" for individual CPUs.

Every major CPU review is a review of the architecture the CPU is built on, because all the CPUs built on that die are released together.

So PCIe lanes, changes to the memory controller etc. don't matter? Saying "28GB/s is better than 24GB/s" is going to get you in a world of trouble when you buy DDR3-2133 instead of DDR4-2133.

I don't think anybody here knows the complete ins and outs of these architectures. What matters are the performance charts, that tell you how they compete against other CPUs in similar applications.

 

[Someone Somewhere]

I would be surprised if anyone here knows 'the complete ins and outs' of switchmode design, either.

.1V of ripple is quite a lot

What matters are the performance charts, that tell you how they compete against other CPUs in similar applications.

Again, that is how you get "oops I don't have enough PCIe lanes", "what do you mean it's not compatible with my motherboard", and other similar issues. If you want a bargraph of how fast each CPU is, we might as well just throw you at Anandtech's benchmarks, or CPUBoss.

 

[Onus]

...
.1V of ripple is quite a lot
...

Absolutely my bad; I meant voltage regulation. Yeah, that's a lot of ripple; enough to make speakers hum.

 

[turkey3_scratch]

I would be surprised if anyone here knows 'the complete ins and outs' of switchmode design, either.

.1V of ripple is quite a lot

What matters are the performance charts, that tell you how they compete against other CPUs in similar applications.

Again, that is how you get "oops I don't have enough PCIe lanes", "what do you mean it's not compatible with my motherboard", and other similar issues. If you want a bargraph of how fast each CPU is, we might as well just throw you at Anandtech's benchmarks, or CPUBoss.

Knowing if a CPU is compatible with the motherboard is pretty simple and doesn't require architectural knowledge.

 

[Someone Somewhere]

Oh, sure, it'll "work". You know what we say about PSUs that "work".