Author Topic: switching-PSUs: why can't you usually exceed the 50% of declared power?  (Read 11756 times)

0 Members and 1 Guest are viewing this topic.

Offline 0dbTopic starter

  • Frequent Contributor
  • **
  • Posts: 336
  • Country: zm
You're really worried about it taking 1 second to trip?

I am more worried about the AEROCOOL VX PLUS 550W, because it didn't even shutdown; at 250Watt of load it made smoke until it exploded, and at this point the lifesaver in my appartment interrupted the energy, but before it happened, the measured p-p ripple was over 300mV on the 12V line, and this for sure can damage things like harddrives and VRMs.

The Superflower Leadex costs 120 euro. As soon as I tried to connect a 650 Watt load it immediately cut of the power, and during all the tests the p-p ripple over the 12V line was a flat line, always about 30mV!
- Fast OVP!
- flat line p-p ripple!
- fair rating on the the label
It's a true 650Watt PSU!

Now you have two examples. One very bad PSU, one very good PSU.
 

Offline wizard69

  • Super Contributor
  • ***
  • Posts: 1184
  • Country: us
Scam artist and ethics problems from manufactures in China.    If you are going to buy hardware form China you need to know the ins and outs of the industry and who the reliable manufactures are.     Obviously there are a lot of good manufactures in China but there are also many that are nothing but scam artist out to make a quick buck.    The big problem for the rest of us is separating the ethical manufactures from the dealers in Junk.   


There is also no legal structure in China that effectively polices the manufactures there.   There has to be a massive public outrage for the Chinese government to take action.    Look up the problems they recently had with vaccines for Children.

The Superflower Leadex is a great series of Switching-Mode PSUs, and if you buy a 650 Watt unit, you can be sure the unit will serve you up to 650 Watt without dropping voltage, or increasing the voltage ripple.

But let's talk about cheaper PSUs. I have recently seen a lot of 500 Watt units where you cannot exceed th 50% of the declared power, thereore if you buy a 500 Watt, it will serve you only up to 250 Watt, and beyound it might get damaged.

What I cannot understand is why. Which are the technical reasons for this? And what one should look into the circuit to have some enlightenments about the quality of the product concerning its behavior under load?

Thanks
 

Offline tautech

  • Super Contributor
  • ***
  • Posts: 29812
  • Country: nz
  • Taupaki Technologies Ltd. Siglent Distributor NZ.
    • Taupaki Technologies Ltd.
You cannot assess a PSU with light bulb loads OR its capability if loading only one rail !

Each output can only be loaded to its rated output and several output loads may be necessary to reach its full rated output in watts.....but specs may be only in watts consumption NOT total output.

Further, tungsten bulbs have high initial current draw until the filament reaches operating temp where its resistance is much lower higher than when cold. This inrush current could send the SMPS into shutdown.

All ^^^ traps for the unwary !

Edit to correct brain fart.  |O
« Last Edit: May 24, 2020, 09:19:28 pm by tautech »
Avid Rabid Hobbyist.
Some stuff seen @ Siglent HQ cannot be shared.
 
The following users thanked this post: Someone, 0db

Offline rsjsouza

  • Super Contributor
  • ***
  • Posts: 6106
  • Country: us
  • Eternally curious
    • Vbe - vídeo blog eletrônico
You cannot assess a PSU with light bulb loads OR its capability if loading only one rail !
100% agree, but an exploding power supply is totally unacceptable. So, the lightbulbs are not necessarily useles as you can still separate the wheat from the chaff.
Vbe - vídeo blog eletrônico http://videos.vbeletronico.com

Oh, the "whys" of the datasheets... The information is there not to be an axiomatic truth, but instead each speck of data must be slowly inhaled while carefully performing a deep search inside oneself to find the true metaphysical sense...
 

Offline engrguy42

  • Frequent Contributor
  • **
  • Posts: 656
  • Country: us
I know folks just LOVE to make sweeping generalizations and dismissals about stuff, but usually they are incorrect.

Anyway, as I said, decent power supplies don't blow up, for many reasons. Look at the ATX requirements. And realize those are MINIMUMS. Decent manufacturers add to those to make sure they don't provide crap and thereby ruin their reputations.

Nobody here knows why your power supplies failed. Nor can we since we don't know the internals of the power supplies or what you did to test them and so on. If you want to believe that all power supplies that aren't the Superflower are junk, and can't be loaded past 50%, go right ahead. If you want to believe that if it was manufactured in China it's probably junk, go ahead. You'll get a lot of fanboys here agreeing. And you'll certainly get a lot of anti-ATX fanboys agreeing with anything that anyone says that's negative towards ATX supplies. You'll even get people saying they're bad because a bug can crawl in and zap it. 

But if you're interested in the truth, do some research. Look inside the failed power supplies and see what happened. And see what the supervisor circuit is, and what it's designed for.

Without that it's all just vague speculation about stuff nobody understands.
« Last Edit: May 24, 2020, 10:41:24 am by engrguy42 »
- The best engineers know enough to realize they don't know nuthin'...
- Those who agree with you can do no wrong. Those who disagree can do no right.
- I'm always amazed at how many people "already knew that" after you explain it to them in detail...
 

Offline Mechatrommer

  • Super Contributor
  • ***
  • Posts: 11714
  • Country: my
  • reassessing directives...
You cannot assess a PSU with light bulb loads OR its capability if loading only one rail !
yup on 12V rail, i can see its less than half the 450W PSU rating, ie 18A, so its only half the story. btw, last time i bought Gigabyte brand ATX PSUs for my PC, decent price. both died prematurely in less than a year, the second one only for few weeks, they only have nice nylon wires grouping sheathing, cute big fans and quality like exterior black paint, but functionally abysmal. then i thought what the heck, let me just buy the cheap $20 one (TronMonxter from Tatu) its surprised me when they can survive for like 4-6 years. the second one i just dismantled not because it blows up dead, its still working, its just the wiring near connectors becomes stiffen and broken in few pins and places making my PC shut off from time to time for sometime now. now i cant stand it so my PC is on a new cheapo PSU again only starts working for few days now, no more intermittent shut off no more corrupted hibernated data on power up. i keep reusing the nice nylon sheathing power connectors that i stripped from Gigabyte PSU to get more connectors for my many HDD/SDD/DVD etc because the cheap one has too few on it, so the dead Gigabyte is not entirely useless. i will keep 1 or 2 PSU unit in stock bought every year or 2, coz i know the working one will sometime failed unexpectedly, so keeping in stock will not leave me in the dark for a day or 2 while finding a shop that can supply me a PSU. some of them are even dirt cheap like $10 i can buy 2 or 3 units in one receipt if i want to, but i dont want to occupy much of my storage space there are many more items to store, not just ATX PSUs. so on rare occasion, this is where you buy cheaper to get better product..
« Last Edit: May 24, 2020, 11:16:53 am by Mechatrommer »
Nature: Evolution and the Illusion of Randomness (Stephen L. Talbott): Its now indisputable that... organisms “expertise” contextualizes its genome, and its nonsense to say that these powers are under the control of the genome being contextualized - Barbara McClintock
 

Offline wraper

  • Supporter
  • ****
  • Posts: 17952
  • Country: lv
yup on 12V rail, i can see its less than half the 450W PSU rating, ie 18A, so its only half the story.
Particular PSU has 42A/504W rating on 12V rail out of 550W total. Any modern half-decent ATX PSU has at least 80-90% of total power rating on 12V rail.
 

Offline PKTKS

  • Super Contributor
  • ***
  • Posts: 1766
  • Country: br
You cannot assess a PSU with light bulb loads OR its capability if loading only one rail !

TRUE.

The ATX  sticker ratings are meaningless - they just sum the rails.

Making tests on a single rail is also meaningless - different supervisors
sample the output in different ways.

Cheap ATX up to 500W **VERY CAREFUL CHOSEN**  can deliver
up to 300W on a well balanced load.

Apart this you have a bad setup.
I do have a lot of those cheapos - they work fine if proper balanced

Aerocool VX series can deliver 350W just fine in well balanced systems

Half bridge cheap designs they all use cheap components and require
a proper and careful load balance for proper operation.

IT IS NOT LIKE:  "HEY 500W in a single 12V rail ... exploded?"
obvious it will explode

Paul
 

Offline strawberry

  • Super Contributor
  • ***
  • !
  • Posts: 1199
  • Country: lv
What can be found in cheap psu: They use BJT even today. Input filter cap size less than ~1uF per watt. Poorly designed control loop (oscilation, voltage overshoot !) and controller. Stadard instead of low ESR. Poor quality enemaled copper wire. Heat and loss is not a thing (POWER). Rock and sand do not improve encapsolated psu heat disipation. Poor conductivity silicone rubber thermal pads.

500w psu have to disipate ~100w@80% of heat.
 

Offline wraper

  • Supporter
  • ****
  • Posts: 17952
  • Country: lv
Photo of this trash

 

Offline PKTKS

  • Super Contributor
  • ***
  • Posts: 1766
  • Country: br
What can be found in cheap psu: They use BJT even today. Input filter cap size less than ~1uF per watt. Poorly designed control loop (oscilation, voltage overshoot !) and controller. Stadard instead of low ESR. Poor quality enemaled copper wire. Heat and loss is not a thing (POWER). Rock and sand do not improve encapsolated psu heat disipation. Poor conductivity silicone rubber thermal pads.

500w psu have to disipate ~100w@80% of heat.

There is nothing wrong using BJTs in FULL BRIDGE and HALF BRIDGEs

MY BENCH  1200W (60V@20A) is a HALF BRIDGE uc38xx current mode
based one - and does the job just fine.

Under 600W ATX half bridges are cheaper
they also DON'T REQUIRE TRUE SINE WAVE INVERTERS
YOU CAN COMBINE 2 to have affordable high power in case
they cost less and do the job

drawback is that you have a lot of bad ones and you need
to choose very carefully you also need to balance the system

And cheap systems (like those used SOHO or POS) just
do not require a 600W PFC with true sine inverter to operate.

Below 350W real power a cheap ATX with much more cheap
non -sinus inverter will more than half the budget.

REASON to use them?  COST.

Paul
 

Offline wraper

  • Supporter
  • ****
  • Posts: 17952
  • Country: lv
There is nothing wrong using BJTs in FULL BRIDGE and HALF BRIDGEs

MY BENCH  1200W (60V@20A) is a HALF BRIDGE uc38xx current mode
based one - and does the job just fine.
Again you promoting and defending trash PSUs. In bench PSU efficiency it is no so critical, we often use even linear PSUs. What's wrong with BJT is guaranteed subpar efficiency / high heat dissipation. And when talking about ATX PSUs in particular, that nowadays you will find them only in trash models.
EDIT: and when you find them in driving transformer in bench PSU, usually it's either because PSU is old or model is produced for a long time. Or simply because it's a cheap trash too.
« Last Edit: May 24, 2020, 12:07:55 pm by wraper »
 

Offline PKTKS

  • Super Contributor
  • ***
  • Posts: 1766
  • Country: br
Again you promoting and defending trash PSUs. In bench PSU efficiency is no so critical, we often use even linear PSUs. What's wrong with BJT is guaranteed subpar efficiency / high heat dissipation.

Yes, I have literally dozens of them without problems.
I use them literally for decades including several small
PSUs for bench work - all half-bridge design - all BJT based

I use HIGH COST PFC based ones in HIGH END PC GAMERs
where the cost literally explodes any budget.

A TRUE SINE WAVE INVERTER au pair with that ATX is also required
and considering benefits - IMHO  they do not worth the money.

For the record: I AM NOT "PROMOTING" garbage.
Stop saying such shit

I am putting simple facts. Where they are not needed
those expensive solutions are expensive crap

My bench uses 5 HALF BRIDGE BJT based PSUs for decades
daily based - nothing wrong with that.

SOHO and POS give a shit if 70% 80% or 90% ratings
they care about monthly TOC (ownership cost)

Me too.  I give a shit if 80% plus gold or shit.
I care about total monthly cost
Paul
« Last Edit: May 24, 2020, 12:03:17 pm by PKTKS »
 

Offline wraper

  • Supporter
  • ****
  • Posts: 17952
  • Country: lv
SOHO and POS give a shit if 70% 80% or 90% ratings
they care about monthly TOC (ownership cost)
Yeah, and garbage you advocate for has higher cost of ownership in any country where electricity price is not dirt cheap.
Quote
For the record: I AM NOT "PROMOTING" garbage.
Stop saying such shit
Then why everyone laughed at you in the last thread where you praised PSU which internally is basically the same as trash discussed in this thread?
 

Offline PKTKS

  • Super Contributor
  • ***
  • Posts: 1766
  • Country: br
I have yet to see a BJT PSU anywhere other than in very old industrial framed SMPSs and those cheapest 5V 1A adapters.

Even the cheapest junk phone chargers now use MOSFET-based monolithic flyback ICs.

In China, you can get a 5V 2.4A primary chip for somewhere around $0.15, so why bother with BJTs?

My opinion about those small cheap patented supervisor
adapters is unprintable.

I have a box of failed ones - supervisors (OBxx and TOPxx) are mostly
RCC (very poor ratings) designs..

patented and almost dispensable garbage.

Last month I just ditch 7 failed IOT of those.

They just do not worth the hassle...
Paul
 

Offline PKTKS

  • Super Contributor
  • ***
  • Posts: 1766
  • Country: br
SOHO and POS give a shit if 70% 80% or 90% ratings
they care about monthly TOC (ownership cost)
Yeah, and garbage you advocate for has higher cost of ownership in any country where electricity price is not dirt cheap.
Quote
For the record: I AM NOT "PROMOTING" garbage.
Stop saying such shit
Then why everyone laughed at you in the last thread where you praised PSU which internally is basically the same as trash discussed in this thread?


Suit yourself - the money you put in your hardware is all yours.

There is nothing to laugh on my simple well put facts.

Just the cost which includes repair and probably replacing
those advocated "better" ones.

Being there several times - they are just as bad and more expensive.

Paul
 

Offline engrguy42

  • Frequent Contributor
  • **
  • Posts: 656
  • Country: us
FWIW, it's a myth that higher efficiency power supplies save you a lot of money on your electric bills. Do the math.

In the US the average cost of electricity is about 13 cents per kWH. In Brazil it's like 14 cents (US). If you calculate the difference in kWH ACTUALLY used with a 90% efficiency power supply versus an 80% power supply it amounts to maybe a few US $$ per year. A bit like the cost savings by running around your house and turning off a 50 watt light bulb when it's not being used.

Unless you run your computer at maximum load 24 hours a day, 365 days a year.... 
- The best engineers know enough to realize they don't know nuthin'...
- Those who agree with you can do no wrong. Those who disagree can do no right.
- I'm always amazed at how many people "already knew that" after you explain it to them in detail...
 

Offline Mechatrommer

  • Super Contributor
  • ***
  • Posts: 11714
  • Country: my
  • reassessing directives...
Photo of this trash..
the topology is not much better than cheapo one. they can make nice fan and stickers, enclosure paint, nylon sheath and any spec list they want and put a sound price to it... the way they cut corner on surge protection/PFC stage is just similar to a $10 unit, maybe thats where they outsourced/get their supply from. just for fun, i dig back the $40+ damaged Gigabyte PSU, 1st pic, no cut corner on surge protection, plastic insulation on both sides, look at that nice heatsinks, it can be reused for other project thats why i keep them (i keep more damaged PSU for parts than the working one). but even that beauty, its a PoweRock my ass only worked for about less than a year iirc. 2nd picture is the $20 TronMonxter, no surge protection (cut corner), less 50W power, but worked for 3-4 years. go figure... cheaper is better! ps: both pic below shows better/beefier main transformer than your shown trash.
« Last Edit: May 24, 2020, 01:00:33 pm by Mechatrommer »
Nature: Evolution and the Illusion of Randomness (Stephen L. Talbott): Its now indisputable that... organisms “expertise” contextualizes its genome, and its nonsense to say that these powers are under the control of the genome being contextualized - Barbara McClintock
 

Offline engrguy42

  • Frequent Contributor
  • **
  • Posts: 656
  • Country: us
A bit like the cost savings by running around your house and turning off a 50 watt light bulb when it's not being used.

I do that all the time. It's not just about money. Saving is just a plain virtue.

That's up to you...as long as you realize the facts, and how it's no different than, say, leaving your air conditioner on a few extra minutes a few times a year when you don't really need it. And so on.

Power supply efficiency, for most, amounts to maybe a few $$ per year. If that matters, then fine. As long as people are aware, and not just handwaving on stuff they don't really understand.
- The best engineers know enough to realize they don't know nuthin'...
- Those who agree with you can do no wrong. Those who disagree can do no right.
- I'm always amazed at how many people "already knew that" after you explain it to them in detail...
 

Offline coppice

  • Super Contributor
  • ***
  • Posts: 10035
  • Country: gb
As well as all the issues of cheaper designs being stripped to the bone to sell at crazy low prices, there are also issues of how the load is spread among the output rails. You can only get up to the rated maximum with any PC supply if you load each of the outputs by the right amount.

Older supplies focussed a lot of their capacity on the lower voltage rails, and not enough on the 12V outputs to suit modern PCs. Modern supplies have limited capacity on the lower voltage outputs, and most of the capacity is on the 12V outputs. The supplies for most rack mount servers only output 12V these days, and PCs will probably go his way. However, supplies they often have multiple 12V output sections, and you need the load to be balanced across those sections to get close to the rated capacity.
 

Offline PKTKS

  • Super Contributor
  • ***
  • Posts: 1766
  • Country: br
FWIW, it's a myth that higher efficiency power supplies save you a lot of money on your electric bills. Do the math.

In the US the average cost of electricity is about 13 cents per kWH. In Brazil it's like 14 cents (US). If you calculate the difference in kWH ACTUALLY used with a 90% efficiency power supply versus an 80% power supply it amounts to maybe a few US $$ per year. A bit like the cost savings by running around your house and turning off a 50 watt light bulb when it's not being used.

Unless you run your computer at maximum load 24 hours a day, 365 days a year....

Thanks...  for the clever comment.

ABSOLUTE TRUE.

This is a total wrong argument. The savings on the bill
promoted by such "ratings" 80% 90% gold plus shit are garbage.

Those PSUs are targeted to high end PCs where you use
hungry power GPUs which waste awesome 700W of true power
plus ownership costs.

SOHO and POS are well below 300W and the cost focus
is more on hardware wearing out soon than the mains bill

Their air conditioner is ORDERS OF MAGNITUDE above...
on the mains bill  and their pile of PCs are a real problem.

So saving more than double the cost on EACH OF THEM..
is a very good argument with their managers.

Paul
 

Offline wraper

  • Supporter
  • ****
  • Posts: 17952
  • Country: lv
Power supply efficiency, for most, amounts to maybe a few $$ per year. If that matters, then fine. As long as people are aware, and not just handwaving on stuff they don't really understand.
If computer works 8 hours a day, when comparing decent and cheapo PSU, power consumption difference over about 1-3 years will cover price difference and in some places even buy you additional good quality PSU.
« Last Edit: May 24, 2020, 01:06:13 pm by wraper »
 

Offline coppice

  • Super Contributor
  • ***
  • Posts: 10035
  • Country: gb
FWIW, it's a myth that higher efficiency power supplies save you a lot of money on your electric bills. Do the math.
For most people moving from an 80% efficient supply to a 90% efficient supply in their PC will not save much in electric bills, because their PC is only working hard for a small percentage of the time. However, for server, telecoms and other high power applications running flat out 24 hours a day every extra percent of efficiency brings a big enough reduction in power bills to justify significantly higher complexity in their power supplies. Try looking at the materials used to promote multi-kilowatt supplies in the 95% to 97% efficiency area. Its all about cost breakdowns over the life of the equipment.

For consumers a shift from 80% efficiency to 90% efficiency is not useless. Halving the supply's heat output helps compact machines to run cool with a very quiet fan.
 

Offline engrguy42

  • Frequent Contributor
  • **
  • Posts: 656
  • Country: us
Power supply efficiency, for most, amounts to maybe a few $$ per year. If that matters, then fine. As long as people are aware, and not just handwaving on stuff they don't really understand.
If computer works 8 hours a day, when comparing decent and cheapo PSU, power consumption difference over about 1.5-3 years will buy you additional good quality PSU.

Again, you're free to believe whatever you want, but if you actually look at the numbers for your ACTUAL usage you may find out that's not the case.
- The best engineers know enough to realize they don't know nuthin'...
- Those who agree with you can do no wrong. Those who disagree can do no right.
- I'm always amazed at how many people "already knew that" after you explain it to them in detail...
 

Offline Mechatrommer

  • Super Contributor
  • ***
  • Posts: 11714
  • Country: my
  • reassessing directives...
please also remember a PC is not the only thing in a (full family) house. TV, hifi, wifi router etc use SMPS. and sometime 3-5 PC/laptop in the house. promoting for extra 10% efficiency and making a habit to turn off appliances when not in use can buy you a good toy within a year.
Nature: Evolution and the Illusion of Randomness (Stephen L. Talbott): Its now indisputable that... organisms “expertise” contextualizes its genome, and its nonsense to say that these powers are under the control of the genome being contextualized - Barbara McClintock
 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf