UPS recommendation for late 2013 Mac Pro

[ActionableMango]

Yes, my external HDs would spin down when they are not used. Does it have benefits to shut off power to them when the computer is in sleep?

I wouldn't think there is much in the way of benefits. Once they are spun down they user very little power.

In fact I think turning them off while the computer is asleep could cause problems. It's something you could test now with your current setup.

 

[vett93]

I wouldn't think there is much in the way of benefits. Once they are spun down they user very little power.

In fact I think turning them off while the computer is asleep could cause problems. It's something you could test now with your current setup.

You are right. I put the Mac Pro in sleep mode and then power off the adapters for HDs. This woke up Mac Pro immediately.

 

[AidenShaw]

I wouldn't think there is much in the way of benefits. Once they are spun down they user very little power.

One thing to consider is what technology is used in the power bricks for the external HDs.

If the power bricks are very light - most likely they are switching power supplies and "Once they are spun down they use very little power" is a good call.

If the power bricks are heavy, they probably contain big transformers and will tend to use a lot of power even when the disk is spun down.

If the power bricks stay warm even when the disks spin down - that's a bad sign.

 

[westom]

I put the Mac Pro in sleep mode and then power off the adapters for HDs. This woke up Mac Pro immediately.

A computer has settings to select what does and does not wake up a computer. These options include the keyboard, mouse, ethernet interface, and selected USB ports.

For each device should be an option to disable that wakeup function.

Meanwhile investigate (search for) 'smart' power strips so that a powered down computer in its first receptacle causes all other connected devices (receptacles) to power off.

Meanwhile, sleep is not power off. Hibernate or shutdown does a power off. Then unpowering a a USB connection does not wake up that computer.

Disk drives use very little power when fully operational. Typically single digit watts (typically less than a night light). Explains why an external disk drive could even be powered directly from a USB port. Once a disk drive's reduced power mode is implemented (if it exists for that disk drive), then power consumption is even less.

 

[AidenShaw]

Meanwhile investigate (search for) 'smart' power strips so that a powered down computer in its first receptacle causes all other connected devices (receptacles) to power off.

A quick search for these only found smart strips with surge protection - which has been assumed to be questionable downstream from a UPS.

https://www.cableorganizer.com/smart-strip/?gcssku=BTL-SCG5&utm_source=Bing&utm_medium=cpc&utm_term=&utm_campaign=73307699&msclkid=552c8a9827ba1acee41dd132eb1143ed&utm_content=Everything Else - Alt

...and is rather useless if the smart power strips' power off of the disks immediately rewakes the Apple.

 

[westom]

A quick search for these only found smart strips with surge protection - which has been assumed to be questionable downstream from a UPS.

Again, please read what was posted before replying. What is the purpose of that UPS? Why does it even exist?

Again, a wakeup function that causes a computer to wake up can be disabled. And do not use sleep; use hibernate or shutdown.

Unread paragraphs (and others) already answered those questions.

Meanwhile, all power bricks must be switching power supplies. There is no massive transformer inside. Those technologies have been obsolete for many decades due to costs, design standards that require electronics to be more robust, basic electrical reasons, and even national energy standards.

 

[vett93]

...

Again, a wakeup function that causes a computer to wake up can be disabled. And do not use sleep; use hibernate or shutdown.

...

How do you get a Mac to enter hibernate mode?

 

[Ludacrisvp]

How do you get a Mac to enter hibernate mode?

There is a pmset option from the command line to do it. Can’t recall offhand. Someone also made a prefpane to manage sleep but also can’t recall the name of it.

 

[westom]

How do you get a Mac to enter hibernate mode?

Search in Terminal Applications in /Applications/Utilities/.

I don't remember. But wakeup functions may also be disabled there. Somehow I remember it being in something called Setup. That is why computers (should) always have a help button. I can never remember, from computer to computer, where necessary hardware controls are located.

 

[crjackson2134]

Read this for an explanation of how it works.

https://computers.tutsplus.com/tutorials/how-to-hibernate-a-mac--cms-23235

 

[AidenShaw]

Meanwhile, all power bricks must be switching power supplies. There is no massive transformer inside. Those technologies have been obsolete for many decades due to costs, design standards that require electronics to be more robust, basic electrical reasons, and even national energy standards.

Do you have any links to support "all" (vs. "most" or a less absolute claim), and the "decades" assertion?

When I look at peripherals that I've bought in the last few years - most of them have lightweight power bricks that are most likely switching, but some have rather heavy power bricks that obviously have a "massive" transformer inside. ("Massive" for a 10 watt supply would be 100g or so.)

(A Craftsmen cordless screwdriver that I bought two years ago has a heavy 10 watt power brick that stays warm even when the screwdriver is fully charged. I'll assume linear, not switched.)

Actually, do you have any links to support anything that you've claimed? Any? And how about "national energy standards"? Any?

And, to your point of reading what has been written, many posts have talked about active power factor correction ("active PFC") PSUs as a significant factor in UPS compatibility - not a naïve distinction between linear and switching supplies.

 

[westom]

Do you have any links to support "all" (vs. "most" or a less absolute claim), and the "decades" assertion?

I am not the one claiming that some power brick is not a switching power supply. Since you (believe you) have one, then we are reading that brick's specification numbers that define it as not a switching supply. And you have said what kind of supply it is. All we have is some personal feelings because you think it is heavy.

Making a conclusion only because you feel[.i] it is heavy is a fool's errand. Honest replies always cite numbers. Cite the model and specification for a power brick that is not a switching supply. Even the weight (in pounds) and voltage a/ current outputs were not cited. Only cited was a speculation.

Little hint. If it is not a switching supply, then it cannot meet the so many energy standards, power requirements, have robust internal protection, operate over a wide range of AC voltages, and be sold anywhere near to switching supply prices.

Post this model number, specification number, and other spec numbers that were read before making a recommendation.

 

[westom]

So we tested many computers and many UPS units and discovered mostly that there isn't really a consistent rule as to whether computers would work with simulated sinewave UPS or not.

At one point, the EU demanded very high power factor correction circuits. Those would be confused by waves that, for example, exceeded 0.2% THD. Those supplies were not required in North America. But many more innovative manufactures marketed some with that high power factor correction.

Since then, Europe has backed off on that standard. And North American computers with that extremely high power factor correction are no longer marketed.

Your experiments say little useful if numbers for power factor correction and numbers for 'dirty' AC power (ie %THD) are not in that study. There are consistent trends if numbers are provided. Terms such as simulated or stepped sinewave are subjective - not informative or useful.

It was a short period when some computers had extremely high power factor correction. 'Dirty' AC clearly did not damage any hardware - as implied in myths. Some computers sometimes would not power on when 'dirty' power confused that extremely high power factor correction circuit.

Does 'dirty' power harm electronics? Of course not. But the myth is promoted by fears - without facts or numbers. That 'dirty' power is problematic to less robust motorized appliances; not electronics.

 

[ActionableMango]

Your experiments say little useful

Our tests were extraordinarily meaningful to us. Stepped UPS vs sinewave UPS was the difference between many computers staying powered or not in a communications building.

Whether or not these are subjective terms might be very important to you, but it is completely irrelevant to us. You could call them UPS (A) and UPS (B), and we'd still pick the ones that don't cause computers to turn off or loudly squeal.

 

[MarkC426]

When I had an APC ups (non-sinewave), if there was a spike/drop in power my mac would wake up.
With my Cyberpower this does not happen.......

 

[westom]

Err deleted

 

[crjackson2134]

When I had an APC ups (non-sinewave), if there was a spike/drop in power my mac would wake up.
With my Cyberpower this does not happen.......

^^^^^ Same^^^^^

Just sayin’.

 

[westom]

No such thing as a pure sine wave. This UPS was marketed as a sine wave UPS. It outputs 200 volt square waves with a spike of up to 270 volts. They did not lie. We all learned it in high school math. Square waves and spikes are only a sum of pure sine waves. Pure sine wave does not exist. Honesty recommendation demands numbers such as %THD.

One did not know which UPS has a high or low %THD. A conclusion is only relevant to that UPS with that computer. It says nothing useful here for anyone else.

No relationship exists between a pure sine wave output and drop that might occur when the UPS responds to noise or voltage changes on AC mains. That drop can occur with a pure or non-pure sine wave UPS. And can occur with some model UPS from any manufacturer.

No hardware damage was created by 'dirty' UPS power. And that is the point. Pure sine wave UPS is promoted to protect hardware. Even the OP admits operating on a not-pure UPS for four years without damage. Hardware is perfectly fine (unharmed) even with 'dirty' power. At worse, 'dirtiest' power only means a computer (with high PF correction) does not power on. Nothing more.

Clean or dirty power from a UPS does nothing to protect or prolong hardware - despite UPS manufacturer and wild speculation promoting that fable. UPS also does nothing to protect hardware from transients on AC main. Another myth often promoted only on fear; not by facts with numbers. UPS is temporary and 'dirty' power so that unsaved data can be saved. Nothing more.

Other anomalies require other solutions.

 

[strawbale]

No such thing as a pure sine wave. .

Some UPSs that have been tested to produce a pretty good sine wave:

https://www.hardwareinsights.com/database-of-ups-output-waveforms/6/#Victron
https://www.hardwareinsights.com/database-of-ups-output-waveforms/6/#Hewlett-Packard
https://www.hardwareinsights.com/database-of-ups-output-waveforms/5/#Emerson
https://www.hardwareinsights.com/database-of-ups-output-waveforms/5/#Chloride

APC Back-UPS Pro 1000 VA (BP1000I) in here:
https://www.hardwareinsights.com/database-of-ups-output-waveforms/4/#APC-Schneider-Electric-other

some APCs in here:
https://www.hardwareinsights.com/da...#APC-Schneider-Electric-Smart-UPS-stand-alone

 

[westom]

Some UPSs that have been tested to produce a pretty good sine wave:

Demonstrated by that citation are so many 'dirty' UPS outputs. And again the point - which is not about a few UPSes with cleaner outputs. Relevant fact:

No hardware damage was created by 'dirty' UPS power. And that is the point. Pure sine wave UPS is promoted to protect hardware. Even the OP admits operating on a not-pure UPS for four years without damage. Hardware is perfectly fine (unharmed) even with 'dirty' power.

OP does not need a pure sine wave. That *need* is hyped by hearsay. UPS does not protect hardware. That myth also exposed previously.

Any of those UPSes (without or without a cleaner sine wave) is just fine for the OP. Since a UPS has one function: temporary and 'dirty' power so that unsaved data can be saved.

 

[AidenShaw]

Demonstrated by that citation are so many 'dirty' UPS outputs. And again the point - which is not about a few UPSes with cleaner outputs.

Wow - when presented with evidence that destroys your argument you claim that it supports your argument.

Dirty UPS:


Sine-wave UPS (the APC 1500 that I recommended):


You continue to claim "square waves with 270volt spikes" when these "oscilloscope" graphs show very nice approximations of a sine wave without any spikes.

Of course, you are technically correct that there are no "perfect sine waves", there will be some THD in any device. However, the output from the APC 1500 in the second chart is close enough to "pure" that the pixilation of the jpeg is more than the deviation from "pure".

Jesus H. Christ riding a bicycle.

Any of those UPSes (without or without a cleaner sine wave) is just fine for the OP. Since a UPS has one function: temporary and 'dirty' power so that unsaved data can be saved.

It's also been clearly proven that many high PFC power supplies shut down with dirty UPS power - so that the "one function" of the UPS fails if the UPS has dirty power. (And high PFC power supplies are by far the norm today.)

 

[westom]

You continue to claim "square waves with 270volt spikes" when these "oscilloscope" graphs show very nice approximations of a sine wave without any spikes.

Why put up a strawman and then attack it? Please read what was posted - not the strawman you have created.

Why would anyone spend $hundreds for the pure sine wave when most all computers work just fine on that 200 volt square wave with 270 volt spikes? That is my UPS - not all UPSes. Did you read what was posted or only what you wanted to see?

Why confuse everyone with fears of extremely high active power factor correction when almost nobody has that. Furthermore, 'dirty' electricity does not damage hardware. What are we discussing here? Protection of hardware - not a PFC that exceeds 0.99.

Failure to grasp the point is apparent. Almost nobody needs pure sine waves. 'Dirty' power does not harm any electronics. Power on AC mains needs no 'cleaning' from a many $hundreds UPS. 'Dirty' power also does not damage those rare and extremely active PFC circuits.

Start with this. What are we discussing here? Protection of hardware. Please grasp a topic before replying. Please tell us how and why a UPS protects computer hardware. And yes, I expect honesty. So that explanation will include basic electrical concepts and with specification numbers. How does that UPS protect hardware? What hardware? And why?

 

[AidenShaw]

Start with this. What are we discussing here? Protection of hardware.

I'll start by disagreeing here. We're protecting data, not hardware.

Please grasp a topic before replying.

I agree completely - and feel that you fail this test. The topic is protecting data - particularly non-persisted data in memory that is being processed.

Our group often runs jobs that run for days or weeks - we have UPS systems so that a random 1 second or 60 second power outage doesn't lose weeks of calculations.

None of my hardware has been damaged by power outages. I spend a lot on APC UPS systems (only $10K this month) - because we get occasional power glitches of a few seconds. Without a UPS, weeks of data science or machine learning could be lost (potentially of more $value than the computer systems themselves).

Please tell us how and why a UPS protects computer hardware. And yes, I expect honesty. So that explanation will include basic electrical concepts and with specification numbers. How does that UPS protect hardware? What hardware? And why?

Please get a clue and realize that some of us buy UPS systems for something much more important than protecting hardware from damage. If a power hit destroys a few hundred thousand dollars of equipment I'll re-order the same day. If a power hit kills a running job - it could hit the stock price.

Honestly. Get a clue.

And while you're at it, do some research on high PFC systems shutting down with dirty UPS power.

 

[westom]

I agree completely - and feel that you fail this test. The topic is protecting data - particularly non-persisted data in memory that is being processed.

Please admit you did not read the OP's post. He asked for a UPS to accomplish two things.

First:

I want to shut off the power to external HDs when the Mac Pro is in sleep mode.

That is obviously only protecting hardware. Power off does not harm any data on a disk drive. So he is asking for hardware protection.

Second:

Sinewave output:

Sine wave does nothing to protect data. So again, he is asking for hardware protection. He obviously is not asking to protect data.

UPS only protects unsaved data. It does nothing to protect hardware (ie hard drives) and does nothing to protect saved data. Sine wave is clearly only about hardware - not data. Please learn the topic. Clearly the OP did not ask to protect data. Only AidenShaw made that erroneous assumption.

Please get a clue before posting your emotions. Apparently AidenShaw does not have the necessary computer knowledge to make an informed conclusion. Now he can end cheapshot insults by posting in an adult manner. Meanwhile only AidenShaw cannot understand what the OP requested. He can stop all cheapshot by posting as an adult. And learning how computers work before making a conclusion.

 

[AidenShaw]

Any of those UPSes (without or without a cleaner sine wave) is just fine for the OP. Since a UPS has one function: temporary and 'dirty' power so that unsaved data can be saved.

Hmmm. So, is he asking for hardware protection, or saving data? You seem to be saying both, while denying both.

And you miss a very important third case - one that is most important for me. Most of our power outages last for a few seconds, not minutes or hours. (The power grid has lots of circuit breakers on high voltage lines that trip during a fault, and then automatically reset after a few seconds. If the fault was transient - the end user sees zero voltage for a short time, then back to normal.)

I get APC UPS systems to allow the computer systems to basically ignore these brief "auto-reset" interruptions.

• I don't have generator backup for the UPS, so if the power doesn't come back in 5 to 30 minutes the battery goes dead and the systems lose power.
• I don't have people watching the systems 24x7 - so there's no one to manually save the data and shutdown the system. (Actually, a few critical systems are tied into the APC monitoring systems and will save and shutdown when the UPS gets to 5 or 10 min of power remaining.)
• The vast majority of interruptions are very brief, so the UPS keeps the systems running without any issue of saving data.

A couple of examples where the APC systems have prevented any lost work/data from complete (but brief) power outages.

• A tree rat was playing in the overhead lines, and made the mistake of touching two wires of the local high voltage 3 phase feed. A "pop!" and a cloud of squirrel-flavored smoke - and about 5 minutes before things reset and the power came back to about 2000 hectares of Silicon Valley's finest.
• For a number of months we had a brief outage a couple of times a month - without any visible cause. This was escalated to the power company, and they installed serious monitoring on the local neighbourhood grid. The problem was in the local substation - a power factor correction system was bad, and was allowing the phases of the 3 phase distribution to get out of sync and trip the safety systems.

But, go ahead and call me names and question my intelligence, experience, and maturity. I don't care what you think about me.