6.5W in standby!

[tom66]

Ouch! When I saw that I was really surprised. What could be eating so much power? Is the scope getting any standby power for features like wake on LAN? 6.5W per hour works out to 1kW per 153.8 hours, if you put that into a year of standby (ignoring the time it is on, of course, where it uses ~60W), it works out to 57kWh, or about £7.41 per year where I live... not much, but not something you can ignore.

I like my HP 54501A, because it has a true power switch. The switch is actually embedded into the power socket and has a long plastic lever, like the one in the Agilent, going to the front for the power switch, and I know it's a real switch because there are only three wires exiting the socket (live, neutral, earth.) Why not just make it a tactile switch and make it wake up quicker? If it already draws 6.5W in standby I can't think it would take too much to put the processor and the rest of the board in a low power sleep mode so it can wake up in a few seconds compared to the 30 seconds for Win CE to boot up...

[Zero999]

Unless it has features such as a timer and wakeup, it's probably laziness - it's easier not to route cables with insulation thick enough for mains round the 'scope case to the front panel.

[Rufus]

Ouch! When I saw that I was really surprised. What could be eating so much power?

It uses a 13v OEM version of this PSU module .
http://www.lineagepower.com/oem/pdf/cch125.pdf

The scope power switch despite being mounted right next to the mains input filter controls the 'REMOTE' input of the PSU. All the 'standby' power dissipation is down to the PSU module.

I wouldn't take 6.5W as gospel, it doesn't feel like quite that much to me. Hard to measure low standby power accurately. Anyone have a scheme for that, preferably one that doesn't require a scope - lol.

[tom66]

Ouch! When I saw that I was really surprised. What could be eating so much power?

It uses a 13v OEM version of this PSU module .
http://www.lineagepower.com/oem/pdf/cch125.pdf

The scope power switch despite being mounted right next to the mains input filter controls the 'REMOTE' input of the PSU. All the 'standby' power dissipation is down to the PSU module.

I wouldn't take 6.5W as gospel, it doesn't feel like quite that much to me. Hard to measure low standby power accurately. Anyone have a scheme for that, preferably one that doesn't require a scope - lol.

Hmm interesting, perhaps that explains all the secondary power supply stuff on the mains input board. Seeing all those TO-220's gave me the idea that it had more than just a 12V or 13V rail (was expecting at least 5V and probably a negative rail too.)

[sacherjj]

Ouch! When I saw that I was really surprised. What could be eating so much power?

It uses a 13v OEM version of this PSU module .
http://www.lineagepower.com/oem/pdf/cch125.pdf

The scope power switch despite being mounted right next to the mains input filter controls the 'REMOTE' input of the PSU. All the 'standby' power dissipation is down to the PSU module.

I wouldn't take 6.5W as gospel, it doesn't feel like quite that much to me. Hard to measure low standby power accurately. Anyone have a scheme for that, preferably one that doesn't require a scope - lol.

An easy way to measure this kind of stuff is with a Kill-A-Watt power meter.  It can measure an AC device for power over time.

[EEVblog]

I wouldn't take 6.5W as gospel, it doesn't feel like quite that much to me. Hard to measure low standby power accurately. Anyone have a scheme for that, preferably one that doesn't require a scope - lol.

The Gossen meter is perfectly capable of uW measurements. Believe it.

Dave.

[Rufus]

The Gossen meter is perfectly capable of uW measurements. Believe it.

You didn't say what you measured it with, and now only that it is a Gossen, so I assume a Metrahit energy.

I looked at the data sheet for that and couldn't figure out what it was able to measure taking frequency and harmonic content into account, although I expect it would probably read low for something it couldn't properly handle.

Anyhow, I stuck 90, 100, and 110v dc into the scope and measured 5.16, 5.31, and 5.33 watts so yes now I believe 5 or 6 watts rather than the 2 or 3 I guessed at from how warm it got.

[EEVblog]

The Gossen meter is perfectly capable of uW measurements. Believe it.

You didn't say what you measured it with, and now only that it is a Gossen, so I assume a Metrahit energy.

I looked at the data sheet for that and couldn't figure out what it was able to measure taking frequency and harmonic content into account, although I expect it would probably read low for something it couldn't properly handle.

Anyhow, I stuck 90, 100, and 110v dc into the scope and measured 5.16, 5.31, and 5.33 watts so yes now I believe 5 or 6 watts rather than the 2 or 3 I guessed at from how warm it got.

It's all in my latest video.
Yes, I used the Metrahit Energy.

I have another Silicon Chip power meter kit that also goes down to 0.01W resolution with reasonably accuracy.

Beware some of the cheap Kill-A-Watt type devices, they can't measure low powers accurately.

Dave.

[alexwhittemore]

This made me want a Metrahit Energy, so for giggles I looked it up on Conrad's website. HOLY CRAP, 970 EUR?! DAMN!

[Rufus]

It's all in my latest video.

I hadn't seen the video - wondered why the OP started a new thread - lol.

There is a service guide you can download which tells you how to check the mains input, the power switch, and the voltages coming off the filter board.

There are a lot of similar 3x5" power 'bricks' available. Couldn't find one that specified disabled or no-load power consumption.

[AdShea]

Though it would void your warranty it doesn't look like it would be too hard to swap out that switch to open the mains instead of running that remote pin.  Thankfully Agilent already gave you a nice post to the front panel for the button.

[EEVblog]

This made me want a Metrahit Energy, so for giggles I looked it up on Conrad's website. HOLY CRAP, 970 EUR?! DAMN!

Err, yeah, Gossen meters aren't cheap at the best of times, but quite simply the best meter I've ever owned.
Try their 6.5 digit multimeter.

Dave.

[EEVblog]

It's all in my latest video.

I hadn't seen the video - wondered why the OP started a new thread - lol.

There is a service guide you can download which tells you how to check the mains input, the power switch, and the voltages coming off the filter board.

Yeah, some people watch the Youtube channel were the video gets posted first, sometimes 24hours or so before it makes it to the main eevblog.com site and feed.

Where is this service guide?

Dave.

[jahonen]

I can confirm that MetraHit Energy measures even idle consumption of mobile phone charger with ease, likewise it handles non-linear loads just fine. Anyway, measuring non-linear loads should not be rocket science nowadays anyway.

Regards,
Janne

[mikeselectricstuff]

I really don't see what the fuss is about.
Their solution is going to be more reliable as you don't have a switch handing inrush currents or switch-off arcs, and there is no inrush stress on the PSU.
Mains-rated fully approved double pole mains switches are pretty large, and would have been a struggle to fit in the space on that board while maintaining clearances to the shield.

There will never be enough of these scopes in the world for the wasted energy to be an issue, and the cost to individual owners is negligible compared to other stuff that people leave on all the time.

My only minor concern is the physical robustness of that switch, but it's not  going to be a big deal to find a replacement if it eventually falls apart - finding exact replacements for a mains switch could be a lot harder. 

[hacklordsniper]

Dave since you hace contact with Agilent, did you try to ask them about their follow up?

[Fryguy]

Dave mentioned in the video that this is not the 70's - Dave , you're exactly right about that !
My vintage 70's Tek 7603 Scope which i still use has got A REAL MAINS POWER SWITCH ! 
You push it and it draws zero .
My 35 year old Tek uses a lot more power than a brand new Agilent and the switch still works perfectly after all these years - it can't be that hard to find a suitable power switch for a small SPSU .
And when i see the price you have to pay for these Agilent scopes , a real mains power switch should be well inside the budget . . .

[comox]

An expansion on my post on your blog....

Nice one again!  I am obsessed about standby power consumption and have a couple of Yokogawa WT210 power meters which I use to analyze appliances and such, which are very accurate down to the 5ma range.  I have a Tektronix MSO4032 which has a big mechanical pushbutton switch on the front which I ASSUMED isolated the power supply circuit from the mains when off.  After viewing your blog I hooked the MSO4032 up to the WT210 and here are my observations:

MSO4032 on: 238V (UK Mains), .361A, 77.54W   (Mode RMS)

MSO4032 standby: 241.92V (UK Mains), 3.57ma, 0.0007W (Mode RMS)

Pics attached.

Sorry Dave, but I am not going to take my scope apart and look at the PSU circuitry, however I imagine that there may be some line filtering circuitry of such (e.g. MOVs) which result in some sort of measurable leakage current.   The Yokogawa WT210 is an approved meter for Energy Star compliance and can be trusted to measure standby power consumption, and cost about £3000 each.  (Amazingly the second unit was purchased on eBay for about £600.)

I agree that 6W standby is out of order in today's times.  We are all fully aware of the environmental impact of energy consumption, and companies should incorporate features such as mains isolation into their product designs.  Unlike a PC, TV or a set-op box, there is no reason to have a scope on standby.

I have gone as far as measuring the standby power consumption of supposed "green gadgets", such as a power strip that would shut off devices (e.g. DVD, set-top box) when your TV is turned off (i.e. UK Dragon's Den gadget from a few years back), and was saddened to discover that the cheap PSU that powered it drew 1W in standby, which was more than my TV consumed in standby!  Oh well..........

Comox

[mikeselectricstuff]

I agree that 6W standby is out of order in today's times.  We are all fully aware of the environmental impact of energy consumption, and companies should incorporate features such as mains isolation into their product designs.  Unlike a PC, TV or a set-op box, there is no reason to have a scope on standby.

This is the sort of politically correct greenery that really annoys me. Standby consumption is only an issue in products like TVs, phone chargers  and set-tops because there are so many of them. It is a totally insignificant issue for niche products, as there just aren't enough of them to ever have any significance.
Long-term, the effect of product standards in high-volume markets will trickle down in the form of improved controller chips etc., and there will probably come a point where all off-the-shelf PSUs will have good standby and no-load performance, but there's no real justification to spend more money on it now.

And don't forget that a significant proportion of standby and active wasted power will be partly offset in any environment that is normally heated.   

[EEVblog]

Nice one again!  I am obsessed about standby power consumption and have a couple of Yokogawa WT210 power meters which I use to analyze appliances and such, which are very accurate down to the 5ma range. 

That's a nice bit of kit!
Thanks for the measurements.

Dave.

[EEVblog]

This is the sort of politically correct greenery that really annoys me. Standby consumption is only an issue in products like TVs, phone chargers  and set-tops because there are so many of them. It is a totally insignificant issue for niche products, as there just aren't enough of them to ever have any significance.
Long-term, the effect of product standards in high-volume markets will trickle down in the form of improved controller chips etc., and there will probably come a point where all off-the-shelf PSUs will have good standby and no-load performance, but there's no real justification to spend more money on it now.

And don't forget that a significant proportion of standby and active wasted power will be partly offset in any environment that is normally heated.   

The actual numbers and whether or not it matters is pretty moot.
The point is that it's just generally poor engineering, and doesn't help drive the industry in the right direction IMO.
If Agilent insisted on a low standby supply, then they help create need for better low standby supplies, and at the same time lower the demand for high standby supplies, thus helping drive the industry forward.
As designers we all have a part to play in this, it's just poor form for such a reputable company (or any company for that matter) to rely on others to do this and think it's their problem and it will "eventually trickle down".

Dave.

[comox]

This is the sort of politically correct greenery that really annoys me. Standby consumption is only an issue in products like TVs, phone chargers  and set-tops because there are so many of them.

Okay, lets look at it in terms of beer.   6W standby for 1 year @ 13p/kWH = 2 pints of beer!

Now are you convinced that standby power consumption is a problem?

[Rufus]

Where is this service guide?

http://www.google.com/search?q=Service+Guide+for+2000_3000+X-Series+Oscilloscopes

Will find it. Agilent site links are a bit too horrid to post.

This is the sort of politically correct greenery that really annoys me. Standby consumption is only an issue in products like TVs, phone chargers  and set-tops because there are so many of them. It is a totally insignificant issue for niche products, as there just aren't enough of them to ever have any significance.

If 1% of a household's power consumption is 'wasted' standby power the (in)significance is 1%. There being one household or 100 million the (in)significance is still 1%. The only bearing on cost benefit analysis is how many units development costs can be spread over.

Agilent screwed up. The problem is the mains switch. All reliability arguments about keeping the PSU warm or inrush currents are invalid because any problems caused by a real mains switch are also caused by turning the scope on and off at the wall socket.
 

[tnt]

IMHO:
* Should agilent have put a proper mains switch: Hell yeah, they should know better
* Will I resell my scope or not recommend it or loose sleep over it: Hell no ...

[mikeselectricstuff]

This is the sort of politically correct greenery that really annoys me. Standby consumption is only an issue in products like TVs, phone chargers  and set-tops because there are so many of them. It is a totally insignificant issue for niche products, as there just aren't enough of them to ever have any significance.

If 1% of a household's power consumption is 'wasted' standby power the (in)significance is 1%. There being one household or 100 million the (in)significance is still 1%.

It may have 1% significance to that household, but negligible in global terms, and also negligible in terms of affordability to any household that owns a scope.

Agilent screwed up.

That implies an unintentional error. I doubt very much this decision was unintentional - more likely a business decision based on cost/benefit and/or availability of suitable PSUs. Every product is a compromise between cost and performance. This is simply not an important performance issue for a product like this. 

The problem is the mains switch. All reliability arguments about keeping the PSU warm or inrush currents are invalid because any problems caused by a real mains switch are also caused by turning the scope on and off at the wall socket.
 

Except that it gets turned on at the switch a lot more often than at the wall. The issue isn't so much reliability of the PSU, but the switch under inrush conditions.

My bet is they just didn't have space for a suitably robust mains switch (including track creepage distances to route it on the PCB and clearance to shields). Maybe the extra EMC filter on the power board was a late addition which used up space originally intended for a switch.

It wouldn't surprise me if it's actually quite hard to find a PSU like this with very low standby draw- last time I looked it was hard enough to find one with remote on/off - the one linked earlier doesn't even spec it at all. Low-energy directives are increasing the availability of more efficient PSUs, and a look at websites of people like Power Integrations will leave you in no doubt that efficiency and standby draw are headline features of new PSU semiconductors. It takes quite a while for these to percolate down to the market, and PSUs have high approvals costs which means they don't want to change designs if they can avoid it.

[MrPlacid]

Wanna save electricity and not waste time invalidating your warranty, everyone? Use a multiple outlet strip and your problem is solved. Actually, all of you should be using one on your bench and make sure to turn everything off when you're not working on anything.



This is also how I combat wall warts. I even have loads of small single switches that I use for other stuffs.

[saturation]

Yep, simple $1 solution.  I use surplus surge suppressors and old style obsolete PC power stations popular in the <= 1990s which are good for instrument stacks; just watch out for the magnetic field it can create.



Since so many devices or various vintages have phantom power in so many flavors, just shutting down the entire bench is a simple solution.  If you're doing a design that requires 24/7 monitoring data collection, well its powered up and not on standby, so I guess that doesn't count.

Wanna save electricity and not waste time invalidating your warranty, everyone? Use a multiple outlet strip and your problem is solved. Actually, all of you should be using one on your bench and make sure to turn everything off when you're not working on anything.



This is also how I combat wall warts. I even have loads of small single switches that I use for other stuffs.

[Rufus]

I can confirm that MetraHit Energy measures even idle consumption of mobile phone charger with ease, likewise it handles non-linear loads just fine. Anyway, measuring non-linear loads should not be rocket science nowadays anyway.

Accurately measuring power into non-linear loads is not rocket science but for 'difficult' loads is difficult and expensive.

Blind faith in any measurement is piss-poor engineering.

[jahonen]

Here are comparative figures for Agilent MSO6034A, it seems to have a real power switch.

Regards,
Janne

[Rufus]

That implies an unintentional error. I doubt very much this decision was unintentional - more likely a business decision based on cost/benefit and/or availability of suitable PSUs.

They absolutely screwed up. They deliberately designed the power switch with a plastic actuator rod, they cut away the main PCB for the rod, they cut away the shielding over the main board to allow the rod to pass through it. They mounted the switch right next to the mains inlet. It is blindingly obvious it was designed mechanically to support a switch switching mains power. If 'soft' power control was an original design intent there would be a push button on the main PCB instead of a cut out for the actuator rod.

When and why they decided to use the same mechanics to operate a switch controlling an enable to the power supply module we can only make guesses about. There was obviously some unforeseen problem with the original design intent and of course a 'business' cost/benefit decision produced the solution, but, it was an unintentional error/screw up which created the need for a solution. 

[jahonen]

I can confirm that MetraHit Energy measures even idle consumption of mobile phone charger with ease, likewise it handles non-linear loads just fine. Anyway, measuring non-linear loads should not be rocket science nowadays anyway.

Accurately measuring power into non-linear loads is not rocket science but for 'difficult' loads is difficult and expensive.

Blind faith in any measurement is piss-poor engineering.

That is of course true and it is always wise to question if one has performed valid measurements (and preferably to try obtain figures for comparison using different methods), but I don't think that usual AC power supplies are particularly difficult load, as fundamental (where the active power is delivered) is at low frequency.

Regards,
Janne

[mikeselectricstuff]

That implies an unintentional error. I doubt very much this decision was unintentional - more likely a business decision based on cost/benefit and/or availability of suitable PSUs.

They absolutely screwed up. They deliberately designed the power switch with a plastic actuator rod, they cut away the main PCB for the rod, they cut away the shielding over the main board to allow the rod to pass through it. They mounted the switch right next to the mains inlet. It is blindingly obvious it was designed mechanically to support a switch switching mains power. If 'soft' power control was an original design intent there would be a push button on the main PCB instead of a cut out for the actuator rod.

Not necessarily - it could just be they were simply keeping their options open. The mechanical and front panel design may well have been fixed long before they had finalised the power solution.
The way they did it gives them maximum flexibility - a lump of plastic is not an expensive part, so making provision for a rear switch,ande deciding later how to use it seems a very pragmatic solution.
If they'd designed a soft switch and then found they couldn't get a PSU with low enough 'off' mode draw within budget, that would have been a screwup.
Bear in mind that they may have been be planning for an internal battery option, which they could do without changing anything but the PSU. Again the decision on whether they would do this may well have come way after the mechanical design had to be frozen for tooling.
A company that spends millions of bucks developing ASIC-based top-end instruments does not make 'screwups' over something as trivial as a power switch.

[Neilm]

I have designed switch mode power supplies at work when I had requirements that could not be satisfied by an off the shelf supply. There are many modern ICs that can be forced into a standby mode with a logic input - when they are in standby they consume micro Watts of power.

The only question is does the scope still do things when turned off, for example if you have a network connection will it wake on LAN activity. This would require some power to impliment but I would have thought it would be much less than 6.5W

Neil

[DavidDLC]

That's why I have one of those:

http://www.tigerdirect.com/applications/SearchTools/item-details.asp?EdpNo=4823710&CatId=5313

[metalphreak]

We have one of those Monster Cable power conditioner boards (don't worry we didn't pay anything near retail price for it).

Was measuring standby power consumption of the hifi rack, and it was in the 100+ watt range 

So I started unplugging stuff to work out what was drawing all the power. Got down to 60w and I ran out of things to unplug. The power board, was consuming 60W! WTF? It's got a little voltage panel meter in it, and I don't know what else could be using all that power :| What a joke. It's like having a standard light bulb on all the time.

[comox]

So I started unplugging stuff to work out what was drawing all the power. Got down to 60w and I ran out of things to unplug. The power board, was consuming 60W! WTF?

WTF Indeed!  Are these things still on the market?

[EEVblog]

A company that spends millions of bucks developing ASIC-based top-end instruments does not make 'screwups' over something as trivial as a power switch.

I'm inclined to agree.
I'm sure everything they did had a reason behind it.

Dave.

[EEVblog]

The only question is does the scope still do things when turned off, for example if you have a network connection will it wake on LAN activity. This would require some power to impliment but I would have thought it would be much less than 6.5W

No, it doesn't. It switches all the PSU outputs off, there is nothing left going to the main board.

Dave.

[Neilm]

The only question is does the scope still do things when turned off, for example if you have a network connection will it wake on LAN activity. This would require some power to impliment but I would have thought it would be much less than 6.5W

No, it doesn't. It switches all the PSU outputs off, there is nothing left going to the main board.

Dave.

I remember a product I worked on that pulled a lot of power when turned off. It turned out that manufacturing was fitting a load of components that were marked do not fit - originally they were to allow power an emulator which alloed us to debug the board during development. I wonder if it is something similar.

Neil

[Lawsen]

Dave Jones:
I cannot afford such an expensive oscilloscope.  I do not own an oscilloscope like this.  Stand by power supplies were used commonly in RCA color television sets in my childhood era.  I guess there is memory back up, clock, or calendar or as you mentioned, "instant on," feature.  Video cassette recorders have to be on all the time to record television programs, when you are not at home to view it.  The VCR in the early days allowed you to not missed a television show or program (programme) by recording it by time.   Agilent admits to this always on set up in their instruction manual in pdf.  

http://cp.literature.agilent.com/litweb/pdf/86100-90117.pdf

Lawsen

[bilko]

Maybe it helps temperature stability and prevents condensation forming in the scope 

[Zad]

Okay, lets look at it in terms of beer.   6W standby for 1 year @ 13p/kWH = 2 pints of beer!

Now are you convinced that standby power consumption is a problem?

6W * 24 * 7 * 52 = 52.416kWH
@13p = £6:81

I would start buying your beer somewhere else!

On it's own that looks no biggy, but if you have 10 in a lab, and they have a lifetime of 10 years, that is a lot of beer.

It must be a massively lossy inductor that is losing >6W on standby. Can't be helping reliability in hot conditions.

[justinc]

Dave,

Has Agilent responded to you in anyway over this issue?

[EEVblog]

Has Agilent responded to you in anyway over this issue?

Actually, no, they haven't.
It was mentioned that they look at it.
But given that it's not an issue as it was clearly a deliberate design decision, they probably see no need to response like they have for problems.

Dave.

[gregariz]

If you're like me... I'm getting forgetful... so I have a master switch in the lab that automatically turns on/off the lights.. if I see the lights on I know to turn the switch off at night... otherwise I used to frequently leave iron etc on

[Rufus]

Actually, no, they haven't.
It was mentioned that they look at it.
But given that it's not an issue as it was clearly a deliberate design decision, they probably see no need to response like they have for problems.

Not switching mains might have been a deliberate design decision but such a crap PSU design can't have been. I have a 30A 30V Manson bench power supply (like the one on the shelf in your vids?) which with the output set to minimum voltage only pulls around 8W and that is with the displays lit and fan running in the back.