Which computer power supply produces little or no dirty electricity?

classic Classic list List threaded Threaded
23 messages Options
12
Reply | Threaded
Open this post in threaded view
|

Which computer power supply produces little or no dirty electricity?

black-sun
Antec NeoEco Gold Zen 600W produces up to 935mV of EMI according to TriField EM100.

I want to reduce it below 60mV. My cheaper power supply doesn't produce nearly as much EMI.

Which computer power supply do you recommend for low EMI?
Reply | Threaded
Open this post in threaded view
|

Re: Which computer power supply produces little or no dirty electricity?

flux
So far I've only found certain laptops produce a comfortable amount of EMI for me, which is 100mV or less. My laptops all run on 65W adapters and can idle around 10W. Older laptops with dual-core Intel "U" chips are tolerable for me, but quad-cores are not comfortable for me even if EMI is low. With some computers I've found that EMI produced scales with the wattage the machine is using at a given time, so EMI increases as CPU load increases. An Acer gaming laptop I tried produced 400-600mV EMI while idling at 15W, but quickly jumped to +1000mV once pulling 40W. On the two desktops I've tried recently that wasn't the case; EMI was high (+1000mV) whether at idle or under load and they're very draining to use them because of that. My Dell Latitudes measure around 100mV at idle or full load, which is ideal for me.

Laptops can use smaller power supplies, and the lower TDP chips means less watts needed. Less watts means less AC/DC conversion, which should produce less EMI, but maybe this isn't always true. This works for my ES, but it puts a cap on how much computing power I have access to.
Reply | Threaded
Open this post in threaded view
|

Re: Which computer power supply produces little or no dirty electricity?

Marc Martin
Administrator
Did you know in Windows that you can throttle the CPU?  I was once given a work computer that was giving me problems, so I throttled both CPUs to 50%, and that was better for me.  Of course, that meant that some things were running twice as slow, but so be it.

I think the next computer they'll be giving me at work may have 8 cores.  That'll be interesting.

Marc
 
Reply | Threaded
Open this post in threaded view
|

Re: Which computer power supply produces little or no dirty electricity?

davidrsnell
Here is an article on Power Throttling.  It may not work on older computers (doesn't on my old ThinkPad 140.)

 https://blogs.windows.com/windows-insider/2017/04/18/introducing-power-throttling/#:~:text=With%20%E2%80%9CPower%20Throttling%E2%80%9D%20%2C%20when,is%20spent%20on%20that%20work.

Marc, I sent you a message via forum email, please check your spam folder if you didn't see it in your inbox. The regular price of the supplement I mentioned is $49.95, this week only it's 30-40% off I think, must call to get the Special price.
Reply | Threaded
Open this post in threaded view
|

Re: Which computer power supply produces little or no dirty electricity?

black-sun
In reply to this post by black-sun
Throttling CPU or power doesn't do much.

I theorize that a less efficient power supply (80PLUS Bronze or below) dissipates a lot more EMI as heat than 80PLUS GOLD power supplies.
Reply | Threaded
Open this post in threaded view
|

Re: Which computer power supply produces little or no dirty electricity?

black-sun
In reply to this post by black-sun
Some people suggested that a better power supply topology can reduce EMI drastically.
Reply | Threaded
Open this post in threaded view
|

Re: Which computer power supply produces little or no dirty electricity?

Marc Martin
Administrator
In reply to this post by black-sun
Yeah, when I've custom-configured my Lenovo Tower PCs, I've always gone for the less-efficient power supply option.  As energy efficiency generally cause me issues.

But I've just ordered an energy-efficient ethernet switch, so I'll be testing out that theory yet again, soon.

Marc
 
Reply | Threaded
Open this post in threaded view
|

Re: Which computer power supply produces little or no dirty electricity?

black-sun
In reply to this post by black-sun
I bought a new power supply that is less efficient and has a built-in EMI filter.

However, the new power supply leaks more EMI than Antec NeoEco Gold Zen 600W.

It seems I need to buy a better mainboard that has at least 10 power phases(10 VRMs).

Each VRM has a choke that is responsible for filtering EMI.
Reply | Threaded
Open this post in threaded view
|

Re: Which computer power supply produces little or no dirty electricity?

sailplane
I just tried a Corsair RM850x because it's the best in category for EMI's on Tom's Hardware. on the Line EMI meter though, it's around 1000mV.. so it's not low at all.. with a video card, you'd get 1500mV.
Corsair HX1000i same thing,..
I have another 2 year old supply either 450 or 550 I will have to check details, that emits no noticeable difference on the Line EMI!.. Antec or Thermaltake, I'll check tomorrow. It seems the lower the wattage the less line EMI. Another Fractal 660P (Too smelly/horrible) does quite a bit better than the RM850x.
Reply | Threaded
Open this post in threaded view
|

Re: Which computer power supply produces little or no dirty electricity?

black-sun
Please test. I'm looking forward to your test results.
Reply | Threaded
Open this post in threaded view
|

Re: Which computer power supply produces little or no dirty electricity?

black-sun
In reply to this post by black-sun
I tested my old power supply which is known to drastically reduce EMI from my old computer.

I hooked it up with my new computer, and it reduced differential EMI from 1280mV to 120mV.
By using, an in-line EMI filter, differential EMI was further reduced to about 80mV.

Its maximum efficiency is 82%. I guess efficiency plays a role?

I'm going to order a low-efficiency power supply.
Reply | Threaded
Open this post in threaded view
|

Re: Which computer power supply produces little or no dirty electricity?

Marc Martin
Administrator
When custom ordering Lenovo towers, I've always gone for the low-efficiency power supply option.

So what exactly are you using for an in-line EMI filter?

Thanks,
Marc
 
Reply | Threaded
Open this post in threaded view
|

Re: Which computer power supply produces little or no dirty electricity?

black-sun
I bought this in-line EMI filter.



Reply | Threaded
Open this post in threaded view
|

Re: Which computer power supply produces little or no dirty electricity?

black-sun
In reply to this post by Marc Martin
I just ordered a newer model of my old power supply. It is sold by the same company that sold the old one.

Its maximum power efficiency is 83%, and it has passive PFC instead of active PFC.

According to https://www.electronicdesign.com/power-management/article/21801141/whats-the-difference-between-passive-and-active-powerfactor-correctors

> The passive techniques have certain advantages, such as simplicity, reliability and ruggedness, insensitivity to noise and surges, no generation of high-frequency electromagnetic interface (EMI), and no high-frequency switching losses.
Reply | Threaded
Open this post in threaded view
|

Re: Which computer power supply produces little or no dirty electricity?

black-sun
In reply to this post by Marc Martin
My new theory is that passive PFC creates less EMI, and passive PFC's lower power factor provides buffer for EMI from my computer.

power factor = power demanded by computer / power provided by power supply

Active PFC components promise power factor of 99%. Power factor of 99% means fluctuations in computer's power demand are reflected in fluctuations in power supply.

Passive PFC components usually have power factor of 70~83%, lower power factor means fluctuations in computer's power demand are reflected much less in fluctuations in power supply.

That's my theory.
Reply | Threaded
Open this post in threaded view
|

Re: Which computer power supply produces little or no dirty electricity?

black-sun
In reply to this post by black-sun
I further theorize that a motherboard with more VRM(voltage regulator module)s of higher qualities filter EMI from computer more effectively. VRMs filter EMI.

CPUs with more cores generate more EMI and thus need a motherboard with more VRMs of higher qualities for low EMI environment.

The theory is that more VRMs of higher qualities on motherboard + passive PFC power supply + in-line EMI filter = very little EMI.
Reply | Threaded
Open this post in threaded view
|

Re: Which computer power supply produces little or no dirty electricity?

black-sun
Scrap my theory. It turns out that my old power supply has active PFC components.

I will report test results tomorrow.
Reply | Threaded
Open this post in threaded view
|

Re: Which computer power supply produces little or no dirty electricity?

black-sun
In reply to this post by black-sun
A power supply that lacks active PFC components and has 83% maximum efficiency turns out to filter EMI even better than my old power supply which has 82% maximum efficiency and active PFC components.

Power supply with Lower efficiency in combination with passive PFC turns out to filter EMI better than any other type of power supply.

With the old power supply, EMI from my computer was reduced to 120mV.
With the new power supply, EMI from my computer was reduced to 45mV.
With in-line EMI filter, it was further reduced to 42~43mV.
With power supplies that have active PFC components, EMI from my computer was between 800~1200mV.

Remember low efficiency and passive PFC.

I dare you to reproduce my experiment.

I expect that a better mainboard with more VRMs may reduce EMI further.

However, there is still common-mode EMI which I guess can be reduced substantially by installing a sturdy common-mode choke at my house's main breaker panel. TriField EM100 can only measure differential-mode EMI in millivolts.

You could install common-mode chokes at various choke points? FDIC runs operation chokepoint to choke people into submission.
Reply | Threaded
Open this post in threaded view
|

Re: Which computer power supply produces little or no dirty electricity?

black-sun
In reply to this post by black-sun
I'd like to add that even if a power supply has active PFC components, low efficiency(~82%) filters out most EMI.

But, a passive PFC power supply with low efficiency filters even more EMI.
Reply | Threaded
Open this post in threaded view
|

Re: Which computer power supply produces little or no dirty electricity?

Marc Martin
Administrator
In reply to this post by black-sun
On April  5, black-sun [via ES] <[hidden email]> wrote:
> I bought this in-line EMI filter.

Ah, so you are saying that the quality of the power coming in from the AC line is impacting the quality of the power coming from the PC internal power supply.  I often wonder how independent/dependent these two things are.

BTW, I just got a new desktop tower PC from my employer while working at home.   Really awful.  A Dell Precision 3630.  Avoid at all costs.   So far I have found it helpful to move it away from me (longer cables, making it 9 feet away instead of the usual 3 feet), draping it in a grounding sheet (the kind you're supposed to use on a mattress), and surrounding it with RF absorption material.  I also found that putting sheet metal nearby made things worse, presumably it's reflecting some of that noise back towards me.

I still have some other things to try, but so far my groundrules are to NOT open it up and mess with things inside of it.

Marc
12