Future of Harmonics in Data Centers and Office buildings 3

[rbulsara]

Is there any truth the rumours I have heard (its coming from a client and I can not directly challange it without a firm knowledge myself) that the harmonics issue in the data center and/or in office buildings will go away in next 5-10 years?

I am sure the power supply desing for comuters may have improved overtime, but are they contribuiting to minimize harmonics in the power system?

Which harmonics are more prevailant (triplen, 5th,7th, 11th 13th?).

Anyone thinks that applying K-13 transformers and double neutrals will be thing of the past anytime soon?

I will appreciate input on this subject.

 

[logbook]

New electronic equipment taking more than 75W is CE marked for harmonic current compliance, amongst other things. Thus in 5-10 years when all old computer kit has been flushed out of the system, the power factor of a bank of computers should be pretty good.

Mind you the earth current will still be an issue because at 3.5mA each, 100 workstations can generate a shed load of earth leakage!

 

[jraef]

From what I understand, there is a shift coming here in the US in that instead of measuring THD at the PCC and mitigating it, all new components will be required to address harmonics up front, thereby (theoretically) eliminating the need for total system mitigation. That basically tracks with what logbook said above, because that is the way is has been going elsewhere for a while nw, hence the CE requirements. This will eventually apply to everything with an SMPS.

 

[Skogsgurra]

The EU demands that all power supplies that consume more than 75 W and draw less than 16 A* from the mains shall have a current wave-form that stays outside a shape that looks like an olympic "prize-stair" (do not know the proper name of it, never stood on one) a certain percentage of time. That is a non-mathematical way of saying that the harmonics contents shall be low.

Many power supplies already have a sine current shaping input circuit. It is a rectifier and a switcher that uses the rectified voltage as a reference for the current delivered to the rest of the power supply. So, if voltage is sine shaped, the current will be sine shaped. And not only that, it will also be in phase with voltage, resulting in very low distortion and Power Factor = 1.00

Unitrode, Siemens, Linear Technology and TI (I think) have been doing this for many years now.

* There will be a standard for devices with more than 16 A sometime in the future. It may, in fact, already be there but escaped me.

Gunnar Englund

http://www.gke.org