Questions regarding project in PE, Feb 2025

[PM3295]

I saw this project featured in the latest edition of Practical Electronics, which was posted in:
http://www.sonsivri.to/forum/index.php?topic=74.msg208310#new

In this issue, there is a practical project for a "Mains Power-up sequencer" to reduce inrush current when switching on devices simultaneous containing large mains frequency power transformers.(pg 62).

The author seems to claim that the least amount of inrush current will occur when the transformers are switched on at the zero crossing of the mains voltage, and the maximum will occur when switching at one of  the peaks. He presents some wave-forms to illustrate his claim. I have some doubts about the current wave-forms shown.  

It is well known that the volt-sec product will be less when switching at one of the peaks and have the best chance of preventing core saturation. Switching near the zero-crossing will double the volt-sec, with increased chance of core saturation.


This issue was discussed in length some time ago in this thread:
http://www.sonsivri.to/forum/index.php?topic=68555.0
  

[optikon]

I didnt read the article, while it may be true that transformers are "closer" to saturation at the peak voltage, that doesnt mean that they are entering saturation. The transformer could be designed to have adequate margin there. Its more likely the surge load is from AC/DC rectifiers turning on supply startup load current to the devices attached.

And then there is consideration for other types of reactive loads that may move that surging point to different areas of the mains waveform.... and then there is the topic of the mains waveform itself which can be quite distorted with non-monotonicity, flat spots, HF noise further complicating when the surge actually happens. Seems to me it could be difficult to decide ahead of time where that bad turn on location actually is. I would strive for a device that would allow for adjustment so it can be fashioned to the actual loads... seems easy enough to design IMO.

[PM3295]

I didnt read the article, while it may be true that transformers are "closer" to saturation at the peak voltage, that doesnt mean that they are entering saturation.

It is all about accumulated volt-secs. If you switch on at one of the peaks, you will only move half the distance on the BH slope from near zero. There may be some small offset due to residual magnetization in the core. If you switch on at the zero crossing you will move twice the distance, because you accumulate volt-seconds over the full half cycle, which is double the amount with switching at one of the peaks. Thus you run the risk of entering the saturation area of the BH curve, depending on the design limit of the transformer.

In this old video, the professor explains this very well.
https://www.youtube.com/watch?v=ZbWpc_TfMLM

[optikon]

I am well aware its all about the volt seconds, but I also know that substantial margin can be built into the transformer design by way of adding more turns than what may be needed for the nominal voltage-time product. If say, the number of primary turns were doubled, then you could run at either twice the peak V or for twice the time for a given volt-second product. Thats an extreme impractical example of course, but margin can and is built into mains transformers so, knowing how close its getting to saturation, one either has to know the details of primary design or measure the current waveform going into the primary looks like..

The transformer is presumably designed to support the *full* volt-seconds it will be exposed to in the main application.. and have margin.... which is why I think the statement
"Switching near the zero-crossing will double the volt-sec, with increased chance of core saturation." - is not justified.

doubling the volt seconds would be a scenario like installing a 110V transformer in a 220V circuit etc... nothing to do where it is switched at in the wavform.

[PM3295]

The transformer is presumably designed to support the *full* volt-seconds it will be exposed to in the main application.. and have margin.... which is why I think the statement
"Switching near the zero-crossing will double the volt-sec, with increased chance of core saturation." - is not justified.

doubling the volt seconds would be a scenario like installing a 110V transformer in a 220V circuit etc... nothing to do where it is switched at in the wavform.

Transformers are often designed to get the most volts/turn to save on copper costs. This will naturally result in higher flux swing extremes under normal conditions. You can make assumptions that it should have wide safety margins, but from a cost saving view that is not guaranteed and corners are cut to be price competitive with many transformers out in the wild. I have seen the core saturation problem in real lab tests with beefy transformers. It is normally not a problem with small transformers as the winding copper losses limit the maximum current.