Author Topic: Transformer quiescent current  (Read 382 times)

0 Members and 1 Guest are viewing this topic.

Offline radar_macgyver

  • Frequent Contributor
  • **
  • Posts: 448
  • Country: us
Transformer quiescent current
« on: October 24, 2020, 04:17:45 pm »
I extracted this transformer from a Yamaha home theater that developed a protection fault:

[attach=1]

It has a 40-0-40 secondary (with a 20-0-20 tap), plus a number of additional secondaries to derive +/- 12V, 6V and +/- 5V; VA rating unknown, but I estimate it's 200VA from the weight. I was hoping to use it to build a linear bench power supply. When testing it, I had it plugged into a Kill-a-watt which told me it was consuming about 10W, and about 40VA (PF of 0.25). I also tested the Kill-a-watt with a different transformer (toroidal, 50VA, 18V output) and the quiescent current was about 0.25W. Why it would consume so much no-load power?

Also, it has what looks like a weld bead between the metal frame and the laminations, same bead on both sides. Wouldn't this short the laminations together and increase eddy currents?
 

Offline jonpaul

  • Regular Contributor
  • *
  • Posts: 222
  • Country: fr
Re: Transformer quiescent current
« Reply #1 on: October 24, 2020, 04:57:19 pm »
Suggest to study "primary magnetizing current" and Eddy current loss.

Varies with construction and input voltage/freq.

Also the kill a watt metes are NOT accurate at high inductive/low power.

Jon
Jon Paul
 

Offline radar_macgyver

  • Frequent Contributor
  • **
  • Posts: 448
  • Country: us
Re: Transformer quiescent current
« Reply #2 on: October 24, 2020, 07:20:20 pm »
Some of the less expensive Kill-a-watt clones only measure reactive power, but mine does measure both reactive and real power, reactive was 38 VA, real is 10W. I verified the reactive power using two true RMS DMMs, I measured 120V and 0.32A.

I don't have differential probes for my scope, but I did try measuring voltage/current with a 25 ohm sense resistor and using the math function on the scope, the current waveform looked like a triangle wave, see attached. CH1-CH3 measures the voltage across the sense resistor. CH1-CH2 measures the voltage across the transformer primary.

Such a high magnetizing current for a 200VA transformer seems abnormal to me, hence my post. It could very well be due to eddy current loss, because of that weird weld bead across the core laminations. Never seen that before, but folks here are a lot more knowledgeable.
 

Offline akimpowerscr

  • Regular Contributor
  • *
  • Posts: 207
  • Country: be
Re: Transformer quiescent current
« Reply #3 on: October 24, 2020, 10:05:45 pm »
The intensity of the magnetizing current (purely inductive) depends on the inductance of the primary.

This inductance depends on the number of turns of the primary and on the reluctance of the magnetic circuit.

The reluctance of the magnetic circuit depends on the existence or not of an air gap, the area of the core and the chosen induction level, considering the quality of the magnetic sheet.

These are therefore essentially parameters of the transformer design.

But there is also another, and an essential one : the manufacturing cost !!!!!

A toroidal transformer has practically no air gap and therefore, has a large primary inductance and a low magnetizing current.

In a conventional transformer, the residual air gap depends on how the magnetic sheets are assembled.

It is difficult to say if a magnetizing current is too high, at most we can deduce if it is a good quality transformer or not.

I advise you to let this transformer operate at no load and at nominal voltage for an hour or more and check if its heating is reasonable because this current will generate losses in the copper of the primary.
 

Offline jonpaul

  • Regular Contributor
  • *
  • Posts: 222
  • Country: fr
Re: Transformer quiescent current
« Reply #4 on: October 25, 2020, 06:41:09 am »
Bonjour radar_macgyver

The comparative  measurements between the large EI welded transformer and a toroidal are quite reasonable!

The losses and magnetizing current depend on many variables, chosen during the  transformer design, eg  regulation, duty cycle,  efficiency, cost target.

Your large EI (probably 500...1000 VA rating) uses a weld  at the EI gap, to avoid any clamps or hardware. Low cost solution but increases losses!
 It was designed for high volume, lowest cost production, in an application where efficiency, Imag and eddy loses were not important.
Minimum Lpri= fewer turns P and S. steel is low quality normal silicon steel.

The toroid transformers are usually much better steel, lower losses, and have NO gap. MUCH more costly per watt!
Coupling P-S is tighter so Llkg lower.  It may have more turns per volt to avoid stauration.

Extreme Examples:

Large power transmission and distribution transformers: (25 kW ...1000 MW) Optimized for lowest possible core and copper losses, duty cycle is 100% 24/7, efficiency ~ 99 - 99.8%

Microwave Oven transformers, lowest possible  cost, very high volume production, only minutes of operation at ~ 1 KW, thus very high losses, efficiency unimportant~ 85%.

Finally as Vpri increases you can approach saturation at line peak, so current becomes non-linear and losses increase rapidly.

For better explanations, of transformer theory and practice, see old reference texts  like Smith - Magnetic Components and Hunt, - Static Electromagnetic Devices.

Just the ramblings of an old retired EE, been designing magnetics and power electronics since 1970s.

Enjoy,

Jon


Jon Paul
 

Online Kleinstein

  • Super Contributor
  • ***
  • Posts: 8038
  • Country: de
Re: Transformer quiescent current
« Reply #5 on: October 25, 2020, 10:09:51 am »
The transformers no load loss depends on the core material, lamination thickness and also the peak magnetization used.
40 VA , 10 W looks still reasonable for a transformer this size. It may not be the lowest loss one, more like a cost optimized one. Thicker lamination can increase the eddy current losses: 0.5 mm is more like low cost, 0.35 mm is more like used for low loss. As crude approximation the no load loss is proportional to the weight of the iron (if the same grade material and magnetization is used).

The losses with ring core transformers can be lower, because they can use higher grade grain oriented material (similar to what the large transformers use), that has lower loss. Still only 0.25 W sounds quite low for me. There may be some error from the meter at this low power.

Larger transformers get better and the small ones really bad. Below some 2 VA for EI transformers things start to turn ugly, with no load losses going up when the transformer gets smaller, up to the point of getting no load loss higher than under nominal load.

Transformers for welders are often very much made to costs, a little like microwave ovens. This includes using aluminum windings and using high magnetization despite of higher loss. In this case additional leakage inductance is also wanted.
 

Offline jonpaul

  • Regular Contributor
  • *
  • Posts: 222
  • Country: fr
Re: Transformer quiescent current
« Reply #6 on: October 25, 2020, 11:55:32 am »
Hallo agree with Kleinstein 100%

Most 50 /60 HZ laminations eg M6 are 0.35 mm thickness.

Indeed the toroids may be grain oriented or amorphous material.

Lamp ballasts and welders need current limiting and have a magnetic shunt design, thus high losses and inherent current limit.

Finally the low cost Kil a Watt are very unreliable at low current and low PF as in unloaded transformer Testing

We use a 0.1 to 1 ohm 1% resistor or current transformer eg Pearson 411.

Finally pro grade wattmenters like Yokogawa WT110 or WT200 series are fine at very low power and all phase angle loads.\\\Enjoy,

Jon
Jon Paul
 

Offline radar_macgyver

  • Frequent Contributor
  • **
  • Posts: 448
  • Country: us
Re: Transformer quiescent current
« Reply #7 on: October 25, 2020, 08:51:58 pm »
Thank you for the replies.

I ran the transformer no-load for a couple of hours, and it's quite warm to the touch. Given that this transformer came from a consumer electronics source, it was likely a cost optimized construction.

I guess my main reason for posting this was to see if the quiescent current was not normal for a transformer of this size. Looks like it is, so I could use it for something rather than recycling it.
 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf