Unloaded Power Transformer Current Draw/Janus-Aikido snipet

[Seraph]

I just put together John Broskie's Janus Feedforward (tubed) power supply with Aikido 12AU7&12BH7. I used three separate transformers for the B+ (460 V AC), heater supply, and the rectifier tube (5V) filament to test the combo. Absolutely quiet beautiful sound with the Russian 5Y3 but even better overall sound with a 5AR4. With the set up I ended up with a B+ around 250 V. Ripple almost non-existent at about 1.5 mV max down to 0.80 mV min. In trying to increase the B+ and supplying the rectifier filament voltage from one transformer to reduce clutter, I tested some old power transformers I had on the shelf. Of the three transformers I tested I found that all three draw anywhere from 25 to 40 ma under no load with mains applied to the primaries. All secondary windings checked out at the rated voltage but why the current draw when the tranny is not loaded? Sure that could be due to a leakage of some sort somewhere but is that necessarily an indication of a bad transformer or one that is going to be noisy or bound to fail? These are hefty transformers. One is a Paeco made for a Hewlett Packard tube regulated power supply. The one I used for the set up above is a cheapy. It does not draw any current that I could measure in the order of a milliamp.

Regards

[HollowState]

Originally posted by Seraph
Of the three transformers I tested I found that all three draw anywhere from 25 to 40 ma under no load with mains applied to the primaries. All secondary windings checked out at the rated voltage but why the current draw when the tranny is not loaded? Sure that could be due to a leakage of some sort somewhere but is that necessarily an indication of a bad transformer or one that is going to be noisy or bound to fail? These are hefty transformers.

As far as I know that quiescent current draw through the primary is normal do to core losses. The bigger the transformer and current rating, the greater the unloaded primary current. It also has to do with wire size, turns per volt used and VA rating. This is not an indication of a bad tranny. However, much higher currents coupled with heating and odor would be.

[boywonder]

Originally posted by HollowState


This is not an indication of a bad tranny. However, much higher currents coupled with heating and odor would be.

Which is usually quickly followed by a ever increasing plume of acrid, yellow smoke that lingers over the workbench for hours even after the transformer is unplugged.

[Seraph]

Thanks for the explanation Hollowstate.

The lingering yellow smoke you describe Boywonder, reminds me of when I tried to use my newly purchased solder pot without adequate ventilation for the first time. An open window was just not enough! I left the house in a hurry with two window fans placed in two windows luckily directly across each other with the fans in intake and exhaust. Was so glad spouse was not home. Since then enameled wire tinning is done with effective ventilation in place. Capacitor failure or other unforeseen mishaps notwithstanding, one should anticipate and try to always be ready for the unexpected. Best is to follow instructions and advice of others along the way.

[zigzagflux]

The small current draw is excitation current, consisting of a great percentage of VARs and harmonics in order to establish the magnetic field in the core. Only a small percentage is actually related to core losses in the form of watts.

For typical power transformers, the exciting current is less than 2% of the full load current. Small (1000 kVA or less) transformers is even less than that. The lower the excitation current, typically the better quality the design, influenced by the core material and geometry. Of course, transformers operated closer to their core saturation point will exhibit greater excitation current for a given applied voltage.

Third harmonic is the predominant harmonic in exciting current, consisting of about 40% of the total.

Less is better, basically, but some must be present, or you don't have a field.

[Seraph]

Thanks. That is one good bit of something good to know

[zigzagflux]

I should have stated, "Only a small percentage should be related to core losses in the form of watts." If excitation current increases over time, it may be that core losses are increasing, a bad thing. To get really exotic, you could resolve the phase angles of an FFT spectrum and see how much of the excitation current is reactive vs. real power. Ideally, this will not change over the life of the xfmr.