Hello there,
about a year ago I purchased a pair of mains transformers for my Aleph J amp. According to the production data they must be from the same batch. Today I accidentally discovered that they are different, I mean the inductances are different.
The inductances were measured with a DER EE DE-5000 LCR meter at 100Hz. The values are as follows:
Transformer 1:
Transformer 2:
Although it is difficult to measure an iron core inductor, the error should be identical for both transformers. Is it not? Moreover, the measured inductances does not match the voltage ratio of the transformer (L1/L2=(V1/V2)^2).
How shall I interpret the results? Is there anything to worry about? Shall I try different measurement methods?
Both transformers are specified for 230V input and 2x18V output. I connected them and measured the output voltages:
Transformer 1:
Transformer 2:
As far as I can tell, if the inductance were different, the output voltages would be different. I am definitely missing something here.
P.S. I noticed that the "transformer 1" is clearly noisier. You can tell that it has more hum.
about a year ago I purchased a pair of mains transformers for my Aleph J amp. According to the production data they must be from the same batch. Today I accidentally discovered that they are different, I mean the inductances are different.
The inductances were measured with a DER EE DE-5000 LCR meter at 100Hz. The values are as follows:
Transformer 1:
230V input - 6500mH
18V out 1 - 121mH
18V out 2 - 119mH
Transformer 2:
230V input - 8900mH
18V out 1 - 190mH
18V out 2 - 188mH
Although it is difficult to measure an iron core inductor, the error should be identical for both transformers. Is it not? Moreover, the measured inductances does not match the voltage ratio of the transformer (L1/L2=(V1/V2)^2).
How shall I interpret the results? Is there anything to worry about? Shall I try different measurement methods?
Both transformers are specified for 230V input and 2x18V output. I connected them and measured the output voltages:
Transformer 1:
230V input - 228.7V
18V out 1 - 18.8V
18V out 2 - 18.9V
Transformer 2:
230V input - 229.3V
18V out 1 - 18.8V
18V out 2 - 18.9V
As far as I can tell, if the inductance were different, the output voltages would be different. I am definitely missing something here.
P.S. I noticed that the "transformer 1" is clearly noisier. You can tell that it has more hum.
Attachments
Core material will vary, so the inductance will vary.
Output voltage for a fixed input voltage depends on the turns ratio and ohmic losses due to load current,
and can drop as much as 30% or more from open circuit values in consumer grade transformers with full loading.
Output voltage for a fixed input voltage depends on the turns ratio and ohmic losses due to load current,
and can drop as much as 30% or more from open circuit values in consumer grade transformers with full loading.
The reading of inductance from a trafo does depend on the flux induced in the core. To get a comparable reading on both primary and secondary inductances you'd need to have the same magnetisation (H) in both measurements which isn't easy to arrange with a simple LCR meter.
An anecdote:
A cop on patrol is bored so he is thinking: "Two plus two is four, and two multiplied by two is also four. But three plus three is six, and three multiplied by three is not six, it is nine. How come?" Unable to figure it out, he calls his superior and explains the problem. "You dumba..", answers the superior, "your job is not to add and multiply, but to subtract and divide!"
A cop on patrol is bored so he is thinking: "Two plus two is four, and two multiplied by two is also four. But three plus three is six, and three multiplied by three is not six, it is nine. How come?" Unable to figure it out, he calls his superior and explains the problem. "You dumba..", answers the superior, "your job is not to add and multiply, but to subtract and divide!"
Does transformer 1 have some residual magnetism in the core? Not sure how you would test for it, other than seeing if it moves a compass needle (unpowered) or if there is second harmonic produced small signal.
All true, but if these are identical (production) transformers, and measured by the same meter with the same setting (assumed), they should only differ slightly, not 30%.
Jan
Have you measured the DCR of each winding? And what happens if you change polarity of your LCR meter? The reading shouldn't change but it's still interesting to verify.
I'm not sure the compass needle can detect the remanent induction, because the resistance of magnetic pathway too low inside the core.
The difference is caused remanent magnetization. It will probably change each time you connect/disconnect them because the instant is random wrt. to the mains waveform.
If you magnetize them in the same way, by connecting briefly a 1.5V battery to one of the secondaries, you will probably have more consistent readings
If you magnetize them in the same way, by connecting briefly a 1.5V battery to one of the secondaries, you will probably have more consistent readings
There could be an issue with measurement of inductances i.e. the short circuit test that does not give perfect answers. From the picture below, the secondary short circuit doesn't get rid of the magnetisation inductance Lm, and the measured leakage always has a reflected component from the other winding.
The peak magnetic flux (and saturation level) is clearly different for both pieces. Now, it could be permeability related, area related or turns related. Some extra energy is lost by magnetostriction and sound / vibration hum and quite likely as extra core losses.
| Factor | Impact on Core Loss | Potential for Loss Reduction |
|---|---|---|
| Core Material | Materials with lower hysteresis and eddy current losses lead to better efficiency. | Use of soft magnetic materials, grain-oriented silicon steel, etc. |
| Core Design | Laminated cores inhibit eddy currents; shape optimization reduces excess material use. | Strategic lamination and geometric design innovations. |
| Operating Frequency | Higher frequencies increase core losses exponentially. | Operational optimization to utilize lower frequencies. |
| Flux Density | Increased flux density enhances core losses. | Maintaining lower flux densities within design constraints. |
Last edited:
The other option to demagnetize the cores is to connect the primary to a variac, and slowly turn down the ac voltage to zero before disconnecting. toroids a known for residual magnetization, and subsequent huge inrush currents upon reconnect. I always joked to people complaing about inrush, that they switched off the unit at the wrong time.
It could have gotten magnetized a bit more than the applied line voltage normally does, too. Too close to a speaker magnet perhaps. That will leave a little residual that normal operation won’t remove. When it happens with tube output transformers the result is high 2HD on a push pull unit. Degaussing is still the same - but might require higher than 230V applied initially.
I applied 5VDC briefly to both transformers and this equalised the magnetisation of the cores. Now the values are as follows:
Transformer 1:
Transformer 2:
The conclusion is that it is not possible to measure the inductance of such a large iron core transformer/inductor with such an LCR meter.
Thank you all for your helpful hints!
Transformer 1:
4700mH
51mH
50mH
Transformer 2:
5000mH
47mH
47mH
The conclusion is that it is not possible to measure the inductance of such a large iron core transformer/inductor with such an LCR meter.
The DCR was the same for both transformers. Changing the polarity made no difference.
Thank you all for your helpful hints!