http://dangerousprototypes.com/forum/index.php?topic=7218.msg74079#msg74079Here they are.
This is toroidal, JBC C105, 1/8 power:
X
There is no peaks, except really narrow negative one when the power is switched on. This is how it should be.
This is EI, on the same iron, 1/4 power (the leakage inductance is so large that it cannot power it with 1/8 power eny more):
X
Notice the HUGE negative peak when the power is switched on, and the HUGE positive peak when the power is switched off later. The second peak also adds some error to the measurement of the temperature, although this is eliminated in firmware with wave shaping.
So, again, USE TOROIDAL TRANSFORMERS with this project, please. You will save both me and yourselves much trouble.
Yes, the EI I am using for tests is REALLY NASTY one (in fact the original one from BK3000LF station), yours may look considerably better, but there is no EI transformer that can be even close to the HUGELY less toroidal transformer leakage inductance.
Sorry guys, but I cannot fight phisics. No one can.
P.S: The leakage inductance, measured:
- Toroidal: 3.76mH
- EI: 122.7mH (yes, one hundred and twenty two, point 7)
This is 33 times difference.
I may put a setting in the menu for not using partial power in the future if someone absolutely insists, but I still won't be able to guarantee that it will work with any EI transformer in the world.