Some questions about Electronic transformers used to drive Halogen Lamps

[powerelec]

Hi,

I have a question about electronic transformer circuits that are used to drive halogen lamps. The questions are based on the attached document on the subject.

Regarding this document in Figure 1:

1) Is the self oscillating frequency constant? It seems that as the rectified AC input voltage changes, the voltage seen across the series inductance of (T1 + Primary_of_lamp_transformer) will increase. This will cause the current slope di/dt to be different every cycle, and the time required to saturate T1 will be different every cycle thus causing a varying frequency. But the author of the document seems to indicate that the frequency of the oscillation would be fixed at 35 KHz.

2) Any idea what typical values for the inductance of T1 and lamp transformer primary should be? Also what should be the ratio of saturation current values for the lamp tranformer primary and for T1?

3) With reference to the .png image I attached what will be the wave shapes of the voltages at node A and B and the currents through T1 and lamp during normal operation? Would greatly appreciate if someone could sketch these out.
I have trouble to visualize what the current shapes especially would look like.

Thanks for the help!

 

[chuckey]

According to the article the voltage could be 6,12 or 24 V (most likely 12 as its in the title) and Pout = 50 W. One point to remember that for a square wave where Vrms = 12V, the actual square wave peak voltage = 12 X 1.4 ~ 17.6V.
Frank

 

[powerelec]

Chuckey, your post is not exactly answering the question. Or did I not catch your meaning?

Another question related to electronic tranformers:

1) If there is no halogen lamp connected on the secondary, will the electronic transformer still have oscillations? I have read a few articles saying that electronic transformers need a certain minimum load on the secondary to sustain oscillations. Why is this required?

Thanks for the help!

 

[chuckey]

According to the article the feedback transformer (T1) must saturate for the oscillations to actually occur. As the output current flows through the transformer, its the output current (or power) that causes this.
Frank

 

[FvM]

None of your question can be easily answered because the answers depends more or less on the chosen design parameters. The output transformer can be e.g. designed either as a resonant transformer with sine waveform at node B or as non-resonant transformer, depending on the capacitors and transformer leakage inductance,

Answering the questions means finishing the design of a circuit, that is only given as a basic toplogy without any dimensioning. You'll possibly find complete cicruit schematics from other sources, or should do the design work yourself.

 

[kripacharya]

One point to remember that for a square wave where Vrms = 12V, the actual square wave peak voltage = 12 X 1.4 ~ 17.6V.
Frank

small correction - for a Square wave, the Vrms = Vpeak.

 

[chuckey]

What you say is correct for when the pulse is present, but it is not there for 50% of the time. So while the pulse is there it has to delivery double the power, because its going to be missing for the next time period. For example, 12V DC, Vp = Vrms. Now chop this up into a squarewave. So 12V for T, 0V for T, so power delivered = 12^2/R X T + 0^2/R X T, so over a 2T time period, power delivered = 12^2 X T /2T = 12^2/2 i.e. half the power because the 12 V is only there for 50% of the time. So you need 12V X 2^1/2 to make the previous sum = 12^2/R.
Frank