I'm in the process of testing a 2KV power supply using an LT3757 device. It has the basic sample high voltage circuit shown in data sheet (see below):
except with a voltage multiplier diode configuration after the transformer:
In the lab, with an open load at the output, load net Out3 measures 1.1KV instead of 2KV as predicted by simulation.
Feedback (FBX) pin reads .9V all the time, and never gets to 1.6V as expected. Is there something I'm missing for proper biasing the LT3757 part?
Again, simulation runs fine with open load conditions.
App note 39 is a good read:
You might want to try increasing the values of the RC that are connected to the SENSE pin.
Increasing the RC sense on the pin does not seem to change the 1.1KV output. Ideal output is 2KV in simulation. Is there another way to increase the blanking period stated on App39 for the LT3757 device instead of the LT1070?
In the simulation, if the transformer is not ideal, e.g. with some leakage inductance, the turn-on current spikes would probably affect the operation as well. Personally, instead of the voltage multiplier, I'd go with the stacked windings. Another good read:
Your suggestion is great. Maybe I can try this next time. However, since I have a breadboard with my original design, I need to make this work.
I found out that changing the frequency resistor on the LT3757 (Rt on page 11 of specification sheet) the frequency pulses driving the flyback transformer do not change change. For example, If Rt = 4.2K (300KHz operating frequency), I measure an average frequency between pulses of 95KHz. This is somewhat confusing. I've also tried other Rt values to change the operating frequency, and I still measure flyback driving pulse at 95KHz. This does not make sense.
Also, upon taking simple transformer input/output sinusoidal measurements over frequency, I found out that the transformer Q is 285KHz, and since pulse are being generated from the LT3757 at 95KHz, driving flyback frequency is inefficient to the transformer. So, how can I change the actual driving flyback frequency of the LT3757 part so that it will fall closer to the optimal transformer frequency (285KHz)?
Is there a simulation file for the stacked winding configuration using the LT8304-1?
For multiple winding configuration, use K L1 L2 L3 L4 L5 1
L1 can be the primary and L2,L3,L4,L5 can be the secondary.