This little project / exercise started as an identified lack of key critical knowledge in various areas of high speed electronics as well as a prerequisite for future SMPS development.Below is an attempt at a project presented in 2004 by Eva, an inductor saturation test fixture, due to the missing images and PCB layout the only design trail left behind was designation component numbers and component values with guidelines and after many failed request to re upload the project I came up with this (I guess to the experienced this is an easy task).
The project was built around a TL494 PWM controller with dead time sweep (1% to 90%) and frequency sweep (10Khz to 300Khz) with mosfet local turn off with an output of 400mA running in single ended mode (as per the datasheet specification).
The design was carefully layout out on two veroboards (signal and power) and I kept all the high current control loops inside a small area. Some issues with the design was anticipated off the shelf junk box components resulted in some “sampled noise” during measurements as a result of using inductive power resistors for the current shunt and “inductor energy reset”.
Power supply source from a modified old ATX power supply, supplying 13.8V @ 10A.
Shows the response from an Arnold core (A-085035-2 MPP -26 material) under test VGS vs current ramp, note the sampled ramp contains parasitic noise as a result of the inductive components. I used a FPQ50N06 and it worked very well and remained rather stable before spike tame and after taming the device ran cool with low heat dissipation.
Vgs close up and under a 2.2ohm / 10watt load with enforced heat-sink with 400mv over the shunt @ 12v, resulted in 0.400/0.1R = 4amps and
PD = 4^2 * 2.2 = 35.2 watts)
Shows the enforced heat-sinking before power resistor fail.
The project was built around a TL494 PWM controller with dead time sweep (1% to 90%) and frequency sweep (10Khz to 300Khz) with mosfet local turn off with an output of 400mA running in single ended mode (as per the datasheet specification).
The design was carefully layout out on two veroboards (signal and power) and I kept all the high current control loops inside a small area. Some issues with the design was anticipated off the shelf junk box components resulted in some “sampled noise” during measurements as a result of using inductive power resistors for the current shunt and “inductor energy reset”.
Power supply source from a modified old ATX power supply, supplying 13.8V @ 10A.
Shows the response from an Arnold core (A-085035-2 MPP -26 material) under test VGS vs current ramp, note the sampled ramp contains parasitic noise as a result of the inductive components. I used a FPQ50N06 and it worked very well and remained rather stable before spike tame and after taming the device ran cool with low heat dissipation.
Vgs close up and under a 2.2ohm / 10watt load with enforced heat-sink with 400mv over the shunt @ 12v, resulted in 0.400/0.1R = 4amps and
PD = 4^2 * 2.2 = 35.2 watts)
Shows the enforced heat-sinking before power resistor fail.
