dc_dc boost converter help

[jk1204]

hi
i am trying to make dc_dc boost converter...vin 5v(old compurt power supply) vout 20v...load 27ohm.....i have 3.3u inductor 15a & 28a saturation ...transistor c5200.. Diode RF307

the problem is i cant reach above 11v. with 33khz reach 11.4v@16% duty and then the voltage fall when duty more than 16%...i had tried different freq 27khz...55khz..atc.. the highest voltage i reached 11.5 then the voltage start to fall when increasing the duty

what the problem?

 

[d123]

(...post a schematic if you can, people who understand what you're making will be far better equipped to help...)

 

[jk1204]

hi
i am trying to make dc_dc boost converter...vin 5v(old compurt power supply) vout 20v...load 27ohm.....i have 3.3u inductor 15a & 28a saturation ...transistor c5200.. Diode RF307

the problem is i cant reach above 11v. with 33khz reach 11.4v@16% duty and then the voltage fall when duty more than 16%...i had tried different freq 27khz...55khz..atc.. the highest voltage i reached 11.5 then the voltage start to fall when increasing the duty

what the problem?

 

[BradtheRad]

With a boost converter, there is a point where increasing the duty cycle does not increase output, but reduces it.

Tips:

For a 3.3uH inductor, try a frequency of 400k.

Expect current waveforms to peak as high as 5A.

The first half of the cycle (the power input part of the cycle) should have very little parasitic resistance. It should be less than 1/20 ohm.

Your smoothing (output) capacitor must endure several Amperes going back and forth through it.

Also consider making twin interleaved boost converters.

 

[jk1204]

hi BradtheRad..i had tried 250khz and 500khz but the the vout fall and never rise..my inductor 15a & 28a saturation.i think that more enough to handle peak current and the resistance 10mohm...i am using electrolytic capacitor 25v 4500u and tried 100u & 1000u with no result. may be the diode not recover fast? in diode datasheet reverse recovery time 500ns @ 0.5A..i am not sure

 

[d123]

How much current can the old computer power supply output? I don't know how to make a boost converter, but maybe if the circuit is drawing/trying to draw a high amount of current (i.e. above the rated maximum for the power supply) then perhaps the power supply output voltage is dropping in response and the converter can't reach 20V - ??
Probably not the cause of the problem, just a suggestion.

 

[jk1204]

computer power supply output 5v 20A

 

[BradtheRad]

For a useful way to pass the time, I like to experiment with simulations.

This screenshot shows the main area of the action. Using your specs:



The inductor is 3.3 uH. To obtain the typical waveform through the coil, the switching speed is around 400 kHz. (See label below top right scope trace.)

The diode will need to be a fast-recovery type.

It is possible to run at a slower speed, but that would require the current waveform to climb to a much higher peak value. Your system sounds like it can handle the power. However parasitic resistance will then take an even greater bite out of performance. Your output may not be capable of 20V into a 27 ohm load.

- - - Updated - - -

Also notice the smoothing capacitor can be just a few uF (since the frequency is high). It is only theoretical because then it must carry several A back and forth. That is unrealistic for a little electrolytic. Therefore a gang of capacitors is a good idea.

 

[KlausST]

Hi,

Standard electrolytic capacitors are not good for high frequency switching applications.
Be sure it is specified for your frequency, current ripple and has low ESR.
Maybe you should add a ceramics capacitor in parallel to the electrolytic. At least 1uF.

Klaus

 

[FvM]

I don't believe that the electrolytic capacitor is the problem here.

But it's impossible to determine it with the scarce information given so far. Nothing has been said how the transistor is controlled. My first guess is insufficient base current, causing desaturation above a few amperes collector current. This would well explain the observed behaviour.

 

[betwixt]

If the driver really is a 'HC08', assing that means 74HC08, it would explain the lack of base current!

Also note that according to Fairchild's data sheet, at 8A collector current the worst case saturation voltage can be as high as 3V.

Brian.

 

[Easy peasy]

Yes, replacing the xtor with a mosfet would most likely cure the problem, plenty of high current (peak) base drive needed to drive a xtor properly, plus the standing current at a gain of 15....

 

[jk1204]

ok ..i think the the problem will be from diode...or transistor...i will buy logic mosfet from digikey next week and some fast diode..but i am really think the problem from the diode.!

as i said i had tried 500khz..250khz..no results


here is my circuit

 

[betwixt]

The transistor would be my first target for improvement. Replacing it with a MOSFET will help but even the you need to be careful to drive the gate to high enough voltage and most importantly, with enough current to give the drive waveform a fast rising and falling edge. From the drivers perspective, the gate will look like a capacitor which has to be charged and discharged quickly so if you deprive it of current you will slow down the ramp up and down of the voltage across it and the MOSFET will spend more time in transit between fully off and fully conducting.

If you want to improve your present design with bipolar transistors, add a resistor (~100 Ohms) between the base and emitter of the 2SC5200 so there is a path for the base charge to dump to when the 2N2222A is turned off and add a resistor (~180 Ohms) in series with the base of the 2N2222A to prevent excess current from the PIC. At the moment you drive a 5V logic output to two series B-E junctions which will sink too much current for the PIC to handle.


Brian.