Boost converter circuit / Electromagnetic Projectile Accelerators / Forums

Forums

4hv.org :: Forums :: Electromagnetic Projectile Accelerators

« Previous topic | Next topic »

Boost converter circuit

1 2 next

Move Thread

LAN_403

j.azz

Fri Jan 30 2009, 12:51AM

Registered Member #888 Joined: Tue Jul 10 2007, 06:52PM
Location: Hannover, Germany
Posts: 40

Hi all!

Based on the bipolar totem-pole driver for mosfets from this paper, I designed the following boost converter for capacitor charging which will use a ÂµController for pulse generation and voltage control.
What do you think about the circuit? (see attachement) Suggestments, improvements? And first: Will it work?

greetz, j.azz

edit: I forgot the resistor in series with the mosfet's gate

Mattski

Mon Feb 02 2009, 06:27AM

Registered Member #1792 Joined: Fri Oct 31 2008, 08:12PM
Location: University of California
Posts: 527

I don't see a need for R1 there with your BJT push-pull circuit. It will slow down how quickly your high-side BJT can bring the MOSFET gate high. Is it in there to limit peak current through the BJT's?

Otherwise it looks like a fine boost circuit, and that resistor won't even stop it from working. You might want a capacitor decoupling your input supply voltage at the top of the inductor, depending on how much current it pulls and where the input voltage is coming from.

j.azz

Mon Feb 02 2009, 02:25PM

Registered Member #888 Joined: Tue Jul 10 2007, 06:52PM
Location: Hannover, Germany
Posts: 40

Thanks for the answer. I will try the booster as soon I gathered the parts. Also I modified the circuit, namely adding a gate current limiting resistor, removing r1, and adding a decoupling capacitor.

j.azz

Wed Feb 04 2009, 02:05AM

Registered Member #888 Joined: Tue Jul 10 2007, 06:52PM
Location: Hannover, Germany
Posts: 40

I now simulated the circuit and found that 5 volts aren't driving the mosfet satisfyingly (does that word exist?). So I tried the following circuit with level shifting mosfet driver but it is inverting, unfortunately. Though it works great, I don't want to take care of the pwm signal being held high when the desired cap voltage is reached. Are there any non-inverting level shifting totem pole mosfet driver? Or do you recommend using a 'cookbook' mosfet driver ic?

KLH

Wed Feb 04 2009, 07:14PM

Registered Member #1819 Joined: Thu Nov 20 2008, 04:05PM
Location:
Posts: 137

I tried simulating this circuit. It works okay, although there were some variables I could not account for, due to lack of simulation models or certain variables, such as series resistance.

Some improvements I suggest:

-Optimize the frequency for your 750uH inductor. Look at this graph:

The graph shows the peak current to be around 70A. This will likely asplode your MOSFET and diode, saturate your inductor, and dry out your capacitor bank, unless your capacitor bank and inductor have very high current ratings. (The two semiconductors inherently can't hold enough current.) To keep the peak currents from spiraling out of control, use this RL time calculator at http://ourworld.compuserve.com/homepages/Bill_Bowden/rl.htm. The current should be limited to about twice the rated inductor DC current and should stay within both semiconductor's ratings.

-Add a snubber from the inductor to ground. Look at this second graph:

After each voltage pulse, which appears when the inductor delivers its energy to the capacitor bank, you can see a thick section of the plot. These are high frequency oscillations, usually in the high kHz to MHz range. In your case, it's about 400 kHz. These oscillations cause core loss, and, less crucially, a long simulation time. The circuit below shows the addition of a snubber.

1233774795 1819 FT62678 Improved Boost Converter

This circuit has a frequency optimized for 8A peak current: 15kHz with 50% duty cycle.

As for the gate driving issue, a dedicated MOSFET driver IC is more reliable and simpler to use. After simulating and checking the gate driving current, it seemed to be only a few milliamps. (Am I doing something wrong?)

Mattski

Wed Feb 04 2009, 08:19PM

Registered Member #1792 Joined: Fri Oct 31 2008, 08:12PM
Location: University of California
Posts: 527

One easy way to get a non-inverting level shifter is by driving your existing circuit with an inverter.

Using a mosfet driver IC would certainly be easier, and there's a good chance it would work better. But it might cost a bit more, and hey, sometimes it's cool to design circuits at the transistor level :)

Also, just curious, what circuit sim program are you using?

KLH

Thu Feb 05 2009, 12:19PM

Registered Member #1819 Joined: Thu Nov 20 2008, 04:05PM
Location:
Posts: 137

Both of us used LTSpice Switcher CAD III. This can be downloaded for free from the Linear Technology website.

GeordieBoy

Thu Feb 05 2009, 03:07PM

Registered Member #1232 Joined: Wed Jan 16 2008, 10:53PM
Location: Doon tha Toon!
Posts: 881

I would bypass the gate-resistor with a diode during the device's turn-off phase. In a boost converter you want the MOSFET to turn off as quickly as humanly possible, but to turn on more slowly to avoid damage to the diode and lots of horrible ringing.

You might also want to include a current sensing resistor in series with the MOSFET's source lead so your controller can monitor the peak current. This would allow your controller to see if the inductor current becomes continuous and starts to rise up to a dangerous level.

-Richie,

KLH

Thu Feb 05 2009, 05:25PM

Registered Member #1819 Joined: Thu Nov 20 2008, 04:05PM
Location:
Posts: 137

Since you're using a controller, you can optimize the timing (without much difficulty) of the MOSFET so that the peak current is limited to a safe level and so that it also never stays at zero for a significant amount of time (preferrably no time at all). Using this method requires a current-sensing resistor (as said above). This will optimize efficiency and charge the capacitor in the least amount of time.

j.azz

Fri Feb 06 2009, 01:52PM

Registered Member #888 Joined: Tue Jul 10 2007, 06:52PM
Location: Hannover, Germany
Posts: 40

Well the parts in the schematic aren't actually the ones I'll use.
The mosfet will be a IRFP450, capable of 14A cont. and 0.41ohm on-resistance.

An inverter on the input is a good idea. Wonder why I this didn't came to my mind...

As for the inductor, the dc series resistance of the one I have is almost negligible (my dmm can't measure -> below 0.1ohm).

What kind of shunt do you recommend for current sensing?
For current sensing, shall I measure the peak current before and hardcode it, or shall I look for 'dI/dt near zero'?

thanks for the answers,
j.azz

EDIT:

Oh, I guess I slightly misunderstood how I should use the current sensing.
You mean I sense the current to shut off the mosfet when the maximum acceptable current is reached, right?

1 2 next

Moderator(s): Chris Russell, Noelle, Alex, Tesladownunder, Dave Marshall, Dave Billington, Bjørn, Steve Conner, Wolfram, Kizmo, Mads Barnkob