Boost converter topology for battery charging / High Voltage / Forums

Forums

4hv.org :: Forums :: High Voltage

« Previous topic | Next topic »

Boost converter topology for battery charging

1 2 next

Move Thread

LAN_403

Storken

Fri May 03 2013, 09:04PM

Registered Member #1960 Joined: Thu Feb 05 2009, 11:57PM
Location:
Posts: 48

Hello 4hv,

I've built some buck converters and some flybacks - but never a boost converter. I want to make a constant current boost converter. The supply for the converter will be a 12V psu, the brains will be a arduino, connected to LabVIEW with USB. I could go the AVR route, but I needed an excuse for getting familiar with LIFA (LabVIEW Interface For Arduino).

Given it's a battery; voltage and current regulation doesn't need to be as tight as in other applications. I plan to sense those this way:
Current: Allegro ACS758 Hall IC current sensor (isolated signal output)
Voltage: Resistor dividers

With that said, here is what i hope to achieve:
- 48 to 80V output, adjustable
- Adjustable current, up to 5A.

I think I have everything except a useable diode - so maybe I'll just use a mosfet with a snappy body diode. Parts list will be updated.

As for boost theory; what happens when you have a low duty cycle and battery with high voltage on output? My understanding is that the voltage over the inductor will rise until its higher than the battery. The current has nowhere else to go. To rephrase my question; Will current flow into the battery at low duty cycles? (I'm guessing yes, in a sort of discontinous mode?)

Steve Conner

Sat May 04 2013, 10:38AM

Registered Member #30 Joined: Fri Feb 03 2006, 10:52AM
Location: Glasgow, Scotland
Posts: 6706

Yes, conservation of energy requires that any magnetic energy put into the inductor must come out again into the battery. Under some conditions, like the one you mentioned, the inductor current will become discontinuous. From the point of view of the control system, the converter is actually easier to control in discontinuous current mode.

Continuous current mode has a "right half-plane zero"- an increase in duty cycle actually causes the output current/voltage to decrease at first, before turning round and heading in the right direction. (Can you see why this happens?) An everyday example of a system with a RHP zero is steering a car in reverse.

So you tune your control loop in CCM, and it should handle the transition to DCM without any fuss.

Storken

Sat May 04 2013, 01:17PM

Registered Member #1960 Joined: Thu Feb 05 2009, 11:57PM
Location:
Posts: 48

Big thanks for that answer! Was looking at a more complicated half bridge design yesterday, as I couldn't find anyone else doing things as I plan to.

As for why output current/voltage decreases before it increases again; *Longish rewritten twice explanation ensues* (not sure if it is correct) The increased time in "energizing" mode means output falls lower than before. When it goes into "off" mode, the higher energy level in the inductor means voltage increases.

As for control loop, I have little experience doing this myself (SMPS IC has been much used) - so I hope the easy requirements of the charger gives me some hands on experience. As for regulators (PID etc), I hope a simple and slow PI regulator will do.

Schematic and parts list will come soon.
Parts I have:
- IRLB4030 100V 180A MOSFET (5V logic level)
- UCC27511 Low side gate driver 4A/8A (asymmetrical)
- 0805 Ceramic capacitors and resistors.
- Green blue toroid (Iron powdered) (DC to 1MHz) and AWG18 enamel wire.
- Arduino

Parts I need:
- Diode, schottky of some kind. Trying to get this sampled.
- 5V source.

Storken

Sun May 05 2013, 11:00PM

Registered Member #1960 Joined: Thu Feb 05 2009, 11:57PM
Location:
Posts: 48

Hopefully ST fullfills my request for some schottky diodes tomorrow. I finished a crude PCB layout, prototyping style.

Storken

Wed May 08 2013, 11:06PM

Registered Member #1960 Joined: Thu Feb 05 2009, 11:57PM
Location:
Posts: 48

This is turning into a build log. Hoped to show a functioning switcher today, but I knew that most likely wouldn't happen. It didn't either.

Did a lot of soldering today, got everything connected. Checked PWM generation with arduino - had to rewrite "Wiring" far down in arduino files to increase PWM freq from 480Hz to 32kHz. Checked the general setup in Power stage designer from TI, looks to be overbuilt in some areas.

With the selected gate resistors, turn on and off resistance is ~7ohms total. The gate is brougth to 5V, but I'm considering increasing it. Switching the gate takes 250ns. After hooking everything up (12V psu in and ~65V battery out) I tried running it at low duty cycles. The current sensor didn't report much current change - although it is very high range and I could have missed it.

Changed the scope probe to Drain-Source of mosfet. It didn't look good, had some spikes at turn off and some at turn on. As I was going to investigate furter, I stupidly managed to put scope ground at switch node(I wish I could say this was accidental, but I totally just spaced out for a second and forgot that scope ground = ground in circuit - doh...). Computer went black, scope went black. For a second thougth I'd made a very expensive mistake, but everything seems to be in order. Not the MOSFET tho, it's now shorted between all three pins.

Some attachments are added here, I'll update with better images of setup later.

Steve Conner

Thu May 09 2013, 10:10AM

Registered Member #30 Joined: Fri Feb 03 2006, 10:52AM
Location: Glasgow, Scotland
Posts: 6706

Welcome to the black and crispy world of power electronics!

Storken

Thu May 09 2013, 10:43AM

Registered Member #1960 Joined: Thu Feb 05 2009, 11:57PM
Location:
Posts: 48

Tell me about it xD

Replaced the MOSFET, turns out the ardunio and gate driver are both damaged, but partly working. The gate driver drives the mosfet between 1V and 4,2V, 5V input. The arduino receives and executes commands - but wont send anything back :P Going to replace both.

This post will be updated if I make more progress today.

Steve Conner

Thu May 09 2013, 10:50AM

Registered Member #30 Joined: Fri Feb 03 2006, 10:52AM
Location: Glasgow, Scotland
Posts: 6706

Sorry to hear it :( In future you might like to try an optoisolated gate driver like the TLP250 series, so a failure in the converter won't take the Arduino with it.

Another thing to ponder: You said that the control loop can be slow because you're charging a battery. This is true, but a good worst-case design should address the possibility of the battery accidentally being disconnected while charging. This implies a means of shutting the gate drive off quickly on overvoltage, and an output filter capacitor big enough to soak up all the energy in the inductor, with a voltage rise less than the difference between the overvoltage trip point and the MOSFET's voltage rating.

Sulaiman

Thu May 09 2013, 12:27PM

Registered Member #162 Joined: Mon Feb 13 2006, 10:25AM
Location: United Kingdom
Posts: 3140

If you're feeling lazy you could get one of these

Steve Conner

Thu May 09 2013, 12:33PM

Registered Member #30 Joined: Fri Feb 03 2006, 10:52AM
Location: Glasgow, Scotland
Posts: 6706

Looks a bit dinky for the claimed power rating. I guess it could manage 600W with 60V in and 61V out.

1 2 next

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