Buck boost converter confusion / General Science and Electronics / Forums

Forums

4hv.org :: Forums :: General Science and Electronics

« Previous topic | Next topic »

Buck boost converter confusion

previous 1 2 3 4 next

Move Thread

LAN_403

DerAlbi

Thu Apr 04 2019, 07:04PM

Registered Member #2906 Joined: Sun Jun 06 2010, 02:20AM
Location: Dresden, Germany
Posts: 727

I would suggest to try LTSpice at this point. If you want CV/CC regulation this becomes very complicated very fast. If you want good regulation, i would suggest a linear output stage with a slower regulating switching pre-regulation. Maybe buy the Chinese module, see how it performs and what you wish to be better, then you have a better specification to verify in Spice.

Inducktion

Thu Apr 04 2019, 07:43PM

Registered Member #3637 Joined: Fri Jan 21 2011, 11:07PM
Location: Buffalo, NY
Posts: 1068

This design is actually inspired by a Chinese module that uses a knockoff LTC1775/1625; biggest difference in mine though is that I'm using dedicated gate drive, bigger inductor and high side current regulation vs low side.

The Chinese module works, but like I said I want to make this my own. I'll try LTspice and see how things go.

2Spoons

Thu Apr 04 2019, 09:36PM

Registered Member #2939 Joined: Fri Jun 25 2010, 04:25AM
Location:
Posts: 615

Your current limit needs to be analog feedback : ditch the comparator, it has way too much gain so is likely to make the loop unstable. You want to feed back the current signal in such a way that it takes over from the voltage feed back - sort of an analog version of an "OR"gate.
I've attached a conceptual schematic - its basically two precision rectifiers connected so that "feedback" will be the higher of V sense or I sense. You will, of course, have to figure out all the phase compensation yourself.

2Spoons

Thu Apr 04 2019, 09:40PM

Registered Member #2939 Joined: Fri Jun 25 2010, 04:25AM
Location:
Posts: 615

I would attach a conceptual schematic - but I can't work out how to do that now.

Inducktion

Thu Apr 04 2019, 11:34PM

Registered Member #3637 Joined: Fri Jan 21 2011, 11:07PM
Location: Buffalo, NY
Posts: 1068

Would using one of these

be out of the question?

Getting rid of the shunt to measure current would be a huge boon, since at the range of currents I'm asking for it'd dissipate too much power, or the op amp would have to have so much gain that I'd fear for stability/accuracy.

It only has 80 kHz bandwidth, so I assume that'd be too slow but still want to pose it as an idea. There are faster versions available as well.

Also, 2spoons, you can take screenshots by pressing the windows key + shift + S, and click and drag across what you want to snip, then go to snag.gy to paste it there and post a link to the image.

2Spoons

Fri Apr 05 2019, 12:50AM

Registered Member #2939 Joined: Fri Jun 25 2010, 04:25AM
Location:
Posts: 615

Well yes you could use one of those, or you could use a 10mOhm resistor and more gain on your sense amp.
The noise figure for that allegro device is a bit crap. Its equivalent to about 100mA of noise signal. Which is ok if you are looking at 10A signals, but not for a 1A signal.
You can do better with a resistor and an op-amp, and you don't need isolation.

Inducktion

Fri Apr 05 2019, 01:33AM

Registered Member #3637 Joined: Fri Jan 21 2011, 11:07PM
Location: Buffalo, NY
Posts: 1068

Oh, I didn't even read the noise figure. Jesus. That's actually pretty bad, and looking at other similar parts they've all got a pretty nasty noise floor of about 100 mA.

Okay.... so.... what about something like this?

I lashed together a very crude buck converter in falstad's simulator (I know it's not LTspice, and I know the limitations of the sim, i.e. doesn't include parasitics etc)

Edit: (realized I was controlling this in a very goofy manner)

The op amp series I'm considering has an incredibly low offset voltage, of only 25 uV.

I know this is spitballing at this point, but I want to know your opinions before I go and actually order PCB's and parts for this.

2Spoons

Fri Apr 05 2019, 03:47AM

Registered Member #2939 Joined: Fri Jun 25 2010, 04:25AM
Location:
Posts: 615

Pretty much typical for current sense in the ground line.

Inducktion

Fri Apr 05 2019, 06:13AM

Registered Member #3637 Joined: Fri Jan 21 2011, 11:07PM
Location: Buffalo, NY
Posts: 1068

My work thus far

LoFpmE

hen918

Fri Apr 05 2019, 11:40AM

Registered Member #11591 Joined: Wed Mar 20 2013, 08:20PM
Location: UK
Posts: 556

You've got some good design features in there, like nice big power planes, but I think the power components should be closer together, to minimise inductance and make mounting more convenient. they could all be on the same side then you only need one heatsink

previous 1 2 3 4 next

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