Linear Current Booster as a possible project for someone (Ghurd)?

[independent]

I've been looking at this for some time but don't really have the necessary skills for circuit board production (or even for myself as yet). But I thought I'd put the idea "out there" for anyone to take up.

Specifically National Semiconductor make a series of PWM buck converter ICs dedicated to this type of thing. They come in various fixed voltages but also in an adjustable version (that is one I am specifically talking about). One I can remember is called a LM2696 but might not be the latest version. It can handle reasonable input voltages. From the reference circuits provided by the manufacturer efficiency is very good (with the right inductor above %80) and the whole unit uses a few components the most expensive (guessing) being a medium sized inductor. The IC supports up to 3A and has an automatic shutoff protection feature (from memory).

I see two main uses.
1- Small scale MPPT for solar. 0.5 or more out of a 5w solar panel for charging 2 AA batteries anyone? Better yet decent current on overcast days. Would be a good circuit used together with Ghurd's dump load controller.
2- High quality voltage regulation. I've seen a variant of this IC used in high quality power supplies for audio equipment. It's one of those things that look easy enough to make but maybe manufacturers aren't too keen on as the actual components are reasonably expensive? A lot of electronic equipment use regulated power supplies. This would be good enough for that sort of thing.

If it's a crap idea let me know :^D

[Opera House]

Built a few.  A power point for for solar is quite easy for a two input  controller like a TL494 and many others.  Even  a single input regulator like the MC3063 can get an extra inut by adding a transistor.   I just manually set the input voltage for maximum power out from the converter.  It doesn't change much and having a circuit that hunts for the power point doesn't buy you much.  Speed kills.  Almost all circuits show operation at the maximum speed with the smallest inductor.  This is for manufacturers that know what they are doing.  The neophyte should slow things down a bit.  Inductors are everywhere.  I even use power transformers and run at 100 Hz. Absolutely forget this will improve anything on an overcast day.

[ghurd]

For Wind, what I expect most people will look at this for-
I believe there is way too much involved for a normal person to accomplish it in any reasonable fashion.
The orientation of inductors, capacitors, and even the way the traces lay on a PCB will all have an effect.

I know a guy who knows what he is doing, has access to all the latest sim stuff, newly educated with up to date info, etc.
And it took him a lot of tries to get it right.
I expect he has more money in prototypes than what the Classic will cost.

For Solar,
It can be done, like OH said, but not sure most people would get much advantage from it.
Small systems (like 5W and AAs) often lose more power in the circuit than they gain in output.
The time involved, and the cost of prototyping something larger could be a LOT more than most people would expect.
Can buy a somewhat dated solar MPPT for $200, or a new SunSaver MPPT for under $250.

G-

[Opera House]

 a possible project for someone......

Can't get around to my own projects.  What I want to try next is a string of those cheap #1 a watt solar panels to makie 150V into an old computer power supply and operating at the power point.  Surprised no one has done that yet.  And build a micro inverter from one of those four pin chips inside a 12V muffin fan.

[hydrosun]

Homepower magazine had a few articles by Chris Greacen in the early 90's on using those chips. I made a few to run 12 volt from a 24 volt battery. I even made a few for customers. They lasted a few years before being replaced with manufactured units with higher reliability. 
One of the articles used them for solar direct use.
Chris

[btrapr]

I am new to all this, but I was recently given 800watts- 50a worth of older panesl, SO I am trying to learn...

PLEASE let me know what you think about the following links I placed here...

I too am on a quest for a solar control system...

I have found the Ghurd LVD/Dump circuit interesting and a lot of good reviews from buyers... $12.95 plus shipping
http://www.ghurd.info/
*The only downside is that the P-fet is now obsolete- NOT sure what can replace the ZM33064?  ANY IDEAS?
reviews- check fieldline.com

OR these are on Fleebay
http://cgi.ebay.com/ws/eBayISAPI.dll?ViewItem&item=320559004290&ssPageName=STRK:MEWAX:IT
$10 shipped!
This company also has other similar
http://cgi.ebay.com/10-32V-35-60V-DC-Converter-Power-Voltage-Regulator-/320571235238?pt=LH_DefaultDomain_2

http://cgi.ebay.com/10-32V-35-60V-DC-Converter-Power-Voltage-Regulator-/320571235238?pt=LH_DefaultDomain_2

THis one is a LM2596

[Rover]

Ok #1 , I probably wouldn't have hijacked someone else's thread. You should have started a new one.
#2 ... Its not terribly difficult to find all the information about Ghurd's controller by doing a little search on here, in fact, you will learn all you need to know about it
#3.. rather than post links , describe the other units/stuff ... not many of us want to spend an hour following links figuring stuff out , etc.... you can always post a link after a description for more information.

However , Congrats on the windfall of panels, have you hooked a voltmeter up to them (one at a time) and seen how they are produing in Sun?

[Rover]

After posting that I realised, you still need to include more information.
1. Type of panels "800watts- 50a worth of older panesl" doesn't really mean a lot, manufacturer?, watts/ voltage per panel, type ? mixed bag? all the same? etc
2. Intended use? battery bank , dreams of grid tie? (<---really important)

[independent]

Just answering my own question.
There are some guys selling made up lm2596 with multi-turn pots, inductors and aluminium caps on ebay. Anyway, I've yet to fully work out how it works but it works! The unit will charge 3x AA from a nominal 12v solar panel. The unit will stop charging the batteries when full (you adjust the unit to stop at a predetermined voltage). Measured the unit will use 0.3A from the panel at 20v and charge 3x AA with 0.7A at 4.5v--success.
There are a couple of minor issues. I've got the manufacturers pdf to read. The problem I get is that the unit is constant current and constant voltage but ramps the amps down as the unit gets closer to the cuttoff voltage. The problem is the start voltage and end voltage are so close I can't seem to get more than 0.8A out of the unit even though it's rated for up to 3A. I had a 15w panel and it would only draw 0.3A. So, ideally the unit will work best for a 5w - 7.5w panel due to not using more than about 0.3A (the output of a typical 5 - 7.5w panel). There might be a resistor or something I can replace to make this more of a slam-bang voltage controller to get more of the potential amps out of the unit (there seems to be an artificial wall of 0.8A when the start (3.6v) and end (4.5v) voltages are within a volt or so regardless of the difference between that and the originating voltage (20v)). Will read the manufacturers spec sheet and see what I can find out.

PS I normally charge NiMH to 1.4v for longevity.

[Madscientist267]

This one works beautifully:

http://fieldlines.com/board/index.php/topic,144675.0.html

There have been tweaks and adjustments since the schematic on page 1, but you get the basic idea.

Steve

[independent]

I don't want to bring this thread to the top but I've been using these little converters lm2596 for awhile now and I thought I'd share some info and make a correction if anyone stumbles across this thread in the future. The LM2596 buck converters ramp the voltage down as they get closer to pre-set voltage. Kind of seems a waste when dealing with such little amounts of current. So, a "slam-bang" cuttoff would be better in some ways for small photovoltaics. LM2596 based converters are very efficient, 85% at a guess even using the more compliated constant current (as well as constant voltage) version. Great for charging batteries off batteries. As far as I know they are used in the best equipment and I can see why, very good buck regulators.
Now for the correction. I don't know how I was getting MPPT out of that other one, as I am fairly sure these need an addon opamp to provide for more volts out than provided at the source.

[Madscientist267]

After a while using mine to do some charging and an experiment with parallel rather than series (why I wanted a buck to start with), a buck (and by extension, MPPT) is of little point with less power unless your target voltages are WAY below the panel's output.

Even then, if the panel is small enough, the losses in the buck will negate the gains. I'm keeping mine around as a now optional addition because there are rare occasions when I need to charge a 6V lead acid from the panels. But for 12V, I've set it aside; I get a tad more juice by just running the panels in parallel.

If the runs are really long, and the wire small enough, it might be practical at 12V. But for me, using 100ft of #14, it's practically a wash, with winning tendencies in the direct connection department.

Shrug. It was a lot of money to throw at something that isn't in the loop anymore, but I have a suspicion it will mysteriously reappear at a later date when I upgrade the panels. When the current starts piling up in the lead-in, the buck will reveal it's glitter.

I'm thinking some additional mods to mine and 48V nominal panels will make for a sweet combo. 

Steve