Hi folks,
I have an idea kicking about in my head and would appreciate some feedback with respect to a general sanity check and if it looks like it may be worth trying, how I might go about building/sourcing the transformer.

I was thinking about cell balancing, and charging in general, and wondered if handling each cell, or parallel bank, with an individual isolated DC-DC converter would be advantageous.

I am assuming, perhaps incorrectly, that using an isolation transformer at several kHz will keep it small and that keeping he input voltage high, relatively, will reduce losses, generally, and make the switching for the primary easier to build.

If you consider a module charging a LiPo or LiFePo4 at a little under 15A then each module would only be handling about 50W and because the output is isolated they could be stacked... When necessary

Multiple modules would offer redundancy and could be controlled by a microprocessor to either maximise cell charge or maintain the supply voltage, performing an MPPT function, all whilst keeping the cell voltages consistent.

SO... As I have never seen a battery manager that works this way am I missing something fundamental that makes the idea silly or is it just a case of it not being done that way, historically.

I cant help thinking that a single module could have several output and input taps to make it compatible with different cell and input voltages , it sounds good.

Am I just being silly or may this idea actually have legs...?
Thoughts with respect to design, if that is worth answering at all?

Cheers,
Al

 

I'm not sure what the aim of all this is ....???

I run my own power wall , charged from solar , over a thousand 18650's , and I have ZERO balancing and no regulator , the panels just feed directly into the bank !!

I have a voltmeter on each parallel pack ... so have 7 voltmeters reading perhaps 3.82 ....3.76 ... 3.81 ... 3.79 etc ... it's a misunderstanding that the voltages have to be "ballanced" ... you just have to make sure no pack goes lower than 3V or higher than 4V ( I keep my top voltage low for long life (li_ion))....

I have been running long enough now to know how the whole set up will perform and never do voltages stray where they shouldn't .. if cells are selected randomly when building the wall statistics determines the (7) blocks will have very close capacities and so will need very minor tinkering.

 

Good points...
I am still on a small-ish lead acid battery, goosed BTY, and about to add panels and fairly soon a different inverter.
I am completely off grid.

Given the relative low cost of panels I did wonder if MPPT was worth while at all.
My interim plan was to add some relays to switch off panels one by one, or bank by bank, and regulate charge current that way.

I suppose I am thinking that a large lifepo4 bank is a significant investment and I wouldn't want to cook it!!

Interesting what you say about paralelled cells, makes sense that they would tend to even themselves out, I like simple.

 

lead acid battery,
I am completely off grid.

Given the relative low cost of panels I did wonder if MPPT was worth while at all.

I have a very bitter experience with MPPT and regulators in general , I've bought many and the average life is about 2 years before some small creature gets inside and blows it , or the damp does the same ....

And you're right , the low cost of solar panels , and their virtual indestructibility makes it sensible to spend your money buying more of them ... even after 25years they still put out over 80% of original output ...

It also make sense to minimize battery capacity by storing cold .... run refrigerators and freezers only when the sun shines , put plastic bottles of water in the freezer which turns to ice in the day , and turns back to water at night keeping everything cold ...

But when it comes to batteries only lithium makes sense they last a very long time , I got all my cells from scrap laptop batteries , many people recycling them in this way , see youtube , or secondlifestorage forum.

 

Point taken RE capacity, and yes we do that, all be it by keeping he fridge/freezer full.
Bought the most efficent one I could get.

However I am looking at storing several days worth of power, rather than just daily cycles, and want to increase capacity over time, as and when I can afford it. Primerrally because of the cost of running the generator o charge the bank.

 

I think it makes sense to ask for comments on both subjects... lets call it passive management with monitoring and as a juxtaposition, my original question.

Thoughts?

 

However I am looking at storing several days worth of power, rather than just daily cycles, and want to increase capacity over time, as and when I can afford it. .

Again , things become much easier with excess panels ... Most people want to have electricity on the days it's quiet cloudy ... You can get 3 cloudy days in a row , when perhaps 30% of the power is generated compared to a sunny day ... this can either be overcome by having a larger battery , or by having more panels ... The cheapness and long life of panels favors the second option.... excess panels will mean you will have excess power which you cannot store , on sunny days ...You can use this for electrolysis , storing the Hydrogen for cooking.. moving over to electric vehicles... a large storage of ice for AC and hot water for domestic use ... It's about seeing where your power is used.

 

I am with you RE an 'oversized', conventionally but not actually, array and to be fair that is where I was heading but that will not stack up with the previous comment about just letting the panels slowly charge the cells.

I am not yet convinced about the 'zero balancing', although I am defiantly not dismissing the idea and will look into it further.
However with an array capable of meeting my needs on a short cloudy day, a long sunny day is, at the very least, going to need some active control, even if that is simply disconnecting part of it.

Using the power to heat or cool is all very well but should the loads fail or the thermal sink reach its capacity something has to step in and stop the cells from cooking and / or exploding.

As I said, I was already leaning towards a very simplistic control system, basically turning of panels, and if, as you say, I can manage balancing by design, monitoring and the occasional tweak to the pack configuration, essentially manually, then switching of panels to reduce the charge current seems like an even better solution.

I tried to build a 50A PWM controller a while back and just generated load of RFI and heat...
To be fair I could probably fix both issues, with time and a little research and thought but life tend to get in the way of days at the bench.

What are your thoughts on batteries?
I like the format of the 38130 cells, although I suspect the pouch blocks will be more cost effective and I don't want 'smart' 12V blocks.

I an essentially considering LiFePO4 and NiFe. Pretty sure I cant use Li-ion or Li-Po because of the fire risk and Pb is simply not robust enough in this application, given the persistent and regular undercharge during the winter.

My battery budget is circa 2K but that will be impacted somewhat by inverter choice.

 

Many people have the insane idea that liion or lipo are fire risks !! they are in billions of phones and laptops .....

There is no need to control rate of charge ... these cell can be charged in less than 1 hour ... in a solar powerwall this is NEVER going to happen in practice , in nearly all set ups the battery takes longer than a day to charge ...

All you need is a relay that is triggered when the battery reaches your chosen top voltage ... these are available on ebay , you set the two voltages levels yourself , I use 28 and 26 .... charging occurs until the battery reaches 28V then the relay opens and charging stops , when battery drops below 26V charging starts again.

 

"insane idea..."
I think that is a little strong.
Look I agree that an adequately protected/controlled cell will be safe and those millions of laptops and phones, they will be using protected cells or at least external circuits to afford protection.

If we are talking about hanging a bunch of cells from a solar panel and connecting a few loads, probably including DC ones that do not inherently have an inverter to shut them off then I dont think worrying about cell abuse leading to fire is insane, in fact I think not considering it is.

Look I get that a system dosn't have to be complex to work well if a cell chemistry is prone to exploding in circumstances that can easily be reached with normal use, let alone abuse, then you have to take notice at the very least.
Your relay solution may not be MPPT or even PWM but it is still a controller... And you are wise to have it.

I want to go with a chemistry that is less prone to exploding in the first place and then still try to treat the cells well.
I am convinced sufficiently with your balancing argument to think more simply, certainly until I have actually tried it but I am not going to get convinced that basic precautions are not worth taking.