Birdbrained schemes needing a plan

[Guest]

Ok, I am in a corner, and I am tired of spending. 

I know this birdbrained scheme is NOT ideal but darn, I am not going to sell this controller easily, nor do I want to spend more money on more controllers and I want to get some ROI out of what I have bought myself into, years ago.

Have a Victron 75/15 MPPT Controller. At the time, with 2 x 200w panels, 24v system, it made perfect sense, as I am not getting more 200w panels.
Want to use a Victron 350VA 12v Inverter to power all the house lights at night.

Problem is that if the batts are connected for 12v, I lose on the panels production. Yes, the controller limits the watts, as Chris caught on the other thread.
Batts are not the issue, as they are 2nd hand.

Now, is this viable? 
Connect batteries 24v and then with pre-connected cables with Brad Harrison connectors, to not make that mistake again(!):
Week one run the inverter off the 1st 12v battery.
Week two run the inverter off the 2nd 12v battery.

If I see the savings I anticipate, then I will swap the controller, for a 75/15 controller is ideal for camping / 4x4 enthusiasts.

[Chris Hobson]

How about a 24/12 V DC  buck converter. Treat your batteries properly.

[___]

2 hours ago, Chris Hobson said:

How about a 24/12 V DC  buck converter. Treat your batteries properly.

Farmers had that problem too. 36V bank in the battery room, but the two-way radio is 12V. So some would just run a second set of cables and wire that across 6 cells, moving it up and down so every cell gets a chance. It's hacky as hell, but it worked. It is definitely not ideal, and you would frequently forget to move it and invariably some would work harder than others over time.

The buck converter sounds like the best idea... but now, you have a MPPT which is a buck converter, and that does a 95% efficiency if you're lucky. Then you have a second buck converter, and that does the same, maybe 90% or so for the cheapies. And then you have a 12V inverter, which at medium loads is maybe 94%. Let's just say 90% to the power of 3, or 72%. And that is before you do any battery charging. Also, buck converter will cost money, and since an MPPT is just a buck converter, what you really want to do is get another 75/15. Which costs money.

So perhaps go with the hare-brained scheme, just for goodness sake, remember that now your "negative" is floating in the middle when you're running from the "upper" battery. So easy to think of negative as ground and short sh*t out...

[Chris Hobson]

14 minutes ago, plonkster said:

So perhaps go with the hare-brained scheme, just for goodness sake, remember that now your "negative" is floating in the middle when you're running from the "upper" battery. So easy to think of negative as ground and short sh*t out...

..... and that is so easily done. We all have a spanner, screwdriver, multimeter,  SCC or inverter that bears testimony to our half-wit brains.

[Guest]

1 hour ago, plonkster said:

what you really want to do is get another 75/15.

How will a 2nd 75/15 sort the problem? Pray tell, for there are cheap ones 2nd hand.

1 hour ago, plonkster said:

MPPT which is a buck converter, and that does a 95% efficiency

Blue MPPT, FWIW, peaks at 98% efficiency, because they are so expensive. 

3 hours ago, Chris Hobson said:

24/12 V DC  buck converter

I have one, 24/12v 20amps converter, but do not like the thin wires and I thought the inverter would draw too much from it?

1 hour ago, plonkster said:

your "negative" is floating in the middle when you're running from the "upper" battery.

Please tell me more!?
idea is to have a cable with a BH connector on batt 1, pos and neg checked, and then another cable on batt 2, pos and neg check, that it is plug and play.

[___]

48 minutes ago, The Terrible Triplett said:

How will a 2nd 75/15 sort the problem? Pray tell, for there are cheap ones 2nd hand.

200W panel on the one... 200W panel on the other. 400W total, at 12V. That was the problem, right?

I don't think feeding the inverter from a buck converter is a good idea.

50 minutes ago, The Terrible Triplett said:

Please tell me more!?

Just saying, I think of the two  batteries as the negative of bat1 being at 0V, the postive-negative link between bat1 and bat2 is at +12V, and the positive terminal of bat2 is at +24V. So if you connect your inverter between +12V and +24V it will work, but then you must not do stupid things like grounding this or that without thinking very carefully. Probably not a problem they way you're going to do it, just the kind of thing you may want to consider if, say, you're using a 12V inverter in a 24V truck and, you know, you put it down on a metal part somewhere and forgot about earthing... it's just one of those things you HAVE to say when people go macguyver on you.

[Guest]

Thank you Plonkster.

Ok, I am going to try it ... with a multi meter glued to my hand to ensure 12v + en 12v - 

If I have to look at a controller, the Morningstar Tristar MPPT30 is eyeing me with interest, the bastard, tell him to look elsewhere!
Look here, then it is eyeing you too  : http://www.morningstarcorp.com/products/tristar-mppt/

Then the little 75/15 will become the portable campsite controller ... 

 

[Guest]

Okay, I chickened out. No not the system, it is in, house lights are now off-grid running off the little 350VA inverter. 

I am did not do the "slap riem" trick with 12v connection on 24v bank. 

Charging will be done by a 12v 500va UPS, connected to the big inverter to charge the batteries. Will switch it manually on till I get the guts to automate it.

@plonkster I may have a BMV600 for you to test. I can connect the RPi you sent me to the BMV, connected to the WWW, and upload the data to Victron if that will help you to test?

EDIT: Bottom line, not one sent spent!!! JIPPIEEEEE!!!

[Weasel]

Jis i don't know hey. I would definitely not go the alternating bank use method. i know what your stance is on batteries but now your'e just asking for it. before you do this let me confirm on what you are saying and play the devils advocate..

On 24V the 75/15 will do ~360W average and peak at 432W at 28.8V. with the 2 200W panels you are not likely going to see that 400W for any significant time and now the first point to make is that you are not doubling the power here by moving to 24V because you're using the same panels, if you added another 200 then yes you got me. But now on 12V the system would effectively be shaving off everything above ~200W right, but given a day how much is that actually going to be? is it not an acceptable loss? you will end up with something that runs 15A into 12V from 9 to 5 on the limiter, no problems, that's like 1.5kw total... if this covers your requirement then there is no problem.

The second point is by alternating the inverter between banks you are halving the storage, and now you want to over exercise one half of the bank and have it charge the following day in series to a full bank? so one half quickly goes up to 14.4 probably overcharges to 15 and your flat half only reaches 13.8 and then the mppt goes into absorb, now moering away that watts you are chasing in the first place. 

the third point is your adding manual changeover,  something will go south somewhere down the line. accidents happen.

so yes i wouldn't do it.i would run the 12V system with panels in parallel to get closer to the battery voltage, this will cut losses in conversion. 

you could use an orion DC to DC converter but they will set you back R750 for 12A, that limits you to 144W on the inverter. Or add  another 75/15 if you have to squeese that last bit out the panels. or you could just look out for a 24V 350Va. Ive actually seen some of the prices on them drop quite a lot because its an odd voltage for a small inverter. well.. for the most part. point is they are almost the same as a 75/15 so yea solves everything if you ask me. 

 

 

 

[___]

Good point on the unbalanced recharge!

Sent from my GT-I9195 using Tapatalk

[Guest]

@Weasel I got cold feet primarily because accidents do happen. Completely forgot about the unbalanced  recharge. Thank you for that heads up.

As it is now, UPS must recharge the batts until I see it working as I think it will, then I will get a Morningstar MPPT30 controller. Want to try one in any case, but not today, today I am gatvol of spending if there are parts lying around that can temporarily fill the gap.

SWAMBO is not affected. There is a plug point direct from the main inverter, separate circuit from the house, that is used by UPS inverter to charge. Just have to switch it manually on / off.

Have an idea to automate it but eish, not really wanting to tamper with my magic changeover box and add more relays.

At a point now where if it works, I leave it well alone. 

[Guest]

My other birdbrain scheme, 2 vastly difference controllers on same bank. Have it, use it I thought.

Look at the values today and with watts generated yesterday, how the MPPT and Morningstar PWM controllers do their thing on the same bank.

Like yesterday the MPPT took the lead early on and pulled a max of 419w out of the 2 x 200w panels, PWM stood back as it sensed all was fine. Took all day to charge batteries due to rain, never could go off grid yesterday. 

Today, once the batts where full, the PWM took the lead when the inverter switched on and now the MPPT is in absorption, PWM powering all including the UPS re-charging the batts from last nights lights. 

 

[Guest]

Klits, I forgot that the UPS takes 24 hours to re-charge.

[Guest]

And what do you guys think of this? ARDUINO MPPT SOLAR CHARGE CONTROLLER (Version-3.0)

[___]

Looked at it about a year ago. He's done a LOT since then! Awesome project with lots of info on how to do the math.

First problem... way too small, and not easy to just scale the design up. Second problem, at least back in the day, the Arduino PWM frequency is limited unless you mess with the low level registers. I eventually figured out you can make it run much faster, but the faster it runs the fewer discrete power steps you have. Here is my example of generating two opposing signals at 100khz:

https://github.com/izak/arduino-mppt/blob/master/mppt.ino

The idea I had was to build something similar, but with two buck converters which you drive 180 degrees out-of-phase. Gives you better ripple values, and you can use smaller components.

Still an awesome project, just a bit small for me :-)

[Guest]

Plonkster, in your opinion, take a Cheap Chinese MPPT. Arduino home made and a brand name, do you think MPPT's will differ in how efficient they are over a period of say a day, week, month?

[Guest]

I googled this this morning, wondering now for years why are brand name MPPT controllers are so expensive, compared to these: https://www.alibaba.com/trade/search?fsb=y&IndexArea=product_en&CatId=&SearchText=MPPT

If you take a brand name controller, and a cheap MPPT, on same array and same load over 1 hours, 30 days or 1 year, will there be a difference in generation I wonder.

[___]

Victron claims up to 10% more output compared to "slow trackers" (more about that below). So there will definitely be a difference in generation, but it's probably going to be single digits, and as usual one might want to weigh up the extra cost against just adding an extra PV module.

I think the biggest reason not to go cheap on an MPPT is because they don't last. One failure within the first 5 years and your entire ROI is shot, because however you look at it, two cheap ones is almost always more than the genuine article.

Re speed of tracking. Two algorithms: Perturb and Observe, or Incremental Conductance.

The simpler method used by the cheaper controllers is P-and-O, adjust the voltage up, if it yields a net increase in power, push it up some more, repeat until the power decreases. Then change direction: Push the voltage down and see if it yields an increase in net power, if it does, push it down again, otherwise change direction again. When it hits the MPP, it basically jumps back and forth between two or three closely-spaced voltage points right around where the MPP is.

The second method calculates the conductance of the panel (that's I/V) and the incremental conductance (how I changes with V, ΔI/ΔV). Some clever people figured out that uf I/V == ΔI/ΔV, then you are at the MPP, because manipulating the math yields the more familiar condition that ΔP/ΔV = 0 at the power point, which when you bring in limits means you want the point where dP/dV is zero. tl;dr: If you detect that you are far from the power point, you can perturb V by larger values and get to the MPP faster.

In cloudy weather, such a controller should adjust a lot quicker and find the new MPP faster, and hence the claims of up to 10% more.

[Guest]

Thanks Plonk.

Agree with cheap controllers that do not last long. Got a T-Shirt for that one too.

Heat also a problem. My first band name controller was a Phocos MPPT100/30 and also my first heat related failure, week after 2 year warranty expired.

If you run the MPPT's near max, they operate with the resultant increase in temp. The teeny weeny 75/15 one frequently on sunny cooler days exceed 400w from the 2 x 200w panels. It heats up obviously but nowhere near what one would expect.

The larger PWM Morningstar's heatsync temp also climbs, but the huge heatsync sorts that pronto.