some old Poly solar panels to mess with options for setup

[phester]

I have acquired 5 panels that are from the 1980's they have a voc of around 11V and Isc of 1.65A. I want to keep a battery topped up and I am wondering what is the best approach.

5 in series results in a good voltage but borderline I/P voltage 50V for some charge controllers. Only going to use cheap so called mppt (most likely) pwm from china.

2 x 2 in parallel results in a more usable V and I  but then I would need a separate solution to use the other panel. Maybe a buck converter feeding a cheap solar pwm ( to cut off charge).

Or could I use a boost reg and feed it all in parallel so 10V approx at 8A approx and convert it to (95% efficiency) say 26V and feed a solar charge controller for a 24V battery. Something like the mpt7210

I am only playing around as I am in Ireland we dont have great solar but after the hurricane I would love a way to keep some lights going off grid. I know I could buy new mono panels that would be more suited.

[Seekonk]

Panels in series. If you want to go cheap and actually learn some electronics, there are non potted 72V converters on ebay for $5 shipped.  These produce almost exactly 12V out so not suitable as is.  Cutting the output sensor line and adding a couple diodes in series can raise that to 13.8V float charge or switch to 14.2V for absorb.  In an emergency just leave it on bulk charge till it is over.  Make it MPPC to keep it at power point.  A TL431 or FET can be used in a pull down mode. That output can be used to fake the converter into thinking the 12V is too high reducing the load on the panels.

[phester]

5 x 11V (VOC) is higher that what I can find on ebay. They may go to 12v on a cold day

most of the board devices are based on the cn3722 which has a 28v max ip.
https://www.ebay.ie/itm/CN3722-MPPT-Solar-Controller-2A-For-1s-2s-3s-4S-cel-LiFePO4-Lead-acid-l-Battery/153175117618?hash=item23a9f2cf32:g:av8AAOSw4n1blsBd
this one looks nice as its usable with modification.  but if i use these then ill have at least 2 of these if not 3. I wonder if the o/p of one will cause an early shutdown of the other charge board.

[Seekonk]

That is not a true MPPT. Search 72V 180W 12V for a rectangular aluminum box. If your idea of DIY is to just click and buy, ignore my comments.

[fourtytwo42]

That is not a true MPPT.

Chinese speak with forked tongue, use moniker MPPT just like they use CE!!

[Seekonk]

Even TI is splitting hairs with their BQ24650. They call it MPPT but it is actually MPPC. The panel power point is fixed. Later they mention that thermal tracking can be added.  Personally I like MPPC as it provides a lot of flexibility for making the entire system more efficient. I'm growing to think MPPT sucks. I did an impulse buy of a makeskyblue everyone talks about. At low power levels of about 100W it can't track at all. It will drag a 60V array down to 16V in some conditions. Disconnect the panel for 5 seconds and the power increases 40% and operates where expected. I first thought it was once in a blue moon. After a week I had to add an automatic circuit to disconnect the panel for 5 seconds when panel voltage drops below 40V for 2 minutes. Every algorithm has its sweet spot. Else you are tracking all the time wasting power.  I've done fixed loads on PV panels and seen power vary 50% in a fraction of a second with wispy clouds MPPC is quite happy with that, tracking gets all confused with MPPT.

[phester]

Ya im not bothered by MMPT as MMPC is close enough for the price difference.

I have decided on versions of the Cn3722 for lead acid (car) or alarm Agm batteries as they are easiest to locate. (being thrifty)


As I have 5 panels  with 20 cells per panel with VOC at 11.6 in current sun conditions that should give a whopping 10V under normal conditions or 0.5V per cell. Isc is 1.65 at present and I think during the summer it was 2A.

So using the CN3722 I could sail close to the wind and try 3 panels to get 30V @2A max (will probably not live too long as its over the chip max I/P)) or else take 2 x 2 20V at 4A and tie both chargers to the same battery. It is 2 mode charger but it will suffice. They seem to suffer a 4ma leak but I might set the voltage higher and use a blocking diode to charge the battery.

With the stand alone panel I suppose its either going to be a spare or I can buy something cheap from ali 20W 12V to put in series to get my third string.

My only concern is the efficiency of these chargers is well tested online but only around the lithium voltages for 1s or 2s. The limited testing of the 5A version at 13V o/p  showed poor efficiency.

The BQ24650 seem to have an issue something along the lines of : going to sleep for too long and not waking up again to charge or was it 30 min to get to a certain voltage and if it fails it aborts the charge and sleeps.

[fourtytwo42]

Even TI is splitting hairs with their BQ24650.

Absolutely nasty marketing speak!! Constant voltage control does not track in any meaning of the word I ever heard but then in solar as in so many other industries most people neither know nor care unfortunately. I agree MPPT can be implemented poorly and not perform very well, I had both software and hardware problems with my early attempts but now it can follow the light down to about 2W (0.2% full power) and that limitation is due to the under-voltage lockout that ensures the mosfets are driven properly. Yes the dithering wastes some power but in my case is minimized to the necessary. Dithering around the ideal operating point is also caused by 100hZ ripple for those using grid tie inverters and I once went so far as to try and calculate the economics of ripple reduction capacitor costs vs the lost panel efficiency from the ripple.

If I was not able to design and build my own stuff then I would definitely avoid Ebay or anywhere else hawking cheap products as they do not have a good reputation and can be a safety risk, one recommendation I have heard is to de-rate there current by at least half!

[phester]

As long as the chip survives: fets, resistors and diodes can be up-speced and replaced afterwards. I dont have board layout capabilities anymore. Vera board is not great for noisy buck conversions.

Heat dissipation would be a concern if I start using an inverter on a sunny day from the battery and have these boards trying their best to dump 4A into the battery.  The heatsink version is better but I would need a temp activated fan too just to be safe.   

[Seekonk]

People will just take all the effort in the world to do something wrong.

[phester]

im sorry but buying a real mppt for 80W panels just seems a waste. If I am doing something wrong then it would be more helpful to point out what you think that is so I can consider it.