18F4550 Solar battery charger / regulator & inverter system

[RYTECH]

I was wondering if anyone could comment on why I should or should not build this (see attached diagram)... it's a step up (or perhaps down) from how I've used my solar array and existing battery bank in the past, which was basically just feeding power directly from my solar array (currently about 33w) into my battery bank (was using 8xparallel 12v 2ah SLA) and connecting my inverters/regulators off that positive rail no current or voltage regulation. The solar array was connected previously via a over-voltage cutoff device which read the voltage level of the batteries. The relay which connects the solar cells to the battery bank is intended to perform a similar duty in this (currently theoretical) circuit.

I did some work last summer experimenting with the max4172 ic's which appear to offer reliable readings as to how much current is flowing between given points in a circuit.

Any suggestions on how to improve this circuit without adding a lot of complexity?
Any comments regarding the use of diode's in the circuit to direct traffic? The intent was to prevent power from coming into the batteries aside from what is flowing through the voltage / current regulated source while still providing any additional power coming in from the solar array directly to the regulators which provide power to the connected loads.

[mackrackit]

I do not have any experience with the max4172, but I do run a portion of my shop from wind and solar.

First I would say you have more solar capacity than you need at you current usage if the batteries are becoming fully charged and power needs to be diverted.

In a perfect set up, the system should be balanced so at the end of the day(literally) the batteries are just at the full mark. This is hard to do so you should shoot for "not quite full". A little over charge now and then will not hurt anything, it will save some one the maintainer's time of equalizing.



The intent was to prevent power from coming into the batteries aside from what is flowing through the voltage / current regulated source while still providing any additional power coming in from the solar array directly to the regulators which provide power to the connected loads.

You need to draw the power from the batteries. A 33watt panel will not run an inverter alone. The panel is for charging.

Sound like you have a simple circuit to disconnect the batteries if the do reach "full", so I would spend my time building a boosting circuit.

The idea is , when the panel is out putting less the the current voltage of the batteries, boost the voltage from the panel high enough to allow current to flow into the batteries. The circuit should have a diode to stop current from flowing in the wrong direction, say at night.

Sorry if all this is of no help, just sounds like you are concerned with the wrong end of things and do not want some one to waste there time like I did many years ago when I started playing with this stuff.

[RYTECH]

Thanks for your reply, all feedback and suggestions / criticism is appreciated and considered helpful...

[RYTECH]

I intend to use the relay to "cut in" and "cut out" the charging circuit depending on a) the state of the battery, and b) the state of the solar array. Let's say for instance it's early morning, and my panel which in full sun open circuit puts about 21v output, is putting out about 13.5. This would be indeed a case where a boost circuit on input charging power would be beneficial, which my "theoretical circuit" does not currently provide for. It does however try to use the solar array power on "load inputs" to reduce the battery output during low light conditions.

I guess one of the things I would like this circuit to do is use whatever renewable power (stationary bike output, solar output, wind output etc) is available to a) charge the battery (if necessary), and b) provide any leftover power directly to the load, taking some of the "pressure" off the battery array. c) report all the specifics on input and output power levels via a pc interface (usb, serial, i2c etc)

I've been working for about 2 years to whip together a relatively simple / low cost solution to manage using solar panels (or other input sources) in a renewable system. I would consider myself an "intermediate level beginner" in terms of my electronics knowledge, PICBasic programming knowledge, if that makes any sense.

What I'd really like to do is get a bunch of us talking about a reasonably scalable low cost circuit that accomplishes a) charging 12v Sealed lead acid batteries (like the ones in UPS units) b) over voltage / under voltage cutoff (maybe with boost, maxim?) c) USB/SERIAL voltage / current status monitoring (vb system tray interface?) provided by a 18F4550 or equivalent and a couple (very small) MAX4172 ic's.

I would be happy to scarf up and share some vb software to interface the whole thinger to a Winblows box (which would use the melabs usb dll, legalities?).

I'm not an electronic engineer, nor do I wish to be (at the moment). I think if there's other's working on similar tasks that we could all benefit from joining forces.

I realize my posted diagram is not a full blown schematic and lacks some passive components.

So far in the course of my electronics projects I've been able to a) setup a 18F4550 to do usb and use a/d inputs b) use max4172's to acquire current measurement. c) use the a/d inputs to aquire voltage via a voltage divider, d) write vb code that uses the melabs easyhid dll to communicate with the usb device.

Anyways let me know if anyone wishes to collaborate or has suggestions as to how to develop a reasonable circuit.

[RYTECH]

You need to draw the power from the batteries. A 33watt panel will not run an inverter alone. The panel is for charging.



I've used my IBM Thinkpad directly off the array in full sun using a (and this is going to seem backwards and inefficient) 12v to 120vac car adapter to my thinkpads power adapter 120v > 16vdc(?), and also in this configuration with just a 12v 7ah intermediate battery.

[DavidK]

Rytech
You might consider using the Allegro ACS712 in the battery leg of your system. It is a Hall Effect-Based Linear Current Sensor that is isolated from your current source. The fact that it is a bi-directional current sensor would allow you to monitor which direction the current is flowing; into the battery or out of the battery. On an A to D scale 2.5 VDC equals zero current and 2.5 to 5 VDC (well almost 5 VDC) is current in one direction and 0 to 2.5 VDC (almost) is current in the opposite direction. Even though I have not tried it, it is supposed to work in the AC leg of your project to monitor AC current flow. Even though it is a SOIC chip, it will handle 5 amps. There is another chip that handles over 20 amps.

[RYTECH]

Thanks for your suggestion, I'm always interested in other IC's / sensors that do current monitoring / reporting as my experience in current metering started and ended with using maxim 4172's...

As for their use in my posted "theoretical circuit" diagram, I've used diodes to "direct traffic" (possibly stupidly) at various points in the circuit... flow should at any any given point should be unidirectional. But perhaps there is a more reasonable "topology" that I should consider.

Nice information.. lets keep the ideas rolling folks hopefully by chatting we can come up with some kind of ideal way to layout this circuit.

(please note, just in case you were wondering, I do not intend to profit from or otherwise abuse any contribution made by fellow members towards this discussion, and if there is anything I can help others with (4172's, 18F4550 hardware etc) I will try to do so)

My vision here is to create a "open source" green power battery system that many people can cheaply use, enjoy, and almost more importantly understand.

[RYTECH]

Rytech
it is supposed to work in the AC leg of your project to monitor AC current flow.

David, I'm more interested in keeping track of the DC elements of the circuit > monitoring the power coming in from the solar array and the power out of the battery banks

[RYTECH]

I dug up my notes for some of the stuff I'd been researching last summer for charging batteries and came across notes for the texas instruments UC3906/2906, i think it would be better than the linear regulator/resistor combo for charging so I'm posting the data sheet in case anyone wants to look at it with me... actually the data sheet won't upload but I did post another document I found on the 2906/3906

The data sheet is pretty straight-forward... although I'm having some difficulty determining the resistor values necessary to indicate to the charger ic the bulk charge rate and for how many cells it's charging... maybe someone can make sense of it...

[Acetronics2]

May be you should have a look to AN U104 ... ( U for UNITRODE )

where it is clearly explained ... with the scheme ( ! )

Alain

[RYTECH]

Thanks I've posted that document in case others wish to review it.

[jcdillin]

Are you still working on this?


Right now i'm working on getting the parts together to start a small solar power system for my shed. At least give me lighting and some ventilation fans. I already have a few of the high power chips from allergo for reading amperage from the batteries and the panels. I am starting to work on the schematics and deciding which pic I want to go with. I would be interested in working more on this.

Jerry

[sayzer]

RYTECH,

Do you need the solar panels ONLY for charging?
I did not get it clear enough with the posts above.


---------------

[RYTECH]

Hi people... sorry for the delayed response, i've gotten somewhat sidetracked designing a perpetual motion machine...lol...

as for the power system.. I would like it to a) recharge the batteries and b) supply the load with any additional unused power from various input sources...

[ofuzzy1]

NEWARK and MOUSER sell the PB137 it is a 13.7V version of the 7815.

13.7V is the perfect voltage for Sealed Batteries, you NEVER EVER WANT TO OVER CHARGE THEM 14.1V is the absolute max.

GEL CELLS will never heal from overcharging, the bubbles form and push the mat away from the plate, the bubbles do not dissolve or move.

I like your idea, you could measure the amount of current flowing IN to the batteries if they are not receiving then you can by pass shunt off the excess.

On ebay many folks are sellling dc-dc converters, you can use these to stabilize the power from the panels