Alternator charge controller

[madlabs]

Hi All,

I've got my lawnmower/alternator battery charger up and running! For starters I used a simple power pot control, but the voltage climbs of course as the batteries charge. I was testing it on a small battery bank, so it took frequent monitoring. I'm picking up ~ 1100aH worth of golf cart batteries in the next week or so.

Anyhoo, I whipped up a simple PIC based controller for it. It can read the battery voltage and set the field voltage to the alternator by PWM'ing a PNP transistor. Works great, and the current ramps down as the batteries charge. The alternator can only put out ~ 50 amps, so I won't need to worry about exceeding C/20, let alone C/10. In fact I'm sure I will wind up scrounging a larger alternator. I also added a current sensor today. I'm going to add a temp. sensor too.

So, my questions are:

Do I need to ramp up the current initially, or is it ok to hit it with 50 amps right off the bat?

With smaller banks, I should set the controller to current limit to a max of C/10, and C/20 is nicer, right? I will use it to charge other batteries than my large bank I'm sure.

I know from our previous discussions that there is no reallly good way to use voltage to determine SOC, and that tracking aH in and out is probably the best way to go. Eventually I will make such a system. However, for now, I need to get a simple contoller going and move on to other pressing projects to get my move to off grid complete. I want this controller to figger out when I am ~ 80% charged and shut down the motor. With voltage and current readings availble, what is the best way to determine when to shut down the generator?

I have read that the best place to put a temp. sensor is on a negative battery terminal. I'm assuming it doesn't matter which battery in the bank, as long as all of the connections are of sufficient gauge. Sound right?

Jonathan

[ghurd]

I tend to think about things differently.

You have the max battery voltage set, and when it reaches that voltage, the charging current is decreased to maintain that voltage?

I think I would set it to shut down depending on the charging amps.  

Maybe 25A for an 1100AH bank? (I didn't put much thought into the number)

Maybe 10 minutes after it drops below 25A?

It would be nice if the initial amps could be ramped up over a few seconds (5?).  Mostly to let the motor get working comfortably.

G-

[madlabs]

Ghurd,

As the PIC maintains the preset voltage, the current ramps down by itself. My controller isn't doing anything other than controlling the output voltage by changing the field voltage.

I can have it ramp up the current no problem. I plan to have the controller wait a couple of minutes to allow the motor to warm up anyway, and ramping up the current is just a few lines of code, no problem.

Jonathan

[ghurd]

I mean use the charging current as an indication of SOC.

The lower the charging current needed to maintain the regulation voltage, the higher the SOC.

Like if 79% SOC takes XX.2A at the control voltage,

Then 81% will take XX.1A.

Just need to shut down when the amps fall to or below XX.1A.

Simply need to figure out the amps the alternator is supplying at 80% SOC.

Maybe not overly accurate, but easy.

I have a feeling it is a pretty good indication of SOC.

Might be able to use the same charging amps data as it relates to the field voltage?

Shut it down when the field voltage drops below X.X1V or something.

Maybe I am way off on that.  Makes sense in my head.

G-

[bob g]

i use a 2 minute delay and a 45 second ramp up, this allows everything to warm up a bit, stabilize rpm, and eases the strain on the drive system and controls smoke polution from the diesel engine. works for me.

bob g

[scottsAI]

Jonathan,

Since you have good control over things look at:

See for charging rates:

http://www.fieldlines.com/story/2009/4/15/2514/70412

Do NOT charge at high rates above 80% SOC.

To efficiently use the power from your ICE, you need to load engine 80% or better.

ICE gen efficiency drops as the load drops.

If using a vehicle alternator remember 60% efficient, few newer can hit 80-90%.

Since the SOC is unknown when starting, always monitor voltage.

We covered the other details with you in recent post: http://www.fieldlines.com/story/2009/4/12/15549/2455

If your doing a daily charge, stop at 80%. Weekly or so need to charge higher.

Once Voltage reached 80% SOC, keep it there and let current fall down.

Voltage for SOC is dependent on charging rate, as the current drops off, so should voltage. Keeping them in balance. Actual numbers will depend on your battery. Links above have links to articles.

If your uncomfortable charging above C/10, charging time is prolonged, cost to charge goes up dramatically using an ICE engine. Grid charging not same issues.

Have fun,

Scott.

[bob g]

on a decent set of flooded lead acid batteries that are around 50% SOC

i would have not issues with hitting them initially with 250amps for a 1000amp/hr set.

actually i would have no problem with a well controlled system going as much as

400amps initially, but that is a very large capability indeed.

voltage monitoring, temp sensors for both the alternator(s) and battery bank

would be of real importance.

the thing is, at 50amps charging, a 1000amp/hr battery bank is going to be an expensive to charge setup with a engine driven alternator.

most especially if you use the battery bank as you should for peak performance.

will you have other means of charging?

bob g

[madlabs]

Well, a 5.5 hp engine is about 4kW of energy. So 80% load would be about 3200 watts. At 60% efficeincy from the alternator, that would be 1920 watts. 50 amps at 14V is 600 watts. So, in theory, I should be able to run a 150 amp alternator or so. Does this math look right? In practice, I am hearing some loading on the motor at 50 amps, but pulley sizes and all that...

Alas and alack, I have no money for a larger alternator, but I'll keep the scrounging radar lit up looking for one. Of course, the scrounge radar is always lit.

The only other charge option I have is a 5kW, 9.9hp 120/240 VAC generator. It is rated for 4.5kW continous. I am going to get a charger for that too, but I don't have one yet. Ideally I'd like to find a 240VAC charger with an output

as high as I can get. Is it possible to use a transformer/rectifier to make a dirt simple charger? I have lots of fairly large xfrmrs laying around, or maybe I could scrounge up an old arc welder and remove some turns? The welder would have built in current limiting.

I can see I need more charging capability. I do have 2X 40 watts PV panels. My plan for those is to just hook 'em directly to the battery bank, they are never going to overcharge a 1100aH bank.

Once again and as always, so many thanks for all the help!!!

Jonathan

[scottsAI]

Madlabs,

Answer to your questions is don't know.

Seams the smaller engines inflate their output numbers, may not hit the 4kw mark.

My lawnmower claims 6.5Hp, yet using electric 1hp can do the job... YMMV.

Vehicle alternators are cheap, get another or two or three as needed use together.

Monitor the temp of the alternators, not designed for long term peak output. May find need to derate alternators to 60% to keep cool for long charging times.

Want to keep below 100C (max rated) Better below 80C. Internal Diodes get hot first, use external to help keep alternator cooler. May be difficult to remove. Use good heat sink on external diodes, cooler the better within reason.

If Engine stalls, overloaded it. Open throttle more or load less.

Transformer/rectifier is fine for bulk charging. Control is required after 81% SOC.

Transformers have their own efficiency issues (90-95%). Output voltage vs battery voltage, mismatch is efficiency loss.

Have fun,

Scott.

[americanreman]

My setup is a Honda 4.5 driving a ford 60 amp 1 to 1.

This is my 3rd one going back to 1981, the 2 previous  setups I controlled the field. On this design I control the throttle and just jump the field to the positive post of the alt.

Using a picaxe ADC input I monitor the bank voltage and raise or lower the throttle based on that voltage using a R/C model servo through the picaxe...very simple and works flawlessly, my power was out for 9 hours saturday after a storm...burned 1/2 gallon gas just floating my bank and load. Engine runs just above idle and puts out about 30 amps, can do more with additional throttle and load.

I pull the throttle gently to idle with a spring and increment the servo with the picaxe code, when the desired voltage is reached the servo goes back and forth gently to keep the voltage near steady.

Kept a full size side by side fridge/freezer running and 36 inch (tube) tv and sat receiver for the whole 9 hours and the engine never broke a sweat on a 1/2 gallon of gas. First time I used it in an emergency, built it years ago.

Startup load on the fridge is 1700, runs at 230...tv is 97 watts.

2000 watt Vector Inverter.

[madlabs]

Man I love this list!

I'm going to try using the arc welder. I have a giant 230V variac, IIRC it is rated for around 50 amps. I could use that to drive the welder to dial down the output to the needed voltage, as it is 30VAC I think. I have some giant diodes on heatsinks to rectify and plenty of big computer grade caps. That way I can leave the arc welder as a welder, although I could just tap and rewind, but that is work and I am short on time so for now I'll accept the greater losses.

I have been thinking of using a servo for throttle control. The Honda GVC160 I am using has some sort of govenor doohickey, I'll have to find out what that is about and disable it. I have plenty of servos laying around, and it should be easy to get that going.

Jonathan

[americanreman]

You should leave the governor in a working state for several safety reasons and to protect the engine.

My honda is for a pressure washer so it came with all the throttle linkage hard mounted on the side of the engine. The governor was set for about 1800 to 2200 (guessing) rpm for the water pump, just loosen the clevis bolt on the arm and adjust it for your desired high rpm.

I don't use stiff linkage between the carb butterfly and the servo so the governor can still do it's job if needed. I used that ball chain stuff, just like on car cruise controls so you can step on the gas to pass someone then let off and go back to your cruise set speed. Same principle.

Depending on how yours is setup, you may have to reverse think a little and pull the throttle back to idle with the servo rather than pulling it to give more gas so the governer can override the servo.

[madlabs]

Great ideas and advice, thanks! For humor I might add reading RPM to the controller. Got it hard wired up today, old school proto board style. I still have a few pins left over to read the RPM with. There is enough room left on the board for an optoisolator, so maybe that with an inductive pickup on the plug wire would do the job.

Jonathan

[americanreman]

I guess you can also read ac or dc from the alternator itself and determine rpm, never done it myself but there is info around on the net to get it done.

Like to see some of your schematics, I currently use up and down buttons to set the pic's desired voltage variable and hook a voltmeter to the battery to visually watch everything.

I have a 1 wire 4x20 lcd all setup, but it's for a ac battery charger project, maybe I'll look around for a cheap 1 or 2 line lcd and do amps, rpm.

[madlabs]

Vid card took a dump, so I haven't been able to reply...

I don't actually have a schematic for the controller yet, I just built it.

I am using an F88, and a 4*4 keypad, which has a F628 reading the keypad and outputting serial data. The LCD is a cheap parallel type. I am using the onbaord ADC to read voltages, through a voltage divider of course. I'm using an Allegro Hall effect sensor for current.

My batteries come today! Woohoo!