Alternator regulator design

[tanglewood]

Adding to requirements....


- Remote voltage sense.


- Support for at least 12V and 24V operation, and 48V for extra credit.


- A PI control loop to throttle field current based on a temp sensor limit setting.


- A not-to-exceed field current limit for alternators that need to be de-rated for continuous operation.


- Enough flexibility in the charge profiles to accommodate both lead and LFP charging.


- A Force Float input would be really nice for LFP in addition to a full shutoff signal.


- Enough field current to drive two alternators. 10A would be the minimum, and 15A would be ideal.

[PaulCrawhorn]

Design one for LFP charging only. Then if successful, and you think enough demand, release a separate model for lead.

[goboatingnow]

Thanks for the suggestions

To avoid mission creep

12V operation only initially
Remote on off
Over Temp step back ( possibly bang bang or PID ) one external temp probe
Field current limits
Inbuilt shunt to accurately switch battery modes
Remote voltage sensing
Field current /output current limits to save belts
Low engine rpm protection
Protection from large step current demands to save engine pulsing
Alarm output
2x16 lcd

Both LA and Li profiles ( which are really just different set points ) can be accommodated.

Single alternator , use two if not you want two !

I ve just placed an order for a 2Hp mains motor to power the alternator test bed. !

I’m of two kinds as to whether to build it all in software or use an alternator regulator IC. I’ll do some investigations into both.

[RaymondR]

"I’m of two kinds as to whether to build it all in software or use an alternator regulator IC. I’ll do some investigations into both."

I'd go with the software option my experience being that code is easier to change than hardware. Microprocessors pretty well all have PWM built into them these days.

[Emmalina]

Don't forget the soft start . It really does make life on the belts better.

[tanglewood]

Originally Posted by goboatingnow

Thanks for the suggestions

To avoid mission creep

12V operation only initially
Remote on off
Over Temp step back ( possibly bang bang or PID ) one external temp probe
Field current limits
Inbuilt shunt to accurately switch battery modes
Remote voltage sensing
Field current /output current limits to save belts
Low engine rpm protection
Protection from large step current demands to save engine pulsing
Alarm output
2x16 lcd

Both LA and Li profiles ( which are really just different set points ) can be accommodated.

Single alternator , use two if not you want two !

I ve just placed an order for a 2Hp mains motor to power the alternator test bed. !

I’m of two kinds as to whether to build it all in software or use an alternator regulator IC. I’ll do some investigations into both.

I would be interested in one of these, but you lost me at the first feature..... oh well.

[Jammer]

Originally Posted by goboatingnow

Following on from my comments elsewhere , I was hoping to design an open source low cost external alternator regulator. [...]

For me the biggest cost is setting up a bench based alternator to test this. I’m looking at acquiring a 1.5 HP motor and fabricating a test rig. ( scrap car alternator )

Godspeed. Really. The yachting community needs this because the commercially available products are overpriced crap. You are correct, you can't test the system without a motor and test alternator rigged up.


For lead-acid chemistries (including AGM and gel) you will need battery temperature because the voltage limit changes -- a lot -- with temperature. You might need it for LiFePO4 in cold regions, because the maximum permissible charge current tapers off as the batteries cool down or overheat.



You can't really get it right without sensing charge current, especially if the load is large and variable. You have to be able to send C/5 or C/10 or whatever to the battery and know that it's getting there regardless of other loads.



You can't built a high performance regulator without sensing the temperature of the diode bridge and/or the alternator windings. You will have to fold back the output as temperature limits are reached.

[maine-cruiser]

I would love to see an input that turns off the field of the alternator. This could be connected to the BMS charge disable output. I have a Balmar regulator that I have interfaced with the Victron VE.Bus BMS. Since the BMS can only source 10mA I put a transistor circuit between it and a solid state relay that connects to the On/Off input of the Balmar. Would be nice not to have all that external stuff to enable/disable the alternator regulator.
If that's too much to ask then at least an input that enables/disables the field somehow. Then with some external glue it can work with a BMS.

[team karst]

In many installs, if you turn off the field, then the tach drive shuts off.
Is there a good reason to turn off ALL field drive?

[SOLAR SUPPORT]

The regulator inside the alternator can be used as it is specially designed for this alternator. I resorted to the old-school method to cut off the Lifepo4 charge to the desired charge level. After removing the alternator, with the help of an auto electrician, we cut one of the two brushes that transmit the energy of the alternator rotor, insulated it, soldered a pair of 2.5 mm2 heat-resistant cables with a length of 50 cm to the mutual cuts and then took them out of the alternator.

I fixed a 90 C° NC 10A thermostat close to the hottest pule side of the alternator. One of the wires goes in and out of this thermostat. A voltage sensitive NC relay with a capacity of 20A was installed between the wires, (VSR). This relay is connected to the alternator battery sense cable to detect battery voltage. The VSR was set to close the circuit in the range of 12.5-14.0 volts.

In this state, Lifepo4 charging was tried with alternator. Since the alternator was charging with its full capacity of 115A, it rose to 90 C° in 7-8 minutes and the thermostat activated and cut off the charge. Powerful fans that cool the alternator and engine compartment come on at the same time, and charging is interrupted until the alternator cools down. No problems so far.

When Lifepo4 batteries are close to being full, the system voltage rises from 13.5 volts to 13.8 - 13.9 volts and reaches 14.0 volts after about 10 minutes. When the Lifepo4 battery voltage is 14.1, VSR safely cuts off the alternator charge before the BMS in the battery breaks the circuit. Since the VSR system voltage is set to close the circuit when the system voltage is in the range of 12.5 - 14.0 volts, the VSR turns on and off continuously after the Lifepo4 batteries reach full while charging the alternator.

To solve this problem, a timer relay (TR) was inserted after the VSR. The VSR and TR start running when the LA crank battery switch is turned on. TR, which counts down for 5 minutes after the engine starts, does not start charging until the engine warms up. After this period, the alternator charge system voltage works smoothly together with the thermal control until the target cut-off voltage is reached. No device other than VSR, TR and thermostat is used in the system. Their total cost is around $100.

Alternator modification, relays, thermostat and their connection diagram are attached.

Cooling Fans video link:
https://youtube.com/shorts/2SW_gLZMJxE?feature=share

[skenn_ie]

It is highly likely that someone has already build an Arduino based one. Even it doesn't precisely meet your requirements, it will have done 80% of the work.

[skenn_ie]

Originally Posted by Rumrace

Yes i get that point. I have a power boat also but the alt has to run 3 hydraulic pumps a blower and 8 huge ignition coils. The two batteries don’t get much.
A diesel doesn’t waste juice on a spark

8 huge ignition coils ? If it is a coil-per-cylinder setup, they total will only consume the same power as an engine with a distributor. That is usually about 4 amps average per engine (12V electrics). Hydraulic pumps, presumable, aren't running much of the time ? Steering ans 2 trim-tabs ?

[goboatingnow]

Originally Posted by skenn_ie

It is highly likely that someone has already build an Arduino based one. Even it doesn't precisely meet your requirements, it will have done 80% of the work.

oh yes the designer of the Wakespeed originally developed it as a open source arduino project . it got rather over specified in my view

http://arduinoalternatorregulator.bl...5/welcome.html

but I still want to roll my own , using a commercially available alternator chip , the ST L9918 and just use a micro to do the interfacing and temp monitoring

[Colin Stone]

I've got 2 of those Heart Interface regulators. Wondered what I was going to do with them??

[Jon Hacking]

Originally Posted by goboatingnow

Thanks for the suggestions

To avoid mission creep

12V operation only initially
Remote on off
Over Temp step back ( possibly bang bang or PID ) one external temp probe
Field current limits
Inbuilt shunt to accurately switch battery modes
Remote voltage sensing
Field current /output current limits to save belts
Low engine rpm protection
Protection from large step current demands to save engine pulsing
Alarm output
2x16 lcd

Both LA and Li profiles ( which are really just different set points ) can be accommodated.

Single alternator, use two if not you want two !

I’m of two minds as to whether to build it all in software or use an alternator regulator IC. I’ll do some investigations into both.

Boating, I like the concept, & will provide what support I can, just as I supported Al in his Very Smart Regulator (Arduino) project. And I agree that you should focus on 12v only & a limited set of features for now. Get the low-hanging fruit first, or it'll never get off the ground.

But IMO as an EE & 27-year liveaboard cruiser, the *correct* way to charge a battery, lead/acid or LiFePO4, is to monitor the current into the battery. When the current into the battery required to maintain the Acceptance voltage drops to 1% of the Ah capacity of the battery, the battery is full & the regulator should drop the system into Float mode. This is what I call a "tail-current charge" & AFAICT is the ONLY way to accurately determine the SOC of the battery. Perhaps your system is already going to do this, since you're wanting to include a shunt, but it wasn't clear.

I wouldn't include a shunt, since most folks who care will already have a shunt for their Ah meter, & that can be shared. Including the shunt means that all of the alternator current has to pass through that shunt, so it should be external. Maybe provide one as an extra cost option, but I think most of your customers will (& certainly should) already have one. But the system needs the leads for this, & needs to be able to monitor that shunt. And ideally should be able to deal with different resistances of shunt, as well as different sizes (Ah) of battery.

Personally, as a geek, I would like the ability to adjust the setpoints. I use non-standard setpoints for my LiFePO4s, to avoid them getting too high. A tail-current charge at 13.8v will charge them fully (& almost as fast) without letting any slightly-out-of-balance cells get into their danger zones.

Lead/acid requires a temperature sensor on the battery, as some of the setpoints are temperature sensitive. I once destroyed a $3,000 bank of Sonnenschein Gels by charging them at only 13.8v, because they went into thermal runaway. Charging LiFePO4 batteries requires a temp sensor for the alternator, so it doesn't melt down. You can use a single sensor for both, I suppose, as long as you tell the system where it's connected. But I suspect it's simpler to just provide 2 sensor ports, one for the battery & one for the alternator.