PIC based DC/AC Mk-II (by Harry Lythall) suggestion

That is very interesting. There are undoubtedly solutions.
It is just to to a bit of thinking and designing.
Different use areas could require different solutions.

My PIC based inverter could perhaps be modified to shut down the oscillator under no-load conditions, so reducing the standby current to just a couple of mA.

I once designed a power socket that monitored the load with 5V AC (chopped DC) so that the 230V AC was inhibited if there was no load or a short-circuit. The idea was to create a child-safe socket. It was not that reliable in that the threshold between inquisitive fingers and a very light load was difficult to set. An SMPSU could not be detected.

But if the same technique was used with occasional "test pulses" at 230V AC then the standing current could be reduced from 100mA to perhaps 2mA. In effect, the processor could decide:
1 - Is there a load requirement?
2 - Is there enough charge in the battery to maintain it?

But this is all the subject for future thoughts and experiments.

Hope you had a nice Christmas
BR Harry

Admin wrote:Hi again,
That is a good suggestion, but there are are a few precautions to take.

I built a similar function into a solar regulator for charging batteries that powered a VHF repeater in Saudi. There was no external power in the dessert when they had the camel races, so we used two solar panels and 2 24v battery banks.

When the load draws current the battery voltage falls. At 11.8V it should cut off the discharge (change to the second battery) but when that happens the battery voltage comes back up to 12.5V or so. The solar panels could not keep the batteries charged since the repeater drew a lot of current during the actual races. My first circuit kept switching between batteries every hour, then more frequently as the batteries became discharged.

If using a single battery, then you need to disconnect the load and be sure that it does not come back on if the voltage rises. That will mean a reset button to start the inverter. Something like this will need a lot of thought.

If a battery is old and the internal resistance is a bit high, then a specific current will have a larger effect on the recovered no-load terminal voltage.

I suppose it could be possible to have a 13.7v switch-ON threshold if the battery is on a constant-voltage charger. I still have memories of relays clacking away as batteries died :-) For the camel-race container I eventually used a pair of jump-start cables from two land-rovers to get a 24V charge for the last couple of hours of the camel race.

I will give this some thought. Thank you for the suggestion. I will apply some of my experience to the problem and see what I can do. It will be a project for when I am next in Spain.

Best regards from Harry - SM0VPO

hi, 
i think shot down can be simply made by hanging in an infinite loop that does nothing. Reset can be  donne automatically  by simply switching off then back on the thing. So no extra switch is needed. 
Since the voltage drop is sensed under the load current It's shore that the battery is discharged and can't keep up. To avoid false shut off i would add in program something like : if the voltage is low(keeps going low) for more then 30 seconds then halt. 
This would be enough for the inverter i think. 
With the solar panels charging the battery is different thing.
Myself i had similar trouble with shut off and on part in 2016 with an arduino based MPPT (maximum power point tracking) solar charger i buil for home. (small 100w test unit) 
The MPPT algorithm works fine but the "what to do when no sun and battery discharged" problem appeared.I also used a 12v  dc power supply in the system so the consumers where switched to that during the lack of power from the battery(and sun) . The solar panels where only charging the battery in this state. But when to change back automatically the system to normal (sunny) operation? 
I started by adding a sub routine which evaluated the 'state of charge" of the battery. I did this both for charge and discharge. I quickly found out that doing that algorithm only based on voltage is not so reliable. I completely sulfated a new 45VA battery.
So i started using also power as parameter in the state of charge evaluation algorithm since i already had ampers  measured (actually the charging current) by the arduino due to the MPPT circuit.
This approach gave best results yet. I also did not have time to deal with it anymore.
Since then i use the 100w solar unit only to collect data with 20w 12v consumers connected to it. 
Untill now the solar potential seems to be just enough to motivate me build my 9kW unit on the roof in the far far future.... but i have my first 100w   so it's time to experiment the inverters a little.
Ps i am the only one to open his gate from 100m away around here 

Hi again,
That is a good suggestion, but there are are a few precautions to take.

I built a similar function into a solar regulator for charging batteries that powered a VHF repeater in Saudi. There was no external power in the dessert when they had the camel races, so we used two solar panels and 2 24v battery banks.

When the load draws current the battery voltage falls. At 11.8V it should cut off the discharge (change to the second battery) but when that happens the battery voltage comes back up to 12.5V or so. The solar panels could not keep the batteries charged since the repeater drew a lot of current during the actual races. My first circuit kept switching between batteries every hour, then more frequently as the batteries became discharged.

If using a single battery, then you need to disconnect the load and be sure that it does not come back on if the voltage rises. That will mean a reset button to start the inverter. Something like this will need a lot of thought.

If a battery is old and the internal resistance is a bit high, then a specific current will have a larger effect on the recovered no-load terminal voltage.

I suppose it could be possible to have a 13.7v switch-ON threshold if the battery is on a constant-voltage charger. I still have memories of relays clacking away as batteries died :-) For the camel-race container I eventually used a pair of jump-start cables from two land-rovers to get a 24V charge for the last couple of hours of the camel race.

I will give this some thought. Thank you for the suggestion. I will apply some of my experience to the problem and see what I can do. It will be a project for when I am next in Spain.

Best regards from Harry - SM0VPO

Hi all,
this looks like a good idea. It would prevent deep discharge of the car battery.

BR from Ivan

hi, 
i looked over PIC based DC/AC Mk-II (by Harry Lythall) and i would suggest to add an additional feature to the inverter. Since a few of those 8pin chips (12fxxx) have also ADC or at least a comparator, it would be useful to use it to monitor the 12 v supply. Only a voltage divider would be needed. When the battery is below 11.8 V for example, the inverter could shut down and save the (car) battery.
I think i can find code written in the past in hitec_c on an old computer which was doing a similar thing.That could be adapted.I used to make little modules for cars to automatically turn on/off headlights. The chip i used was pic12f675 i believe and it was monitoring the charging voltage on the battery. I used a mosfet to turn on lights above 13. something V and turn them off below that. Some wanted also a delay at turn off so they had light untill walking to the house.