High side switch

[mmmalmberg]

That's definitely the idea I'm talking about. I'd add in a low voltage cutout for the cap too just to get it going quicker in the morning and since it's a simple circuit to duplicate on the same board. In your drawing I see the l.v.c. for the battery, but what does the zener do? Thanks Ron.

 

[ronv]

Actually that is a schottky diode. It just allows the load to draw from the capacitor until the capacitor voltage is about .6 volts below the battery voltage then the battery will begin to supply most of the current.
Have you actually tried charging the cap with your buck boost converter? The reason I ask is the cap will look almost like a short for a while till it charges up. We may also have to do something on the battery side. The sequencing could get tricky.

 

[4pyros]

For those who missed it earlier in the thread, here http://www.markmalmberg.com is a link to my website that shows the existing piece "Albireo" in a prior installation in Philly. Two Arduino-controlled sculptures communicate wirelessly by XBee, interacting in motions and sounds. I monitor the voltage state with the Arduino so I can use the motors more when there's more power available, and less when the voltage is low. In the middle of the day, there's lots of activity, the motors are much more involved and the battery gets charged with 1.8A and discharged with bursts of up to 5A or more. The cap will handle this part of the day really well and last just about forever doing it.

For the next installation in Albuquerque, there will be a third similar piece, and I'm rebuilding kind of everything to eliminate some weak links and to formalize the electronics much of which is currently wired point-to-point. I have some things also to get working that didn't get finished last time - encoder reading, some Hall sensors, amplification of the audio etc.

Thats to COOL!!

 

[mmmalmberg]

Thanks 4pyros!

Ron - One thing I was wondering is without the Schottky diode, won't the two sources just work in parallel as their voltages drop together? Haven't charged the caps yet, I was thinking about that but the good news is the regulator has a higher rating than what the solar panels can output so I'm thinking that should be ok.

 

[mmmalmberg]

Or is that diode there to keep the cap from discharging into the battery when it's at higher voltage? I guess I don't need a diode to protect the battery from charging the caps, that makes sense. Schottky vs a regular diode because of lower dropout?

 

[ronv]

Yes, the diode keeps the battery from discharging the cap when the caps voltage is higher. As the system comes down at night or the motors have run a lot in low light they will be in parallel untik the battery is cut off or some more light charges the cap, At the end the cap will take over again untill it is cut off. I think this is the best that can be done with the 2. I'll see if I can't simulate the whole thing today, but I did notice that there are very large currents if for some reason the battery would be switched on without the cap charged. This probably won't happen in real life but may happen in testing, so we need to stay in sync or fix it.

 

[mmmalmberg]

What you've got is closely related to the sketch I posted, as far as I can tell. A diode on the cap would keep the battery from discharging into the cap; I agree it's hard to imagine how it would happen in practical terms, and I'd prefer to not lose that voltage if I don't have to. If you haven't already, take a look at the sketch I posted and if you don't mind, give me some feedback on that; I think we're thinking the same thing...

I have a current-limiting resistor so that I can charge the battery on low current during the day, maybe 100 mA, while most of the current is diverted to the capacitor. Does that make sense? This setup altogether will let me use the battery very lightly rather than hammering on it all day long and I think will really help the battery life. I'll set the voltage cutout for the battery at 12.25V out and 12.5 in or something like that. Thanks again.

 

[ronv]

Don't know how I missed that post but I did. Yep I think we are going in the same direction. I think only one diode is needed since we always want to use the cap if we can. I'm pretty sure there will be a way to interlock the 2 circuits together somehow. The trick is to figure out all the possible combinations and check them.

 

[mmmalmberg]

I have a charge controller for the battery. I'm hoping all I need to do is current limit it so that a) the battery is charged gently and b) charging priority will go to the capacitor when the current output of the cells exceeds what the resistor allows through to the battery.

 

[ronv]

Let me think about that one. Not sure what the charge contloller will think about that. Do you have a part number for it? With some losses we could build a simple charger to limit both the current and final charge voltage to the battery, but if the controller will do it so much the better.

 

[mmmalmberg]

https://www.ecodirect.com/Morningstar-SunGuard-SG-4-12-Volt-4-5-Amp-p/morningstar-sunguard-sg-4.htm

If I limit the current going into it, I don't know what it can do other than charge at the limited current...

 

[ronv]

I think we can fool it. We can put a constant current source in front of it so we can get the most out of the panels when needed.

 

[mmmalmberg]

What I want is to charge the battery only at a low rate so that it will have a maximum lifespan, also so that the cap charges first, and so that no matter what happens with the battery, the piece can still run on the cap.

The other consideration is I don't have much time or resources left for additional circuitry, which is why I might use a simple resistor to limit current to the battery charge controller; I know it's about the least efficient way of doing things... If there's a better way that uses one or two components, that would be fine. I do have a simple current-limiting circuit that I'm using downstream in the power supply to limit current to the motors, could probably recycle that if needed, a couple of transistors and a couple of resistors.

 

[mmmalmberg]

I think we can fool it. We can put a constant current source in front of it so we can get the most out of the panels when needed.

What sort of constant current source are you thinking about?

 

[ronv]

Current source

I think something as simple as Fig. 4 would work. It's a common part and will need a very small heatsink. The resistor will need to be a 10 ohm 2 watt, but 5 watt may be easier to find. That way you will have 120 ma. at 23 volts or 14 volts.

https://www.electro-tech-online.com/custompdfs/2012/05/lm317-1.pdf

I think this is a good way to go. The motors should run off the cap almost all the time with only the battery as backup.

 

[ronv]

1st pass

I'm going to be tied up for the next couple of days but here is a first pass if you want to visit the local Radio shack.

Looks like if you can keep the big motor on for less than 45 seconds the battery will never be used as long as there is light.

 

[mmmalmberg]

That's great, and right in line with what I was thinking. Only thing is I can't quite make out the values on that sketch, maybe in the a.m with better glasses though So the regulator will be somewhat more efficient than just a resistor to limit current, correct? I.e. that's the reason behind it.

The big motor never runs for more than about 5 seconds at a time, often less, and with 20 seconds or more in between. And, like I said I watch the voltage with the Arduino so that I can dial back on the motors.

So that I'm clear, I'll be sending the output of the solar array to both the current-limited battery charge controller and also, in parallel, the buck-boost charger for the capacitor, yes? I'm going to go ahead and board this up...

 

[ronv]

Picture

The regulator is better because if you just use a resistor it has to be sized for 120 ma max at 23 volts or about 100 ohms. So at 15 volts you only get about 30 ma. This way you get 120 ma all the way down to about 15 volts before it falls off.

Yes, charge controller and buck, boost are separate. Input to both.

 

[mmmalmberg]

Fantastic. I REALLY appreciate your help and input. I'm working on the board layout right now Thanks Ron, most generous of you to share your time.

 

[mmmalmberg]

View attachment 64562 View attachment 64563