My Power Engineering/Electronics Project

Thread Starter

Hammad Imtiaz

Joined Oct 24, 2016
9
Hi, I would like some help with a project of mine. See, I want to design a priority based system for the charging of a battery through a UPS (Uninterrupted Power Supply). The system must give the first priority to Solar Panels, so if there is charge available from solar panels then the panels should charge the UPS/battery first. Second priority goes to the Main Grid, so if solar panels cannot supply power, then the Main Power grid should charge the load.

Here is a basic block diagram showing the idea.


I would also really appreciate any guidance on which sensors would be needed for this kind of undertaking.
Also how to deal with high voltages while interfacing the project with Arduino.
Please reply.
 

Kermit2

Joined Feb 5, 2010
4,162
First. You need to sense the solar voltage and whenever it goes below your threshold point you switch input sources to grid.
When solar monitor senses return of voltage above threshold input is switched back to solar panels.

For high voltage, the simplest way is to use a voltage divider and only work with a fraction of the full voltage.
A 900k resistor in series with a 100k resistor would make a zero to 100 volt signal into a zero to 10 volt signal.
 

ErnieM

Joined Apr 24, 2011
8,377
Do read the section on hysteresis. This is because once you release the load on the solar panel it's voltage WILL increase, which renewables it, pulls it down, disables it... It just chatters until the sun goes completely down.
 

Thread Starter

Hammad Imtiaz

Joined Oct 24, 2016
9
Do read the section on hysteresis. This is because once you release the load on the solar panel it's voltage WILL increase, which renewables it, pulls it down, disables it... It just chatters until the sun goes completely down.
So basically there should be a slight delay to account for the fluctuations in solar panel voltage that happen often, right?
 

ErnieM

Joined Apr 24, 2011
8,377
No. Not time, voltage. There should be a voltage limit A where the panel is switched off, and a limit B where it is switched on,
where B > A.

Just waiting some time will decrease the frequency of the chatter, but it still chatters.

These limits may best be found imperially (try it and see).
 

Thread Starter

Hammad Imtiaz

Joined Oct 24, 2016
9
No. Not time, voltage. There should be a voltage limit A where the panel is switched off, and a limit B where it is switched on,
where B > A.

Just waiting some time will decrease the frequency of the chatter, but it still chatters.

These limits may best be found imperially (try it and see).
Thanks, we'll make sure to test this out
 

John P

Joined Oct 14, 2008
2,026
That's "empirically".

But the empirical observations could go further than just determining when the solar-generated voltage is high enough to switch over to the panels. You could have a chart of measured voltages at particular current levels, and predict what the voltage is likely to fall to if the open-circuit measurement is a certain voltage and if a certain current is drawn.
 

Thread Starter

Hammad Imtiaz

Joined Oct 24, 2016
9
I have another question. We need to use a relay, preferably a Single-Pole-Single-Throw, which can handle a 12 V or 24 V DC input.
Does anyone know of a relay or relay IC of this type?
Also, can a relay of this type also give an output of 5V, or will i have to use a buck converter?
A small IC type relay would be very helpful.
 

Kermit2

Joined Feb 5, 2010
4,162
A relay is just a switch. You turn it "on" or "off" with a voltage applied to its coil.

If you want better reliability look at a solid state relay. It does not use magnetic attractions or physical metallic contacts.

If relays are all you can afford look at automotive supply sources for 12 volt relays with high current capacity.
 

Thread Starter

Hammad Imtiaz

Joined Oct 24, 2016
9
I have a question:
I have an automatic charging circuit for a 12V battery as shown in the attachments. The regulator lm317t is adjustable, and the transistors and zener diode help to prevent discharging and overcharging of the battery. My problem is that the regulator LM317t is too lossy. The sheer number of resistors and lossy elemts internally that it has would make my project very inefficient. Does anyone have any ideas as to what I could use to replace the regulator to prevent such wastage? Or any other way to reduce losses? And any other tips for this circuit?
Note: The battery goes where the voltmeter is.
upload_2017-1-7_14-19-2.png
 

Alec_t

Joined Sep 17, 2013
14,335
Whatever series analogue regulator you use it will be lossy. The regulator has to drop whatever the voltage difference is between the generator (solar panel) and the battery. Multiply that by the charging current and you have the power loss.
 

Thread Starter

Hammad Imtiaz

Joined Oct 24, 2016
9
Does anyone know a MOSFET driver IC model that can be used on the gate terminal of the 2N6660 MOSFET safely? The priority is compatibility. Also, is it possible to use a 555 timer IC to operate the driver IC, instead of using a microcontroller? If yes then how?
 
Top