# My DC-UPS - done right ? Advice needed

I'm working from a while ago on a DC-UPS concept, after I've put all the things together and I'm requesting from you an advice, can I start to build it or I have serious flaws in it ?

• Current comes from AC-DC transformer at 27.6V, goes into a distribution box like the one in this image
• Current goes to battery-distribution box which is wired on negative terminal with a resistor (limit charging current) and a diode in order to block current flowing OUT of the battery + blocking current flow other way around (thru load distribution box) - if grid power is available and the transformer works current on this state could only flow INTO the battery and not OUT of it. Due to Schottky diode there is a voltage drop of max 0.55v - so the battery won't charge higher than (27.6v - 0.55v) = 27.05V
• Current also goes to load-distribution box in order to feed the load with power

===

• From battery distribution box perspective, if AC-DC transformer stops delivering current (grid is down) it will feed the load distribution box making it a fully functional UPS with zero delay.
• Due to Schottky diode connected from battery distribution box to load distribution box there is a voltage drop of 1.1V - the battery won't reach a higher voltage than 27.05v so the load voltage (on battery power) will be somewhere between 25.95V (fully charged state) and 21.9V (near fully discharged)

===

• From LOAD distribution box perspective, there are two voltages present > the battery (25.95V fully charged - 21.9 near depletion) AND 27.6V (if there is grid power available) so the load will be feed from AC-DC transformer at 27.6V when everything works in order OR 25.95V from battery when the grid drops out.

===

I'm not 100% positive that I've put the diodes on the right places, electrons flow from battery negative > positive / from PSU should be the same. If the battery discharges when negative electrons are attracted by positive terminal then charging is made by feeding electrons to the negative end ?

I'm willing to create an efficient circuit in order to keep a strip of LED's running all the time 24/7 - total load of the circuit is somewhere around 3Kw (yes, I know that the AC-DC transformer is too small for the job but currently I'm thinking about it upto ~ 800W)

• Both of your diodes are backwards ... The 'arrow' formed by the triangle in the diode symbol indicates the direction of 'conventional current' flow, not electron flow. – brhans Mar 15 '17 at 15:01
• Thank you! Do you see any more flaws in my circuit ? – user3405598 Mar 15 '17 at 15:16

I am working on something similar. I have LEDs that run on 48VDC. I have a 48v supply and also supplied by solar panels with lead acid batteries.

I have not put it into operation yet but I have the diode OR worked out. I do not use Schottky Diodes but rater an Ideal Diode Controller from Linear Technologies. They replace the schottky diodes to reduce their voltage and minimize heat.

Charging batteries with a resistor as the current limiter is not good for the batteries. How bad it is depends on the batteries. I would NEVER do that with my Lead Acid batteries. That is just not the way to charge a battery. My batteries are being charged with a \$400 Max Power Point Tracking (MPPT) charger which is solar powered.

I bought this Linear Tech Ideal Diode Controller Eval Board

which uses this Linear Tech Controller:
LTC4355 High Voltage Ideal Diode-OR with Input Supply and Fuse Monitors

My Vin1 will be the lead acid batteries being charged by the solar panel. Fully charged they are about 56+ volts.

My Vin2 will be a 48V power supply tweaked down to around 44V, depending on the colors of the grow light LEDs.

When the batteries voltage drops below the power supply voltage, the controller board will switch over to the power supply. When the batteries charge back up to greater than the power supply voltage, the LTC4355 board will switch back to solar/batteries.