# Boost circuit using UC3843-IC

I would like others to explain how to make an buck or boost circuit using UC3843-IC.

I am constructing a solar power supply, that can be setup as an alternate power source so that it can be fed to main lines.

For an ordinary solar panel the generated voltage output ranges from 6 to 10V; but it's necessary to step-up this voltage to constant 12V in order to charge a 12V battery that we are using.

As it being our college project, we have to implement the boost DC-DC circuit with IC UC3843. Can any one explain me how to design a DC-DC boost converter circuit, that can able to step up 6-10V DC to 12V DC.

-
The datasheet has a schematic for flyback operation. With which section is your group having trouble? – tyblu Feb 21 '12 at 15:19
eevblog.com/2010/09/10/… – tyblu Feb 21 '12 at 15:21
You formulated your question really poorly; pay more attention if you want to receive attention. – clabacchio Feb 21 '12 at 15:22
MUST you use the UC3843? If so, why? What do they intend you to do about the fact that it will not start until Vin = 8.5V? – Russell McMahon Feb 21 '12 at 23:51
you are right. its 8.5V. can you explain me how can i construct a booster circuit using UC3843 for a constant output of 12V – Ashik Mathew Feb 22 '12 at 15:11

The MC3843 is not suitable for the task as you specify it.

You say Vin = 6 to 10V BUT the MC3843 operates from 7.9 to 8.5V start voltage. ie it will start when Vin is >= 8.5V and stop if Vin falls below 7.9V.

3845 = 7.9 - 8.5 (50% max duty cycle)
3844 = 10 - 16 (50% max duty cycle)
3843 = 7.9 - 8.5 (100% max dc)
3842 = 10 - 16 (100% max dc)

TI datasheet - contains above data.
On Semi datasheet
ST datasheet

Fairchild data sheet a bit sparse.

ie The IC will not start or run properly under some conditions of Vin. There are ways of "fooling it" but unless your course is an advanced one this is unlikely to be what the examiners had in mind.

If the person setting the paper specified the IC and the input voltage then they do not know what they are doing.

The MC384x family is useful if you need its capabilities BUT is messy and annoying to use. Because they are current mode controlled they have an inherent instability mode which MUST be compensated for where relevant. The versions which have 100% duty cycle capability are unconditionally unstable above 50% DC unless you use ramp injection into the FB cct. This is not hard but is annoying. The <=50% dc versions may not achieve what you want at low Vin.

If I was in your position I would try to use the MC34063 (sounds similar :-) ) which is nearly ideally suited to this task - so much so that I am probably going to use one for a real world application that does about what you are doing here - but I will be buck converting.

MC34063 !!!!!!!!!!!!!!!!

ST datasheet - very good - has sample circuits AND PCB layouts.

Fairchild datasheet very poor (again)

For best results in what you are doing you do not want to control the output - instead you want to control the input. There is a point at which your PV panel makes maximum power. Load it more heavily and it will drop in voltage and lose power. Load it too lightly and it will rise in voltage but you will take less current. SO - find the optimum voltage and us that to control your boost converter. Optimum PV panel voltage is about 0.8 x Voc in good sunlight - slightly less at lower sun levels and slightly more at high levels. Also varies slightly with PV cell efficiency. More soon ...

-
+1. OnSemi's datasheet: onsemi.com/pub_link/Collateral/MC34063A-D.PDF (Also very good, and judging by the MC... prefix, they're the ones who introduced the part first...) – zebonaut Feb 21 '12 at 18:21
that was really useful. but my instructor making us to do it with UC3843. as u said may be it will operate in certain conditions. will you help me to do that specified circuit to boost up input voltage to 12V dc – Ashik Mathew Feb 22 '12 at 15:05
@AshikMathew - You need to obtain Vin for IC supply from 12V battery so it is high enough to ensure operation. Look carefully at my comment re getting best resuslts by loading PV panl to a set voltage and NOT by monitoring utput voltage. Google for "MPPT". – Russell McMahon Feb 22 '12 at 21:01