Hi mate,
I chose the 100uS as it was an easy figure and roughly mid way between the shown limits.
Would it be better to change the on time to a smaller one? Also any suggestions on the output cap as however much I read the sheet I can't find any info on choosing it, ripple is not mentioned at all!
Any help is appreciated, thanks Al
Edit. I can't get the Ton above 15 using the formula given without going above the suggested 50 to 500uH values, worked out as follows:- L (uH) = Vi divided by I(pk) x Ton which in my case is 12 / .360 = 33.333r and saying Ton is 15 gives a final value of 500.
Considering they say choose inductance between 50 and 500 uH and Ton between 25 to 150uS, how can this formula ever work out in their expected range? Unless I am making some weird mistake?
Hello again,
The simple answer is that whomever wrote up that data sheet made a mistake or else they meant to imply some sort of other conditions to those values. In any case, you should ignore that very part of the data sheet and proceed according to the equations given and whatever inductance you end up with you end up with, and that's what you use, whether it is 50uH or 5000uH.
The equations, unlike the rather restricted range they quoted, look correct so we'll use them. They should render us some good enough values to use in an actual circuit.
We'll start with a Ton of 100us and then work our way down by dividing that in half twice, giving us 50us and 25us, and then you can pick your favorite set of values
For an output of 28v we set the target output voltage to 29v to make up for some losses.
With Vout=29 and ILoad=60ma we get an iPeak of:
iPeak=2*ILoad*Vout/Vin
with the constraint that iPeak<450ma for some margin of error, and we get:
iPeak=2*ILoad*Vout/Vin=0.290 amps
(and that meets the iPeak constraint so we're ok so far)
and with a Ton=100us that means L comes out to:
L=Vin/iPeak*Ton=0.004138 Henries
and we'll increase that by 10 percent for some margin of error so we get:
L=0.004*1.10=0.0044 Henries.
We can also calculate Toff:
Toff=Ton*Vin/(Vout-Vin)=70.6us
and the duty cycle D:
D=Ton/(Ton+Toff)=0.5862
and just to check the output voltage:
VoutCheck=Vin*1/(1-D)=29.0 volts
and the average input current is:
IinAvg=ILoad/(1-D)=0.145 amps
With a max output ripple of:
Vripple=0.5 volts
we can approximate the value of the output capacitor:
C=(Ton*Vin/Vout)*(iPeak-ILoad)^2/(Vripple^2*iPeak)=30uf
To recap:
Ton=100us
L=4.4mH, and
C=30uf.
All of the above were based on an on time Ton=100us, but if we cut that in half then we can also
cut all of the component values in half too. This would lead to the following:
Ton=50us
L=2.2mH
C=15uf
Since now that is based on a Ton=50us, if we cut that also in half we can then halve all the other
values again. This leads to:
Ton=25us
L=1.1mH
C=7.5uf
So there you have it, three different sets of components to choose from.
In all of these the value of the current sense resistor should be 1 ohm, 0.5 watts.
Another thing to watch out for is the ESR of the inductor. You dont want too high of a series resistance or the output voltage will never be able to reach the required value.
If you are really seeing an output voltage that is just slightly less than the input voltage and all the components are chosen correctly, then there is a good chance that either the chip isnt enabled or the internal power transistor is blown open. This transistor could blow out if the wrong component values were used without that 1 ohm current sense resistor (a short instead of that resistor).