Hey there - somebody recently told me that it wasn't possible to have a very wide input range on a DC/DC converter. Is there any reason why this would be the case? Say I want an input range of 10-75VDC with an output of 5VDC. Couldn't I just find a FET that could handle the 75V, find a driver for the fet, and then find a dc/dc switcher without an integrated switch? Or is there some reason this won't work (like a minimum operating voltage on some FETs or something like that?)
Everything you say is possible and is sometimes done. But if you design for a smaller input range, the minimum efficiency can be made to be higher. A wide input range requires added complexity (cost) and larger compromises and losses, at least, at some operating conditions in that range.
Here is an example with a 4:1 input range (18 to 75 VDC):
10v-75v ->5v would be doable, but you pay for the wide range, basically you have higher peak currents becuase it has to be higher at 10v than 75v but at 75v you stil have the same current but at a low duty cycle.
this it would be less efficient in terms of size and power.
some controller chips have quite a wide range, although at 10-75 i dont think there are that many, i dont know one off hand.
However, A key component in the converter is the inductor. This is set by "delta Voltage" time and current. AS you state a >7:1 input voltage range, all above the desired out voltage range; the optimum inductor would have difeernet values. That being said, you could make a compromise. Also some chips use adaptive control loops,so that you may have adjustable "time" aka frequency and or Pulse width modulation, to absorb the variations in the input voltage,and then there is a whole realm of output loading and loading profiles that can be required! The devil is usually in the real -world details.
also as the VDD increases, 10v,to 20 , 40, 80 etc.. a different cmos process is needed to provide stand off voltage. as an 75 VDC source is an odd level?
Furthermore, it can be worse! There are time where the input voltage range, straddles the desired output range. for example say 5 vdc is output volts.
then you add an input voltage range of 2- 8 .. very different approach
It is possible, only the duty cycle for the power switch will be small (about 6.5%).
You could look at the LM5010 (for example). This is a solution that has the power switch on-board.
Some notes about this chip.
It has a boost capacitor charge pump to drive the internal N-channel MOSFET. When, after a transient, the FET stays off for a too long period, the capacitor runs out of charge and the switchter will not start again until Vout drops very low (to recharge the boost capacitor).
Other thing with such devices, the datasheet specifies a 30ns fast recovery diode. If you want to keep you chip alive, use a high voltage shottky rectifier instead.
Apart from what others have said, a knowledge of which topologies is best in a particular situation helps. A wide Vin/Vout range will make life rather difficult if you are using a voltage mode converter (which is particularly sensitive to Vin/Vout issues in the control loop) although a current mode converter will still have issues when trying to choose the inductor. (Voltage mode converters are sensitive to input voltage variations for a fixed output voltage and current mode converters are sensitive to output current variations as far as stabilising the loop is concerned).
Although it's do-able, most of the time I would do a double conversion if I had a known high voltage at the input.
There were lots of other good replys but one issue didn't come up:
Wide range converters have an extra measure of trouble with the control loop. The changing input voltage changes the loop gain unless you compensate for it somehow.
Abandoning the idea of holding to a constant frequency is usually needed to make a wide range one practical. If you have about the right amount of inductance for operation at the high end, it is way too much for the same frequency at the low end. A "constant ripple" design sort of does the change automatically for you.
ElectronDepot website is not affiliated with any of the manufacturers or service providers discussed here.
All logos and trade names are the property of their respective owners.