High voltage input capable DC-DC converter for solar/PV

[zac]

I'm working on some monitoring (data logging) gear for solar power systems.  The typical array voltage is up to 600V (but more typically 300V) so am trying to find a low cost dc-dc converter or design one.  The measurement electronics and radio transceiver only use a fraction of a watt (typically less than 5V @ 100 ma) so am trying to figure out how to convert 100-600V (DC) into 5-12V. 

This is a commercial dc-dc converter than can do that, but was told it's too expensive:

https://mepospower.com/index.php/products/dc-dc-converters/item/267-100-1200vdc-ultra-wide-input-dc-dc-converter-for-pv

The previous solution used a separate small solar panel to run the electronics which worked pretty well until a bird pooped on it. 

[zac]

I found these linear regulators that can tolerate an input voltage of up to 700V (450V operating) and provide 10 ma of output current.  Since package thermal resistance is 79 degrees C per watt, I'm uncertain how the device can provide 10 ma of output current.  Assuming 450V input voltage and the 15V output version, that means 4.35 watts dissipated by the regulator at 10 ma output.   Based on the 85 C max operating temperature (though shutdown doesn't occur until 145 C) and 20 C ambient, allowing temperature rise is only 65 C which corresponds to 0.82 watts power dissipation.  That implies a maximum output current of only 1.9 ma continuously.  I've never worked with regulators with this high an input voltage, but can't think of why a high voltage linear regulator would work difference than a low voltage one.  Am I missing something? 

https://www.onsemi.com/PowerSolutions/parametrics/366/products



Posted on: July 25, 2019, 07:58:06 19:58 - Automerged
It occurred to me a high input voltage switching power supply wouldn't be much different than one designed for 230/240 VAC input (like utility line voltage in most of the world other than the US, Mexico, and Japan).  240VAC rectified is 340V which is pretty close to what I need.  And most small switching power supplies are designed to work with 100-240VAC input to cover the whole world.  (Japan has the oddball 100 VAC line voltage, except on US military bases where it is 120VAC.)  So, I wonder how low the input voltage can drop before the typical switching power supply loses regulation.  I can't think of why a typical SPS wouldn't accept DC input.  The only problem I can think of is the undervoltage protection in some designs.  I guess I'll try running some small switching from a DC power supply and see how they behave as the voltage drops before the mininum rated operating voltage. 

[PM3295]

It will be quite simple to run a small AC rated switch-mode from a similar rated RMS DC voltage, since the circuit essentially runs from a rectified DC bus. 

[zac]

Unfortunately, it would have to accept a much wider input voltage range (like 100-300VDC) that would be well under the 100VAC minimum input voltage most are rated for, even assuming a 10% undervoltage.  But, I just tested 5 small switching power adapters ranging from 3.5 watts to 40 watts.   All the units rate for less than 12 watts (and in the wall wart form factor) worked down to 50VDC input though current output was much less than rated.  But, none of the larger unit produced any output voltage even with 102VDC input (which was the maximum I could get out of 2 HP power supplies in series), presumably due to a undervoltage protection function.