Paralleling Power Supply Units Made Easy ???
I am considering paralleling a couple of 20a variable adjustable psu's for 40a charging. According to what I have read on the world wide inter-web, it can be done quite easily, provided the following ...
Forum Discussion
So much theory so little experience....
In the real world, the units have to be connected to an IDENTICAL LOAD, one at a time and the voltage set on each unit as close as humanly possible one matches the other.
I don't give a damn how close published specifications are for ripple, the ripple has to be snubbed as best as possible. CONNECT CAPACITORS one at a time testing circuit stability from Shottky inlet to negative. If you see a DIFFERENCE in voltage on your meter STOP. The PSU has reacted to elimination of ripple. The effect of ripple on the psu voltage sense circuit can be ALL OUT OF PROPORTION when an isolation rectifier is used. I am going to start with a .1uf cap and work up to a 22uf cap on each psu. I'll post the results. This will give a person a chance to nullify inequalities in the feedback loops. I am going to control 2 units with a common pot. Means 1 amp diodes plus biasing resistor on the alfa unit. A lot of this stuff is a result of a lot of work using high power generator voltage regulator units for paralleling production.
In the real world, the units have to be connected to an IDENTICAL LOAD, one at a time and the voltage set on each unit as close as humanly possible one matches the other.
I don't give a damn how close published specifications are for ripple, the ripple has to be snubbed as best as possible. CONNECT CAPACITORS one at a time testing circuit stability from Shottky inlet to negative. If you see a DIFFERENCE in voltage on your meter STOP. The PSU has reacted to elimination of ripple. The effect of ripple on the psu voltage sense circuit can be ALL OUT OF PROPORTION when an isolation rectifier is used. I am going to start with a .1uf cap and work up to a 22uf cap on each psu. I'll post the results. This will give a person a chance to nullify inequalities in the feedback loops. I am going to control 2 units with a common pot. Means 1 amp diodes plus biasing resistor on the alfa unit. A lot of this stuff is a result of a lot of work using high power generator voltage regulator units for paralleling production.
