Home > Case Studies > Reverse Engineering of a Dual Redundant Switching Power Supply

Reverse Engineering of a Dual Redundant Switching Power Supply

Reverse engineering a solution to meet specific customer requirements

Client project requirements

The legacy power supply utilized on a customer’s programmable controller was obsolete and needed to be replaced. Due to the fact that the system had been classified as a single-point-vulnerability, the replacement power supply required dual redundancy capabilities. No commercial off-the-shelf unit could be sourced so a reverse engineered solution was required.


Scope of work

The project required Kinectrics to design and qualify a replacement power supply compatible with the existing boards in the programmable controller, meet the legacy functionality and performance, as well as incorporating the added dual redundancy capability and ElectroMagnetic Compatibility (EMC) immunity and susceptibility.

Kinectrics solution

Utilizing commercially available DC-DC converters, the replacement power supply was designed with two independent power supply blocks Or’ed together with diodes.
With each power supply block capable of supplying the full current rating of the legacy power supply, the new power concept had the capacity to double the output current of the legacy unit if the output load decreased. This presented an issue, as the current rating of the motherboard and associated connectors was only 120% of the full current rating.
The challenge in implementing over-current protection in a dual redundant power supply is in detecting which block should be shut down in the event the current increases above the over-current limit. Having each block supply 50% of the output current is the ideal arrangement, as it provides the least amount of stress on each block and the smallest output voltage dip in the event one block fails and the other assumes the full load.
Kinectrics’ solution to this problem was to set the current sharing between the two independent blocks with a 90-10 ratio. The output voltage of the lead block was set slightly above the nominal output voltage and the lag block was set slightly below. The difference in the output voltage of the two blocks was optimized to ensure that the lag block was able to assume the full current output if the lead block was disabled.
Independent over-current monitoring circuits were added to the output of each block, which when tripped would shut down the block. With the overcurrent for each block set at 100% of the full current output, in the event of a decreased load on the output, the lead block would be disabled when its output current reached the 100% level and the lag block would only be adding an additional 10%, effectively providing an overcurrent limit of 110%.

Client Benefits

The Kinectrics design provided the required dual redundant capability to address the single-point-vulnerability concern, in addition to meeting the reliability, performance and functionality requirements.