Dealing with V drop when DC supply is switched between various outputs

[davidd31415]

I am looking for ideas on how to reduce I^2R drop between a DC power supply and a load without shortening or using larger gauge wire. The supply can be configured for remote sensing.

The supply output is switched across four different terminals. Each terminal is connected to a relay. If the relay is open the termminal is connected to the negative supply output. If the relay is closed the terminal is connected to the positive supply output. Only one relay is closed at any given time.

I have been considering routing the supply sense lines through a separate pole on the relays. When no relays are closed the sense lines would be connected to the power supply outputs. A timer would cause a delay before switching the sense lines from the supply output to the terminals.

The timer would keep the sense lines from being switched before the supply itself. This may result in the supply voltage dipping until the fifth relay closed but I would prefer this over a spike.

To shorten or remove the dip I have considered using a PLC or microcontroller but I would prefer a less expensive solution.

Any thoughts on this ?

 

[itsmoked]

Yes. Investigate your supply's sense inputs. Most of them have inherent limits so that unconnected they just go to the supply's local terminals.

I wouldn't expect any large excursions from connection/disconnection of the sense.

You could also use a filter on the sense signal so it "engages" over a short period of time, say, a 1/4s as it seems you just want a correction to a steady DC as apposed to constant tracking of a varying load.

Keith Cress
Flamin Systems, Inc.-

http://www.flaminsystems.com

 

[davidd31415]

This supply turns off if the sense leads are unconnected (TCR 20S30).

David

 

[itsmoked]

hmmm.

Well what if you put a cap across the leads at the supply? One that will carry thu a relay change. Ultimately it will be back to whatever channel the system selects, no big excursions.

It's the turn OFF and ON that will cause the overshoots.

Keith Cress
Flamin Systems, Inc.-

http://www.flaminsystems.com

 

[VEBill]

Could you use diodes (maybe 8?) to straighten out the polarity of the voltages sample at the far end, and then feed that result back to your PS. Adjust slightly for the diode drops.

To prevent PS panic during switching, also provide local resistors (and series diodes for Vd consistency I think) to provide a much weaker current-limited local sample to ride over the gaps.

Maybe a cap for smoothing if it helps and doesn't hurt.

Apologies that the above is more conceptual than detailed.

 

[VEBill]

Here (attached DiodeSteering.pdf file, about 175kB) is the conceptual schematic.

PLEASE double-check because I've not had my morning coffee yet. It could be completely flawed for all I know.

Optional smoothing capacitor not shown. Resistors value depends on Zin of Vs inputs. That path needs to be weak, but not too weak. Maybe 1k?