Capacitor or diode option in latching relay set-up?

I’m building a circuit that involves the use of latching relays. In this circuit, the relay will switch between power supply A or power supply B, to my load. I would like to ask for recommendations of whether to use capacitors or diodes in this case.

To illustrate my set-up, let's say that power supply A is currently supplying current to my load, but I want to trigger a switch to power supply B. There are two possible implementations that I have thought of. I can use either capacitors or diode to ensure that my load continue to function in the event of a switch triggered by the relay. In the former, during the switching process, power supply B turns on, and both power supply A and power supply B will briefly overlap (thus the need for Schottky diodes for each, otherwise problems cause by competing DC power supplies ensue) before power supply A is cut off by a relay, thus ensuring a smooth transition of power supply. The load remains unaffected. In the latter case, I envision a design where the (?)microsecond switch would be covered by a capacitor (or multiple capacitors) that ensures continual current delivery to my load, before power supply B takes over.

Given these two options, which should I use? I am thinking that in terms of power loss, diodes may cause more wastage than capacitors? Also, there might be abit of voltage drop if diode is used. Thus, I am inclined to use the capacitor approach. However, I am not very knowledgeable about this. I wanted to ask which would be the recommended option?

It does depend a lot on the current required for your load.

If the load takes only a few milliamps, then a few 100uF would be fine to hold-over with a relay switches.

If your load is 10 amps, then you'd need a huge capacitor.

Using diodes does give you a 'volt-drop' which is also more of a problem at high current; at 10 amps, a drop of, say, 0.3v will dissipate 3 Watts.

You could do both...
One change-over relay selects supply A or supply B to the load via schottky diodes
A second double-pole relay can be used to 'short out' the diodes.

So your change-over sequence would be: switch in diodes, switch over supplies, switch out diodes.

Yours,
TonyWilk

Thanks Tony. Indeed, the capacitor would need to be huge, hence I have some capacitors on hand to potentially join them up to increase capacitance where needed.

I foresee that the load will draw about 100mA to 300mA, hence might still be sizable in terms of power loss in the diodes.

In the 'both' case, how does the capacitor figure in the set up with the relay?

Can you offer a sketch ?

A latching relay gets one pulse, power and duration as noted on the data sheet.

From that point it will stay in that state until it receives a pulse to change state. [ NO POWER NEEDED ]

IF ( I am guessing) your question is if you lose power at the moment you want to change state of the relay. then your cap only needs to be as large as the power needed to charge that coil for the duration required.

IF (I am guessing) you are not using latching relays, but are holding a relay in a state and want to provide un-interrupted power for that device........

pyren:
In the 'both' case, how does the capacitor figure in the set up with the relay?

Depending on the circuitry of your 'load', the sudden step in voltage when switching in and out the diodes may cause bother... so you'd just need to smooth out the transition a bit.

Yours,
TonyWilk

dave-in-nj:
Can you offer a sketch ?

A latching relay gets one pulse, power and duration as noted on the data sheet.

From that point it will stay in that state until it receives a pulse to change state. [ NO POWER NEEDED ]

IF ( I am guessing) your question is if you lose power at the moment you want to change state of the relay. then your cap only needs to be as large as the power needed to charge that coil for the duration required.

IF (I am guessing) you are not using latching relays, but are holding a relay in a state and want to provide un-interrupted power for that device…

Thanks for your questions. I have attached some rudimenatry sketched diagrams of how the diode method would work.

The capacitor method is attached in the files here, in a step by step format.

pyren:
Thanks for your questions. I have attached some rudimenatry sketched diagrams...

Ooh... do not switch the GND lines, only switch the +ve of each supply
Diagram:
PSU_Relays.png

Yours,
TonyWilk

PSU_Relays.png

TonyWilk:
Ooh... do not switch the GND lines, only switch the +ve of each supply
Diagram:
PSU_Relays.png

Yours,
TonyWilk

I designed for the switching of the GND lines to avoid looping with multiple grounds. By using the relay to absoultely shut one off, I think that would prevent any issues or unwanted noises from occuring. Especially since my two DC power source dont share a common GND.

pyren:
I designed for the switching of the GND lines to avoid looping with multiple grounds.

If you have one common ground, then you 'avoid looping' because you don't have multiple grounds.

What are your two power supplies anyway ?

Yours,
TonyWilk

The primary DC source is a solar panel. The other is from the wall socket.

pyren:
The primary DC source is a solar panel. The other is from the wall socket.

Ok, I think you should have a common permanent ground connection

If, for example, you leave the solar completely unconnected - at some point you have to engage relays to connect it to your circuit. Without a common ground, that solar supply could have 'floated' away from circuit ground and would give rise to voltage 'spikes' when connected.

Yours,
TonyWilk

TonyWilk:
Ok, I think you should have a common permanent ground connection

If, for example, you leave the solar completely unconnected - at some point you have to engage relays to connect it to your circuit. Without a common ground, that solar supply could have 'floated' away from circuit ground and would give rise to voltage 'spikes' when connected.

Yours,
TonyWilk

Thanks for the advice about the need for a common permanent GND. So that would mean the GND of the my wall socket. I have a silly question, but I will ask just to be sure. Would connecting the GND of a solar panel system to the GND of the wall socket affect the power consumption readings of the grid? Since I employed the solar energy system, I wanted to save as much cost (grid energy bill) by only utilising the power from the wall socket if only needed. If the + of the solar panel were to go through the load and drain into the wall socket GND, would that cause any changes to the power meter readings of the grid, such that I may be somehow charged for energy perceived to be taken from the wall socket, when in fact it is the solar panel?

pyren:
Would connecting the GND of a solar panel system to the GND of the wall socket affect the power consumption readings of the grid?

No.

Current can only flow around a closed circuit.

Consider an LED torch, current flows around the circuit inside, through the LED. The body of an aluminium torch is usually connected to the -Ve of the battery inside, if you touched the torch to some GND connection do you think some current could somehow flow out of the torch ?

Yours,
TonyWilk

TonyWilk:
Ok, I think you should have a common permanent ground connection

If, for example, you leave the solar completely unconnected - at some point you have to engage relays to connect it to your circuit. Without a common ground, that solar supply could have ‘floated’ away from circuit ground and would give rise to voltage ‘spikes’ when connected.

Yours,
TonyWilk

I would now like to connect the GND terminal of my home power socket to my solar circuit (+battery), however, I remembered you mentioned about a possible voltage spike on connection due to a floating GND of the solar circuit. How can I start to connect them safely? I was thinking that I could put a diode with the forward direction of diode pointing into the GND of the home power supply. Would that be sufficient to prevent any problems on connection?

pyren:
I would now like to connect the GND terminal of my home power socket to my solar circuit (+battery), however, I remembered you mentioned about a possible voltage spike on connection due to a floating GND of the solar circuit. How can I start to connect them safely?

Just connecting Solar GND to the 'wall socket power supply' GND will be fine.

If you have a multimeter, you might measure some voltage between the GND's before you connected them (either on AC or DC range), but it shouldn't be much, nothing will spark or explode or anything :slight_smile:

Having a permanent common ground avoids this sort of thing:

For the sake of argument, imagine your Solar DC is 12V. That is 12V between Solar GND and Solar+

If Solar GND happens to have 'floated' up by, say, 3V with respect to your Power Supply GND, then if you measured:

from PowerSupply GND to Solar GND = 3V
from Solar GND to Solar +V = 12V
from PowerSupply GND to Solar + = 15V

so if your relays happed to connect + first, there would be a 15V spike on the power rail. Maybe not a lot, but worth avoiding altogether.

Yours,
TonyWilk