🔴 DC Change Over Relay - Lets Make It FASTER ! - No.1329

[@zebo-the-fat]

Interesting!

[@IanScottJohnston]

You could play with a FET and low value series resistor across the relay coil to discharge it.

[@orion310591RS]

My logic tells me you are supposed to change relay, not the capacitor. It works now, but I think you will destroy relay sooner. Conclusion is there are pros and cons when changing capacitor. 33uF at 3:52 and 10uF at 4:49 5:09, difference is visible. 4:15 - This actually means, each time you turn on relay, its turned on 4 times, reducing life of relay effectively 4-5 times. Anyway nice short demonstration.

[@TheEmbeddedHobbyist]

Dump the relay and use active diodes to do the switching. If you need the relay isolation you could try latching relays as they are driven in to both states. But relays are always going to put contact bounce and switching delays in to the system. Also if the DC current gets higher then the welding of contacts due to arc's is going to be a problem. Relays, and switches are never very good at switching high DC currents as there is nothing to suppress the arc formed as the contacts break. Magnetic arc suppression helps a lot in relays and contactors.

[@sparkyprojects]

Ny solution would be to put a decent size capacitor across the output of the device so it acts like a short term battery back up while switching is happening, there's no switching time for the device you are supplying. ;) (might need a diode to prevent the capacitor backfeeding depending on circuit)

[@AnnaVannieuwenhuyse]

Reducing the amount of clamping the output diode is doing could help too. That back emf field is actually to some extent beneficial to your off time.

[@bblod4896]

Is there a snubber diode in parallel with the coil on the relay?

[@AnnaVannieuwenhuyse]

Yes, as part of the step down switching circuit.

[@bblod4896]

@AnnaVannieuwenhuyse 🤔. Well, I assume he wants a faster switch over because the load may reboot due to the momentary power drop. Someone else suggested a supercap across the power output to compensate for the possible power drop. That could work.

[@GeorgeGraves]

Why not use a SPDT relay? When the power is lost on one of the inputs, it will switch to the other one. What am I missing here?

[@TheDefpom]

@@GeorgeGraves that is exactly what this is

[@GeorgeGraves]

@@TheDefpom Hmmm. Then why does it need a power supply? Let the incoming voltage power the relay closed to one set of contacts. When the power goes away, the relay switches over to the other set.

[@TheDefpom]

@@GeorgeGraves because this particular unit can be used on a range of voltages, not limited to only the relay coil voltage itself, I think it was 11v to 30v or something.

[@GeorgeGraves]

@@TheDefpom Hmmm. That doesn't seem like a huge range. Not so much that you couldn't find a relay that would cover that range? Or limit the current? Dont' know.

[@entropyachieved750]

The contract bounce is interesting.

[@anullhandle]

Regular switches do it as well.

[@Conservator.]

@@anullhandle and keyboards too.

[@anullhandle]

@@Conservator. Idk if the computer or the microprocessor in the keyboard handles the key debouncing?? just a guess the keyboard?

[@anullhandle]

@@Conservator. Idk if the computer or the microprocessor in the keyboard handles the key debouncing?? just a guess the keyboard?

[@Conservator.]

@ since a keyboard connector (before USB) only had a few wires, I think we can conclude that the keyboard itself must have had a microprocessor to handle bouncing of the keys.

[@Conservator.]

You could probably mitigate the 30ms further by adding a very beefy super capacitor on the power output as a buffer. Oh, you mention that later on. 😁

[@Hasitier]

It’s always the capacitors 😀

[@poormanselectronicsbench2021]

I don't know if you have a pratical application for that board design, besides as an exercise in modding it, which is good info but I wouls suggest 2 different configurations if "change over time" is crucial. The first, which is practiced by major Telecom companies that need to power equipment ( and I know this because I spent my last 13 years in employment in one installing and maintaining equipment, power systems and batteries for one of them) is, to connect the equipment directly to the battery source, and have the Line / Mains powered DC source directly bridged to the batteries with no relay or switching connect device beteeen them. This will work if the powered equipment isn't voltage sensitive to the point that the difference between a battery float voltage and working voltage after the line / mains voltage cannot power the system anymore and work in a wid range of voltage power, such as, most 48V powered eauipment could work in a range between 44VDC and 60VDC, as the Absorbed Glass mat, or AGM technology batteries we used had a common float voltage of 54.48 volts given to them, not accounting for changes due to tempreature compensation. The direct bridged configuration has no change over time, or, relay that has to take a surge and invoke a switching time between sources, and is basically fool proof. If something was more sensitive to voltage change, designing a bridged system that used a MOSFET, possibly a P Channel depleton mode one in place of a relay, could most likely reduce the connection time to microseconds with no contact bounce.

[@SylwerDragon]

nice video. I would suggest to use only 100nF capacitor not electrolytic and use instead of original capacitor only diode ..that should help with extra voltage in opposite direction