Hi All,

I need to isolate a relay board without using a second power supply, so I have decided to use a isolating DC-DC converter. I have found this at a local supplier - http://www.meanwell.com/search/spr01/spr01-spec.pdf, but I cant workout if it isolates AC.
In other words, if I use this with some opto-isolators, will my relays be "isolated"?

Thanks

 

I don't understand what you mean by "if it isolates AC". What AC? It's a DC-DC converter.

 

Sorry, should of been more clear on this.

The relay board switches 240VAC. I want to isolate the relay board from the logic board using opto-isolators but I don't want to use a second power supply form the relay board.

I guess what I am asking is "If I put 240AC on the output of this DC-DC converter, will the circuit on the input side be safe?". Don't care if the DC-DC releases smoke..
I know only some DC-DC converters isolate the outputs, just can't workout if this one does. I guess I COULD work it out, but I think asking here is safer

Thanks again.

 

That isn't all that much clearer. Are you switching 240VAC power by tying the power supply to the relay inputs? If you are using opto isolators, why would your DC supply have to share the same conductors as your AC power? I think a wiring diagram would help.

 

Sorry, let me try again.

I have got a relay - 240VAC on the output of the relay. Relay is connected to a opto isolators which in turn is connected to a PIC. Something like this (Note this is not my board, just a good example) - http://yourduino.com/p/OptoRelayChannelData-575.jpg

If I use the same power supply for both VCC and JD-VCC the opto isolator is doing nothing as I am sharing a power ground connections. (The above board is useless even if the user would power VCC and JD-VCC from different supplies, it looks like ground is still common).
Normally I would power the relay board form a different power supply therefor isolating the relay from the PIC.
In this case, however, I can't use a second power supply for the relay board so I need another way to split the power up. I know some DC-DC converters do just this, but some don't. This part mentions it has a 1K DC isolation, but makes no mention of AC isolation. I would think that if it can isolate DC it would also isolate AC but I just want to make sure.

The AC and DC would only come together during a fault condition and so I don't care about protecting the DC-DC convertor or the relay. I just want to protect the PIC board.

Thanks again.

 

Just an observation, but if you replace Q1 and the relay with a triac, then you can directly switch your triac (and therefore your AC power) from the optocoupler output. You would want a triac output optocoupler if you want the full AC waveform out of your power triac.

However, I digress. Following the circuit that you have attached, AC is really not your primary concern, as it is fully isolated from your DC supply. The relay coil and contacts are physically seperated in the relay housing, and it would take a serious fault to change that. In fact, I would guess that it is easier to create a short between the input & output of the optocoupler than it is to do the same in the relay (based solely on physical spacing, an IC mfr could verify if that is correct). The reason to isolate the relay coil from your micro is because of inductive kick from the coil, since it is using a magnetic field to move the contacts back and forth, not to mention relay coils usually require more voltage than a 5V output micro can directly supply. However, that flyback diode should be providing more than adequate protection from inductive kick. If you source the 5V opto power from your micro, and you have a well grounded circuit, you shouldn't have any issue with ground noise. A regulator for micro power also wouldn't be a bad idea.

EDIT: After looking at some datasheets, I take back what I said about the potential for a short. The isolation resistance of an optocoupler is roughly \(10^{10}\) ohms whereas a relay is more like \(10^9\) ohms.

 

Since your output is AC, you could replace your proposed circuitry with SSR's (Solid State Relays), which basically look electrically like an optocoupler-controlled TRIAC. You'll still need a resistor to limit current through the input side.

 

I like SgtWookie's suggestion. Much simpler.

http://www.clare.com/home/pdfs.nsf/www/CPC1998.pdf/$file/CPC1998.pdf

Here is an example of an SSR that could be used for your example.

 

The issue I have with a SSR is I need to use the relay as a change over relay, eg I use both throws of the relay, so I can't use a SSR. Unless a SPDT SSR relay that does not cost a arm and a leg exists.

I am going to take your advice b1u3sf4n09. Maybe I am just been over protective buy using a optocoupler here. The only fault condition I could think of is (massive) overvoltage but if that does happen the device I am trying to protect would have be blown anyway...

However, just for the record, would this DC-DC converter provide AC isolation? When the datasheet says 1000v DC, I would think that the isolation of AC would be as good. Would that be correct?

Thanks for the help guys.

 

An 'isolated' DC-DC converter will provide both AC and DC isolation, but will typically have MUCH greater capacitance between input and output than an opto isolator. This usually is not a problem, but it's good to understand, DC-DC converters also can be a nasty source of noise- on both input and output sides.

All isolation techniques have quirks, no free lunch.

 

Is it clear why the relays need to be isolated from the logic board? Relays already provide isolation between the coil and the contacts.

 

Is it clear why the relays need to be isolated from the logic board? Relays already provide isolation between the coil and the contacts.

Agreed. Without a schematic, or even a block diagram it is anyone's guess what is being done here.

That DC-DC converter probably also works for AC, but to what degree and what voltage is unknown. I would not use it for that reason.

 

Is it clear why the relays need to be isolated from the logic board? Relays already provide isolation between the coil and the contacts.

+1

You could use a 240vac coil relay,driven by SSR, and 240vac from load supply.