Do I need a galvanic decoupling between a controller and a power line? 2

[KRYM2006]

Hi,

We would like to develop a simple temperature controller to control a resistive electrical heater. A sketch of projected schematic is attached to this message.

Does that insulating transformer (T1) make a lot of sense? Or it is a waste of money?

The maximum current through a heater (H1) is 20A. I did not estimate expected short circuit current through the heater but I assumed it would not exceed 2-3kA, while I could be wrong.

Thank you for your help!

 

[MikeHalloran]

The simplest temperature controls use bimetallic switches to break the line current directly. Old electric flatirons would be an example. New ones without solid state controls might be hard to find. I think triacs and such are now cheaper than durable high current contacts.

The transformer doesn't buy you anything in terms of touch access limitations, e.g. UL requirements. Shouldn't the secondary be grounded somewhere?

Where is the temperature sensing element? I.e., what's inside the 'PID'?




Mike Halloran
Pembroke Pines, FL, USA

 

[ScottyUK]

The transformer is necessary if you want to be able to connect a standard sensor to the PID controller, because most sensors aren't designed to operate at line potential. If this is a commercial product for industry then most instrument techs won't be happy with the sensor at line potential. Tiffies are afraid of anything higher than 110V. If you allow the sensor to float at line potential then you need to design it specifically for that duty - which means custom-built and expensive. A transformer big enough to run that PID block won't be more than a few pounds / dollars and you might be able to get a small modular SMPS for even less.

3kA sounds low, even for a domestic installation at 230V. My domestic fault level is over 7kA which is pretty high by UK standards, but I'm not far from the substation. If it's an industrial installation you could easily exceed 3kA, unless you use an HRC fuse to limit the fault. Most semiconductor switches are protected by an ultrafast fuse designed for the purpose, although they aren't used as often as they should be because they are quite pricey. Sometimes it's cheaper to blow the silicon to protect the fuse.


----------------------------------

If we learn from our mistakes I'm getting a great education!

 

[OperaHouse]

Making the transformer 24V would make sense. That would make a lot of the wiring touch safe. Lots of controllers are available 24V.

 

[KRYM2006]

>> Shouldn't the secondary be grounded somewhere?
Mmmm... It is not planed to be grounded. Should it be? The heatsink of the SSR will be connected to PE but it is different story, I believe.

>> Where is the temperature sensing element?
PID is a standard temperature controller in DIN format, like OMRON or GEFRAN. A temperature sensor will be connected to it and will be physically placed close to the heater, while it can be anywhere. I omitted the temperature sensor in the sketch.

 

[KRYM2006]

Thanks for your help, guys!

I believe, a sensor at the line potential is not a concern in this case. We are going to use something like GEFRAN 600. 230V will connect to a power supply input and a sensor will be connected to a temperature input.

The only reason for that transformer was protection of control block from surprises coming from faulty load. Problem is I do not clearly understand if a load fault can generate something dangerous for the controller from power line side.

 

[VEBill]

Does the PID have a built-in power supply, or does it use a CPU chip that runs directly on 230 VAC?

;-)

 

[KRYM2006]

SoC with integrated 230V power supply.

 

[LionelHutz]

The GEFRAN would be designed to use the 230VAC line power so there is no need to stick an isolation transformer in there. Well, you might want stick something like a DIN mounted power filter in there but that'd be about it.

 

[KRYM2006]

Many thanks for all your useful inputs!
I removed that transformer from the project as an obsolete part.