Protection diode

[g2c]

Hello everybody,

Can someone please state if D1 (4004) is correctly placed to protect the LJ thing when it starts or stop to conduct? Also, the reasoning is most appreciated

How does this contactless limit switch (for swing door controller) works? : When the door is not by the end of stroke , current flows to the coil and the relay is closed. When the door reaches a limit, the LJ conducts and shunts the current flowing to coil causing the relay to open which indicates to the controller to stop the motor. This thing is tested and works OK. My concern is for the longevity of the LJ metal detector whence my question.

Note - the components data is attached too.

Thanks a lot for your help!

 

[Diver300]

With that arrangement there is no for need a diode at all. When the transistor is off, the current through the coil will be limited by the resistor. When the transistor turns on, there will immediately be 0 V across the coil, so the current in the resistor will be 9/270 = 33 mA. Initially, the coil current won't have changed due to the coil inductance and will still be 11 mA, so only 22 mA will flow in the transistor.

Over the next few milliseconds, the coil current will reduce to very near zero, while the resistor current will stay the same, so the transistor current will increase to 33 mA. There are no voltage surges.

When the transistor turns off, there will be no current in the resistor, and so there will be 9 V across the coil. In a few milliseconds after the transistor turns off, the coil and resistor currents will increase up to 11 mA, while the voltage across the coil falls to around 6 V. Again, there are no surges.

Your circuit isn't like the normal circuits, because it is rare to permanently consume current whether the coil is on or off. In this case it's 33 mA at 9 V so about 300 mW, which isn't much for a circuit running from the mains. The result of your circuit layout is that the impedance seen by the coil is either 0 or 270 Ohms, so the 11 mA in the coil will not create a significant voltage, and no diode is needed.

The more usual circuit has the transistor in series with the coil. That way no power at all is used when the circuit is off. However, the impedance seen by the coil is either 0 (when the switch is on) or infinite, when the switch is off. Then any current from the coil will create an infinite voltage when the transistor turns off.

(Capacitance and other effects will limit the voltage, but it can easily be enough to damage the transistor).

In the more usual circuit, the impedance seen by the coil is reduced by adding a component in parallel with the coil. The voltage across the coil then depends on the coil current and the impedance of the additional component, so can easily be kept small by the right choice of component.

Common components are:-
Diode. Keeps the voltage down to 0.7V more than the supply voltage. Can slow down the opening of the relay which can be bad for the contacts. Needs to be the right way round.
Back to back zenners. Keeps the voltage down, but will be at least twice the supply voltage. More expensive than a diode.
A resistor. Cheap, and doesn't have to be the right way round. Keeps the voltage down, but uses extra power whenever the relay is on, and there is a straight compromise between the power used and how high the voltage rises when the relay turns off.

Proximity sensors will often contain zeners to prevent damage if no component is added, but the manufacturers may not admit that so that users add their own components.

In the place you have the diode, it won't do any harm. It certainly shouldn't be the other way round.

 

[g2c]

Thanks deeply for this detailed discussion. Ill learn it thoroughly. Ive chosen this arrangement to comply with controller's requirement of a NC 'end of stroke' indication

 

[g2c]

I understood perfectly the way to analyse transient voltage while an inductance 'sees' infinitely high resistance.

Now that I have finished (one door of my two doors singing gate) I realize that as is it's not good engineering as for the fault tolerance.

I think It'd be better to have a normally closed proximity sensor so that if any of the 3 wires brakes, the controller will consider an end of stroke and will idle the motors. Unfortunately I found no sensor like the one mentioned above which is normally conducting and stops conducting upon detection

 

[dr pepper]

There are (or at least was lat I looked) steerable proxy's, where a control line selects No or Nc, you might be limited in the style & size though.
Its common practice nowadays for nc machines to detect end of travel by monitoring motor current, you can do this is if the axis has a drive where current can be limited, I guess a door might not have such sophistication.

 

[g2c]

This controller (city 1 evo by V2) too monitors the current and usually stops in overload but my doors are made of panels (polystyrene sandwich) an are too fragile.

If you can remember of a device operating in nc or programmable, please let me know. I only need metal (steel) détection at up to 4mm.

 

[rjenkinsgb]

Something like this?

https://uk.farnell.com/multicomp/mcpin-t18l-011/sensor-m18-npn-shielded-nc/dp/2218041

 

[g2c]

Yes, thanks a lot! Was not aware. Price quite salty compared with the eBay version