:: Detecting an Open Circuit ::

[suby786]

I have a triac that interfaces a PIC output (ttl) to a bayonet 240v AC bulb...isolated by opto-isolater... but what if the bulb blows, or the bulb is removed, i would like to detect this and i know that in these instances, no current will flow so id need a sort of meter to detect current, not accurate measurement just whether CURRENT DRAW or NO CURRENT DRAW, and feed this back to my PIC as an input. Any ideas how i could go about this?

 

[windozeuser]

why not just a relay with the current going through the relays coil. The output will be a NC contact so when there is no fault the detector is not triggered. But if theres an open in the wire the NC contact will close and you can trigger the detector

 

[penoy_balut]

use diode, dropping resistor driving an opto isolator connected accross triac main terminals

 

[kamdy]

Use a current transformer . It will give you isolation from the mains power.Connect a resistor across the output of the current transformer .This will give you a voltage proportional to the current drawn by the traic then use a comparator or ADC to read this voltage By PIC

 

[ljcox]

Use a current transformer . It will give you isolation from the mains power.Connect a resistor across the output of the current transformer .This will give you a voltage proportional to the current drawn by the traic then use a comparator or ADC to read this voltage By PIC

This is feasible, but a current transformer may be expensive and is potentially dangerous.

A simpler, safer and cheaper solution would be to connect a 5V1 Zener Diode is series with the light globe and connect the diode in the optocoupler in series with a resistor across the Zener such that the opto diode points in the opposite direction to the Zener. Thus on the half cycle that the Zener is reverse biassed, about 5.1 Volt will be applied to the series combination of the resistor and opto diode. The resistor will limit the current to a safe level

Connect the other side of the opto to the PIC. The signal to the PIC will be half wave rectified.

 

[suby786]

ok but remember ill be varying the brightness of the bulb hence if i want the bulb2be REALLY DIM, only a small portion of the AC wave will be passed through the TRIAC and hence would i really get a half wave signal?

has any1 got a circuit they cud quickly draw to show me what u mean... as im a bit confused with what u mean?

 

[Hero999]

Try installing a small LDR right next to the bulb, then you can have feedback as to whether it's working or not.

 

[suby786]

although an LDR would work, its not feasable as the electronics will all be in a nice metal box... and just 4 PLUG terminals for the lights..

its for a theatre system, hence my electronics is for lights that are ALREADY there...i can only sense if current being drawn or not...so the zener diode seems2be better idea..but if some1 can explain how... i know a bit about zeners and ive also done a search on their operation...its just linking that to my application...can some 1 help further?

 

[Comcat]

Referring to the attached graphic, I think you could detect a blown or removed bulb by monitoring the voltage at the line identified as "sense" while the triac is turned off. You could periodically but briefly command the triac to stop conducting and monitor the "sense" voltage while the triac off. If the line voltage is present, the bulb is OK. If there is no voltage present, the bulb is blown or removed.

 

[Hero999]

This is feasible, but a current transformer may be expensive and is potentially dangerous.

I disagree, a current transformer is probably the safest option, because it avoids having a direct connection to the mains.

You can make one with a small mains transformer.

Connect the low voltage secondary in series with the bulb with a small resistor in parallel with it (1hm: to 10hm the connect the mains primary to your circuit with a couple of back to back 5.1V zenners, a bridge rectifer and smoothing capacitor and discharge resistor on the output. This will be the input to your monitoring circuit.

 

[Comcat]

I was not suggesting connecting the mains directly back to the PIC. I was attempting to convey the concept of monitoring a sense voltage to sense a blown bulb. Of course isolation is a consideration and in this particular case I would use opto isolation.

 

[suby786]

Hmm i dont understand about checkin the triac when its off.. what if i have commanded the triac to be on all the time...i.e. 100% brightness... and i thought after a triac has been triggered, u cant turn it off until it passes zero (where it turns off anyway)

 

[Comcat]

At 50 Hz, there is a pass through zero every 10 mS and at 60 Hz it occurs every 8.3 mS. You should be able to periodically command the triac to turn off, read the sense line and turn the triac on again without noticing that the light was off for a fraction of a second.
Keep in mind that 50 Hz lighting is constantly turning on and off every 10 mS and is not perceptible to the human eye. Would a periodic 30 or 40 mS blip be a problem in your application?

 

[suby786]

but how do you turn the triac off, all it needs is a small timed trigger for it to turn ON ...and it will turn OFF when the 10ms is up... thats the function of the triac i believe?

also i dont understand the sense line, when the triac is OFF... what would the sense line be showing... 1) when there IS a working bulb and
2) when there ISNT a working bulb

i assumed if u t urn off the triac... you wont get ANY voltage

also i think ur diagram is a bit wrong..

shouldnt the bulb be on the NEUTRAL side of the triac, i.e. when off, there will be NO 50hz signal going thru the bulb...only when the TRIAC is on, is when the LIVE feed goes THRU the bulb?

 

[Hero999]

The only problem with voltage sensing is that would still require a direct connection from the mains to the PIC.

If you want to check the bulb when the traic is off then connecting the primary of a small mains transformer across the triac would be a much safer solution.

 

[Comcat]

The triac can be turn off by removing the gate voltage. When the gate voltage is removed the triac will turn off on the next transition through zero.

With the triac off and a good bulb installed, the full line voltage will be present at the sense line when measured with respect to the neutral.

If the bulb is blown, the sense voltage will be zero when measured with respect to the neutral.

The circuit will work with the triac in the neutral line and I've shown it there specifically to allow measurement of the sense line with respect to the neutral.

I could have shown the triac in the line side but that would necessitate measuring of the sense with respect to the high side of the line.

 

[Comcat]

The only problem with voltage sensing is that would still require a direct connection from the mains to the PIC.

If you want to check the bulb when the traic is off then connecting the primary of a small mains transformer across the triac would be a much safer solution.

The line voltage does not have to be connected back to the PIC. An opto-isolator can be used here. As you suggest, a transformer connected across the triac would provide isolation but unless the transformer primary was a high impedance, there would a current draw when the triac is off, resulting in the bulb be slightly illuminated even when the PIC has commanded the bulb off.

 

[Hero999]

An opto-isolator can be used here.

That's another way of doing it, and might also be more compact depending on how much room the ballast will consume and possibly cheaper too, if you can get two isolators in one package then it would be prefect.

As you suggest, a transformer connected across the triac would provide isolation but unless the transformer primary was a high impedance, there would a current draw when the triac is off, resulting in the bulb be slightly illuminated even when the PIC has commanded the bulb off.

Unless the transformer is huge the current draw will be far too small to light a 100W bulb.

 

[Comcat]

True.
A small power adapter transformer would work fine.

 

[suby786]

id feel more confortable using an opto isolater network...and feed that to the PIC...i only know of a ZERO CROSS opto isolate circuit that creates a pulse when it passes zero.. after 10ms the pulse goes low til the next cycle... can i use this or can i have a circuit that detects a difference in POTENTIAL... but if there is an open circuit, what wou;d the voltage be against the N line? it would be floating wouldn it?