Opto-triac

[su1c1d4l]

Im trying to control 220V 16watt fluorescent bulbs with electronic ballasts through the PIC. pde po ba ang MOC3021 for this application? pano po ung set up nya?

[DigitalMind]

mas ok kung relay na lang gagamitin mo kasi gagamit ka pa nang snubber sa triac. and marami nag sasabi na unreliable sa inductive loads even with a snubber..

but yes i don't see why hindi pwede ang moc3021 as a optoisolated triac driver.

kindly read the datasheet.

[su1c1d4l]

yes, i have read the data sheet. ung actual set up ko my kasamang LCD at 12vdc relay using c9013 as amplifiers.. kaso pg ng deactivate po ang relay (with 220vac 16w bulbs connected) , naapek2han ang LCD, lumalabas mga alien characters. im thinking of using something that would isolate the relay ckt from my PIC and LCD. anu po suggestions nyo?

[DigitalMind]

i think kailangan mo lang nang diode sa closest sa terminals nang relay.
back emf issue lang yan. 

[RaffT]

or search sa internet for SOLID STATES RELAY..

[DigitalMind]

http://e-gizmo.com/wordpress/?p=195

[labgruppen]

if you already have the "snubber" diode and the "aliens" are still appearing, then putting snubber across the load or across the relay contacts might help.

[su1c1d4l]

bale eto po ung schematic ko: i changed the resistors to 330ohms.


mali po ba ang mga components na ginamit ko?

[DigitalMind]

since may relay and transistor driver kna for the relay proceed to that path. medyo mas complex pa ung triac...
aslo lagi namin sinasabi dito na please lagyan nyo nang decoupling capacitor closest sa VCC VDD nang MCU.
a 1uf or .1uf ceramic cap will do.
it adds a TON of noise protection sa mcu mo.

[su1c1d4l]

meron po akong coupling caps malapit sa vdd ng PIC at isa sa LCD. actually po greater than 1uF nilagay ko. kaso electrolytic. mali po ba? LCD timer ksi ung function nya. i encounter no problems when the relays activate, only when they deactivate. my mali po ba sa circuit ko?

[kinyo]

baka makatulong lagyan ng capacitor across C-E ng transistors (1 or 0.1 uF din)

[su1c1d4l]

ok po...ung ibang components ba, like the 330 ohm and the 1n4001, are they okay? how about using electrolytic caps as coupling capacitors across vcc?

[su1c1d4l]

i think kailangan mo lang nang diode sa closest sa terminals nang relay.
back emf issue lang yan. 

sir, anu po ang purpose ng 22k across the protection diode?

[rdpzycho]

meron po akong coupling caps malapit sa vdd ng PIC at isa sa LCD. actually po greater than 1uF nilagay ko. kaso electrolytic. mali po ba? LCD timer ksi ung function nya. i encounter no problems when the relays activate, only when they deactivate. my mali po ba sa circuit ko?

kailangan meron ka pa ring 0.1uF or fast capacitors malapit sa supplies. maganda 'yung malalaking capacitors to localize current loop pero pagdating sa spikes, 'di nila kayang i-absorb kaagad.

pwedeng unang subukan 'yung 0.1uF across the CE (lagyan niyo rin ng resistor papunta sa GND para lang sigurado in case na mag Hi-Z 'yung pin ng MCU).

posible ring radiated emission ang nae-encounter mo dahil merong current sa line then puputulin, magkakaroon 'to ng maraming harmonics (kapag merong spark 'yung contacts, malaki na kaagad ang radiated emission, ganito rin ang ginagamit kong pang-test ng immunity dati sa radiated emissions, self-oscillating relay circuit). pwedeng ilayo 'yung relays sa MCU, lagyan ng shielding, or mag-off ka lang sa zero crossings nung AC line (pwedeng gamitin 'yung info na makukuha sa secondary ng transformer.

to minimize radiated emission, dapat 'yung lines ng ilaw mo e as close as possible sa isa't-isa (remember twisted pair or flat wires). pwede mo ring taasan ang immunity ng circuit by also minimizing current loops sa traces ng PCB.

actually, kahit na opto + triac ang gagamitin mo, may chances pa rin na ma-encounter mo 'yung same problems. malaki tulong nung .1uF sa pins ng MCU, but just remember in future applications, nababago niyan ang slew rate nung output pin ng MCU. kaya in high speed applications, mas madalas mong makikita 'yung common mode chokes in the form of clip on ferrites.

[su1c1d4l]

kailangan meron ka pa ring 0.1uF or fast capacitors malapit sa supplies. maganda 'yung malalaking capacitors to localize current loop pero pagdating sa spikes, 'di nila kayang i-absorb kaagad.

pwedeng unang subukan 'yung 0.1uF across the CE (lagyan niyo rin ng resistor papunta sa GND para lang sigurado in case na mag Hi-Z 'yung pin ng MCU).

posible ring radiated emission ang nae-encounter mo dahil merong current sa line then puputulin, magkakaroon 'to ng maraming harmonics (kapag merong spark 'yung contacts, malaki na kaagad ang radiated emission, ganito rin ang ginagamit kong pang-test ng immunity dati sa radiated emissions, self-oscillating relay circuit). pwedeng ilayo 'yung relays sa MCU, lagyan ng shielding, or mag-off ka lang sa zero crossings nung AC line (pwedeng gamitin 'yung info na makukuha sa secondary ng transformer.

to minimize radiated emission, dapat 'yung lines ng ilaw mo e as close as possible sa isa't-isa (remember twisted pair or flat wires). pwede mo ring taasan ang immunity ng circuit by also minimizing current loops sa traces ng PCB.

actually, kahit na opto + triac ang gagamitin mo, may chances pa rin na ma-encounter mo 'yung same problems. malaki tulong nung .1uF sa pins ng MCU, but just remember in future applications, nababago niyan ang slew rate nung output pin ng MCU. kaya in high speed applications, mas madalas mong makikita 'yung common mode chokes in the form of clip on ferrites.

wow...it packs a lot of info..thanks. i think im going to try first putting 1uF or .1uf as coupling caps near the vdd of the pic and lcd; as well as to the C-E of the transistors. if im going to use optocouplers + relays, there is still a chance that i would encounter the same problem?

pwedeng unang subukan 'yung 0.1uF across the CE (lagyan niyo rin ng resistor papunta sa GND para lang sigurado in case na mag Hi-Z 'yung pin ng MCU).

BTW, panu pla ung set up pg nglagay ng caps across collector to emitter? parang hnd ko ata gets. and panu po mglagay ng shielding sa mga relays?

[rdpzycho]

parallel lang. use ceramic capacitors, mas mabilis response nila, 'yung electrolytics, inductive ang behavior nun sa higher frequencies.

sa suuply mas maganda talaga mataas. sa SMD merong MLCC (multilayer ceramic capacitor), merong matataas na values pero mabilis pa rin response. pero kung wala talaga, high value electrolytic, parallel sa ceramic.

actually, 'di practical i-shield ang relay, test lang kung radiated ang problem. kuha ka lang ng de-lata na walang laman tapos itakip sa relay.

[su1c1d4l]

parallel lang. use ceramic capacitors, mas mabilis response nila, 'yung electrolytics, inductive ang behavior nun sa higher frequencies.

sa suuply mas maganda talaga mataas. sa SMD merong MLCC (multilayer ceramic capacitor), merong matataas na values pero mabilis pa rin response. pero kung wala talaga, high value electrolytic, parallel sa ceramic.

actually, 'di practical i-shield ang relay, test lang kung radiated ang problem. kuha ka lang ng de-lata na walang laman tapos itakip sa relay.

i think kailangan mo lang nang diode sa closest sa terminals nang relay.
back emf issue lang yan. 

@rdzpzycho: ganito po ba ibig nyo sabihin pag lagyan ng caps sa CE? Sir pag ng lagay aq ng 100nF caps near vdd of PIC, tatangalin ko ba ang 2200uF caps ko? or pde i-retain lg?

[rdpzycho]

hindi mo tatanggalin 'yung malaki mong capacitor. reservoir 'yun, kapag kailangan ng current nung PIC or ng any device, sa nearest capacitor muna siya kukuha dahil madalas siya ang may lowest internal resistance sa supply line. ginagamit din siya to localize current loop.

'yung smaller caps naman, para sa higher frequencies. respond sa higher frequency or kung may higher frequency component 'yung current na kailangan nung device.

yup, ganyan nga. tries to maintain the voltage flat at the input of the transistor. though pwede mo rin 'yang ilagay doon sa pin ng MCU.

[nucleus]

merong opto triac na may zero crossing detector para mabawasan ang noise na nagegenerate sa pagswitchon ng load specially inductive ones

[TinTopHack]

Below are my comments/recommendation - hoping you will find them helpful
Ckt # 1

For this circuit, the 100 ohm resistors at the base are too low. Changing them to 330 ohms -> still too low
if you are controlling the transistors using an MCU. With 9013 transistors a base current of 1mA should be enough to drive small relays at the collector so the base resistors should be 3K3 to 4k7 in value.

You mentioned that when the lamps are connected then the relays are deactivated/activated the LCD display is affected. It would be interestign to see how you connect your LCD so we can identify possible noise entry points. In the absence of such info, please check what happens if the lamps are NOT connected and you deactivate/activate the relays. If the LCD is NOT affected this time then your problem is related to the switching currents of the lamps - being flourescent lamps as you said. CFL or flourescent lamps then being inductive are generating a lot of EMI during turn on and off so you need to put EMI filter on your AC input. You may also need to put RC snubber across your relay contacts if your lamps are CFL or flourescent otherwise your relay contacts will not last long.

You also need to review your PCB layout - proper grounding, start grounding, etc. Poor layout makes your circuit susceptible to EMI.

Ckt #2


R1 22K should be removed. It could keep your transistor turned on all the time because it supplies unwanted bias to the base. The cap at the base will only slow down the switching of the the transistor. It will not help much on solving your problem. The diode across the relay coil is enough to suppress its inductive spike and noise. If you are worried about spikes entering the MCU pin then put a 1n4148 diode (anode to MCU pin + cathode to resistor) in series with the base resistor. This diode will block any voltage higher than 5V from entering the MCU pin.