Post by thread:Relay problem

Whenever the 5V output from PIC is applied to relay, the microprocessor will run unexpectedly. If I put relay in a separated board, it's OK. I think the problem is the noise from the inductance of relay. Can anyone tell me how to eliminate this problem? Thank you. ________________________________________________________ NetZero - We believe in a FREE Internet. Shouldn't you? Get your FREE Internet Access and Email at http://www.netzero.net/download.html

When the microprocessor turns off the current through the relay, the magnetic fi eld around the relay coil collapses, and a momentary high voltage appears across the relay. The best cure is to put a silicon diode across the relay, which wil l allow current to flow from the relay coil through the diode starting at the mo ment the PIC turns its output off. The diode should be arranged so that it is b ackwards for current flow from the pic pin to the voltage source. If one end of the relay is connected to +5 volts, then the cathode of the diode should be con nected to the +5 volt end. In other words, the arrow for the diode 'points' to the +5 volts. This 'fix' will make the relay take a few extra milliseconds to 'release', but in most applications this is not an issue. It helps to make sure that the +5 of the relay has a separate wire back to the + 5volt source, instead of sharing the wire or trace that brings +5 Volts to the P IC. Regards, Ron Fial At 10:06 PM 4/6/99 -0700, you wrote: {Quote hidden}> Whenever the 5V output from PIC is applied to relay, the microprocessor >will run unexpectedly. If I put relay in a separated board, it's OK. I think >the problem is the noise from the inductance of relay. Can anyone tell me >how to eliminate this problem? Thank you. > >________________________________________________________ >NetZero - We believe in a FREE Internet. Shouldn't you? >Get your FREE Internet Access and Email at >http://www.netzero.net/download.html >

> Whenever the 5V output from PIC is applied to relay, the microprocessor >will run unexpectedly. If I put relay in a separated board, it's OK. I think >the problem is the noise from the inductance of relay. Can anyone tell me >how to eliminate this problem? Thank you. Hi Peter. I'll try not to avoid the dyslexia that plagued me Sunday when I tried to help someone else with a similar problem. :-( If you could find a relay that would operate on 20mA of current, you could theoretically drive it direct from the PIC, but it would still be a bad idea. Don't drive the relay directly. Drive the base of a switching transistor thru a series resistor. For a typical low voltage relay, use a 2N2222 and put the relay in the collector circuit. Use a 470 ohm resistor to drive the base. In parallel with the relay coil, place a general-purpose diode with the CATHODE toward the (+) supply voltage. It's a good idea to return the (-) side of the relay supply directly to the transistor emitter, which should also be logic ground. Make sure that MCLR is asserted high and that your supply will deliver the current your circuit is demanding from it. If that doesn't stop the resets, email me directly and I'll try to help you further. Reg Neale

I have found that the PIC is very reliable and tolernat to noice. BUT if you don't have the 0.01 myfarad capacitor across it's Vcc/Vdd, it is very sensitive. In your case maybe try«0.1 AND 1000pF instead. Close to the chip. Sven -----Ursprungligt meddelande----- FrŒn: Peter Tran <spam_OUTltranTakeThisOuTNETZERO.NET> Till: .....PICLISTKILLspam@spam@MITVMA.MIT.EDU <PICLISTKILLspamMITVMA.MIT.EDU> Datum: den 7 april 1999 16:12 €mne: Relay problem {Quote hidden}> Whenever the 5V output from PIC is applied to relay, the microprocessor >will run unexpectedly. If I put relay in a separated board, it's OK. I think >the problem is the noise from the inductance of relay. Can anyone tell me >how to eliminate this problem? Thank you. > >________________________________________________________ >NetZero - We believe in a FREE Internet. Shouldn't you? >Get your FREE Internet Access and Email at >http://www.netzero.net/download.html >

Just to be the pest of the day... :) what means myfarad? I don't have any farad, do you? :) of course everybody knows you meant uF (micro Farad), but "uF" is easy and the correct way to write it... I told you... to be the pest of the day... :) Wagner. MILTON MEDICINTEKNIK KB wrote: > > I have found that the PIC is very reliable and tolernat to noice. BUT if you don't have the 0.01 myfarad capacitor across it's Vcc/Vdd, it is very sensitive. In your case maybe try«0.1 AND 1000pF instead. Close to the chip. >

Thank you all for your help. My relay is driven by a 2N3904, whose base is connected to the PIC through a 4.7K resistor. I did use a diode (cathode to +) across the relay, and the 0.01uF cap between Vdd and ground. However, whenever the PIC output high, the serial communication between the PIC and the DS1620(thermostat chip) become unexpected. I think the noise or the magnetic field from the coil of relay might: 1. interfere the communication 2. affect to the normal working conditions of the PIC or the DS1620 Please give me some advises. Thank you. {Original Message removed}

>Thank you all for your help. My relay is driven by a 2N3904, whose base is >connected to the PIC through a 4.7K resistor. I did use a diode (cathode to >+) across the relay, and the 0.01uF cap between Vdd and ground. However, >whenever the PIC output high, the serial communication between the PIC and >the DS1620(thermostat chip) become unexpected. I think the noise or the >magnetic field from the coil of relay might: >1. interfere the communication >2. affect to the normal working conditions of the PIC or the DS1620 > Please give me some advises. Thank you. > Peter: Is this a 5V relay, operating from the same supply that powers the PIC? In that case, if your theory about EMR is correct, temporarily supplying the relay from a separate power supply will make no difference. If that is indeed the case, then you have to consider parts placement, shielding, careful layout of power and ground busses, perhaps an RC snubber and/or varistor in addition to the relay diode, etc. Your 4.7K resistor will give you only about 1 mA of base drive for the transistor. A 2N3904 transistor has a min beta of 30 at 100 mA, so you are probably not driving it into saturation. 470 ohms would be better. What are the other PIC terminals doing? If pins are left as inputs, are they pulled up or down so that they are in an assigned state? Is MCLR driven by a brownout circuit, or tied to +5 thru a low-value resistor? Do you have a scope to look at what the supply voltage is doing during the glitch? 0.01 is not a big enough capacitor for many power supplies. Leave it there and add a 100uF in parallel. Just a few ideas; doubtless others will suggest more. Let us know what happens. Reg Neale

Hmm, Peter, I do this all the time without much trouble. Another post suggested using a lower value drive resistor, I usually use 2K. The diode is good. Is there a way you can avoid doing serial communications when the relay is being driven on or off? Throw away the current reading from the temperature chip and get another one? Tell the chip to shut up for a minute (nicely) ? I assume you are describing a problem that happens when the PIC output changes. are you using any kind of brownout protection at the PIC? Perhaps a temporary low voltage is causing the PIC to go haywire either WITH or WITHOUT brownout. WITH, the PIC skips a serial beat, WITHOUT the PIC may do random things.... {Original Message removed}

>>Thank you all for your help. My relay is driven by a 2N3904, whose base is >>connected to the PIC through a 4.7K resistor. I did use a diode (cathode to >>+) across the relay, and the 0.01uF cap between Vdd and ground. However, >>whenever the PIC output high, the serial communication between the PIC and >>the DS1620(thermostat chip) become unexpected. I think the noise or the >>magnetic field from the coil of relay might: >>1. interfere the communication >>2. affect to the normal working conditions of the PIC or the DS1620 >> Please give me some advises. Thank you. Making that cap larger will probably help. It has to handle the transients when the relay is turned on and off. On isn't so much of a problem, but 10-100uF would be a lot better. Also, when the relay turns off, the diode returns the energy in the collapsing field to VCC, which can cause VCC to spike high if you don't have enough C online to absorb the energy.

On Sun, 4 Apr 1999 Reginald Neale <.....nealeKILLspam.....SERVTECH.COM> wrote: > You didn't say what the design voltage of the relay is. In any > case, there is no way you can drive this relay directly from a PIC > pin. Never say never. Jameco sells some 5VDC DIP reed relays from Excel Cell Electronics with a coil resistance of 500 ohms, and any PIC can source or sink 10mA. There's even a model that has a diode across the coil, so you don't have to add one externally. Brian