Do I correctly guess this is part of a microwave oven? I understand your need to switch the relay during the correct half-cycle of the AC wave. However, you are not turning it on and then off again within the same 8.3 ms time, are you? Bob Ammerman RAm Systems ----- Original Message ----- From: To: Sent: Tuesday, October 29, 2002 9:40 AM Subject: Re: [EE]: Speeding up a relay - zero crossing detection > Hmmm, Roman, this capacitor idea has come up more than once. It sounds > like dangerous medicine, ringing is generally something to be avoided, but > a little RC network near the coil might just ring once (like the postman) > and then quit. I am using a 2n5551 to drive this thing, which is a high > voltage transiostor, so I can tolerate quite a lot of spike without > damaging the transistor. Right now, I am going to try a resistor in the > circuit. I reason thusly: the resistance of the coil is 1600 ohms and > rated voltage is 24 volts. So the moment I turn it off, a 1600 ohm > resistor in parallel would produce a 24V negative spike. If I double the > resistance, this voltage should go up by twice. Quadruple it, and I might > get a 100V pulse out of the thing. Since a 33V zener (which I tried > yesterday) did a lot of good but not enough, this might just do the trick. > I'll try 6400 ohms and see what that does. > > Under ideal conditions this relay is supposed to turn off in 2mS, which > would be peachy, but I fear that means it produces a 250V kickback pulse > in the process! > > Yes there is a PIC in this thing. The whole problem with the relay is, if > I don't turn off in less than one half 60HZ AC cycle, I get big noise > pulses off the relay contacts which get into everything including my PIC > power supply. Eventually one of these pulses either resets the PIC or > makes it do wierd things. > > Imagine the NO and NC contact of a relay. When the relay is firing, an > arc is drawn from the NC contact. If there is 120V from the NO to the NC > contact, as this arc ionizes the air in the gap, and some inductance > somewhere, the arc continues as the contacts move across the gap and these > contacts are momentarily shorted together! Yes this really happens, and > yes it happens every time a relay is fired. Look it up in Aromat (NAIS) > technical literature. In my circuit, the relay is choosing between > half-wave and full-wave power. As long as I fire during the correct phase > of the AC cycle, there is only a diode drop voltage across my relay > contacts. If I fire the relay during the wrong half, it momentarily shorts > the NO and NC contacts, generates a huge pulse of energy at 2.5megacycles > and about 1500 watts, and this literally goes everywhere. Imagine trying > to work your PIC on the same board as a 1500 watt AM station. I think the > FCC will not be happy about this either. No amount of shielding or > decoupling seems to fix it. After loading a blunderbuss full of > capacitors and firing them at my circuit board, I decided this problem > would be solved by subtler and more devious means. Thus trying to fire a > relay on a zero crossing, a quixotic quest at best! > > > -- Lawrence Lile > > > > > > > Roman Black > Sent by: pic microcontroller discussion list > 10/29/02 03:32 AM > Please respond to pic microcontroller discussion list > > > To: PICLIST@MITVMA.MIT.EDU > cc: > Subject: Re: [EE]: Speeding up a relay - zero crossing detection > > > > > To get fast relay opening you need to remove > the catch diode or add the zener in series with > it, or even a resistor in series with the diode > may be fast enough. > > Typically in the "old days" there were many > mechanical SMPS like this, popular in automotive > uses like the alternator voltage regulator and I > have seen some used in old military equipment like > step-up voltage converters. The most common way > of handling the problem is to use a capacitor in > the circuit across the coil similar to a tank circuit > that resonates the relay. This was also how they > made "personal protection" type HV generators > before the mosfet days. Try a search for "tesla > coil design" or something, you should be able to > get it buzzing very nicely once tuned properly. > > The final contact bounce may also be tunable with > some mechanical damping, I have seen weights or > even "blobs" of solder on the relay leafs which were > obviously added as mechanical tuning. > > This sounds like quite an interesting application > you are working on?? :o) > -Roman > > -- > http://www.piclist.com hint: PICList Posts must start with ONE topic: > [PIC]:,[SX]:,[AVR]: ->uP ONLY! [EE]:,[OT]: ->Other [BUY]:,[AD]: ->Ads > > -- > http://www.piclist.com hint: PICList Posts must start with ONE topic: > [PIC]:,[SX]:,[AVR]: ->uP ONLY! [EE]:,[OT]: ->Other [BUY]:,[AD]: ->Ads > > -- http://www.piclist.com hint: PICList Posts must start with ONE topic: [PIC]:,[SX]:,[AVR]: ->uP ONLY! [EE]:,[OT]: ->Other [BUY]:,[AD]: ->Ads