At 06:22 PM 12/30/2011, you wrote: >Hi, > >I'm building some simple pic-based thermostats, which will have to >keep some technical rooms above freezing. >Controller part is fine, and one version of power part is fine, too. >It's based on MOC3062 optocoupler and BT138 triac. Connections are >directly as per MOC's datasheet, with resistors of 360 ohms. >I used ordinary 1/4W resistors, without any problem, that device is >working fine in test mode, switching a 500w heater. >(230 VAC 50Hz here) > >For another application, halogen lamp is used as a heater, mainly >because they can be obtained really really cheap, and will direct the >heat, so I can mount it out of way. But those burn out very quickly >when switching on-off too often. >So, I decided to use one-way thyristors instead of triac, so i could >turn on half power initially, then power another one in opposite >direction after a delay. Looking thru available info, it seemed like >same circuit should be fine, in this case, two of them. But the gate >resistor is letting out magic smoke, it receives much more power! >Thyristors used are BT151, from datasheet their gate currents are even >smaller than for BT138. > >I have forgotten many things about these, so probably I'm missing >something. Probably MOC3062 can not detect zero crossings, and is >feeding gate current differently? Probably thyristor requires >different circuit? Any ideas? The SCR will block in the reverse direction, but the triac-output optocoupler will turn on. The SCR will not trigger, but the gate will break down with reverse voltage (well in excess of the abs max, which is usually just a few volts), so gate current will flow- peak of maybe 0.5A (on the negative half-cycle- think of the load in series with the 360 ohm resistor in series with the optocoupler output in series with a zener diode representing the SCR gate-- all across the 230V for the negative half cycles). The result is too much heat in the SCR gate (maybe > 1W) and WAY too much heat in the series resistor (>50 watts). To fix it, just put something like a 1A 1kV diode in series with the opto output- that should work. In fact, you can retain the triac and add the diode to get half cycles- short the diode to get full power. (I suppose you know that drawing much current with just half cycles is frowned upon- it creates harmonics and can overheat transformers-- since you've only got a few A it shouldn't be a problem). Also, thyristors (including triacs) tend to fail 'on' so please take that into account in your safety considerations. --sp >I really really do not want to just use bigger wattage resistors, >since whole device will be mounted in not-so-big plastic case and >unnecessary heat is unwelcome. > >-- >KPL >-- >http://www.piclist.com PIC/SX FAQ & list archive >View/change your membership options at >http://mailman.mit.edu/mailman/listinfo/piclist --=20 http://www.piclist.com PIC/SX FAQ & list archive View/change your membership options at http://mailman.mit.edu/mailman/listinfo/piclist .