Peter L. Peres wrote: > Hi all, > > I have done some testing and have found out that most (I tried three) > electric ovens and toasters have a max temperature of 150-170C after > which an internal safety thermostat turns off the heater no matter > what. > > Can anyone confirm this ? Esp. Lawrence L. who should be an expert on > this. I seem to remember that even fondue is limited under 200C, > because over 200C something or other is produced that may be toxic to > people (something carbonized or such). > > Along the same lines, would it be safe (from the point of view of > insulation, melting materials etc) to tamper with this thermostat > just a little bit ? Say, 50 degrees up ? > > Peter Hi Peter. I just ordered an electronic thermometer with a type K thermocouple to build the temperature control for my oven. The initial tests show that the thermostat is faaaar from being helpful. The time the temperature takes to get to the thermostat, it bounced very dangerously for the SMD components, I mean, the peak temperature inside t= he oven goes way up until the thermostat reacts, it has a large time delay. Using 3 x 1N4148 diodes in series, I got the following reading of VDROP when exposing them into a professional oven; 80F... 1.82V 350F.. 1.40V 400F.. 1.28V 450F.. 1.19V 500F.. 1.09V 550F.. 1.04V Setting the thermostat of my cheap kitchen oven to 350F, the same 3 diode= s hanging in the middle of the toaster, generated the following VDROP; Warm 1.41V 1.41V 1.41V \ .' \ .' \ \ .' \ .' \ 60s\ 300s.' 60s\ 300s.' 60s\ \ .' \ .' \ \ .' \ .' \ 1.088V 1.088V 1.088V Hot Both elements turn on for 60 seconds, the diodes takes down VDROP to 1.088V, elements off for 300 seconds (5 minutes), diodes takes VDROP up t= o 1.41V, turn on elements for 60 seconds, etc. From the relative VDROP from the prof oven (darn until I don't have the digital thermometer), it seems that the elements heat up the temp to more than 500F for the thermostat turn them off, then the oven cools down to around 220F to turn the elements on again. For cooking food this temperature ping pong should be ok, since a beef ca= n take the temperature average, but for a tinny poor SMD transistor or capacitor, this will be a killing. The worse part is when you take the oven from cool to 200F at the first stage of SMD soldering. Setting the thermostat to 200F and turning on th= e toaster, it takes 70 seconds with both elements full power, then the thermostat turn then off. There is an enourmous temperature delta from th= e oven chamber to the thermostat chamber. From what I noticed, 70 seconds = of full power can take the oven chamber to more than 500F during the last 15 seconds. This can probably force a premature damage in several SMDs in near future. Setting the thermostat for 200F, it seems that the temperature goes this way into the oven; 500F .-'-.OFF OFF .' | '-. -'-. / | '-. .' | '-. 200F---------/----|---------'-.------------/---|-----'-. / | '-. / | 150F .' | '-. / | 130F .' | '-.' | .-' | ON | 80F-'ON----70s--->|<--------9min------->|-30s--| --------- The 3 diodes were connected using small metal clips, since solder would n= ot hold under 450=B0F. Used high temperature wire for 500=B0F to connect th= e diodes to outside of the oven. The same metal clips tied the diodes to t= he wires. One could build a simple microcontroller to drive a solid-state-relay (or TRIAC with a zero-crossing opto-coupler), based on the diodes VDROP. It would react much faster than the horrible thermostat. To get better clos= e to the real PCB temperature, the diodes could be assembled on a real smal= l PCB (no soldering) and it could be sit side by side with the actual board= s being soldered. As the microcontroller job would be to control the oven temperature for 200, 320 and 450F (SMD soldering profile), it needs to keep track of the diodes VDROP for those 3 temps, something around 1.55V for 200F (4 minutes), 1.42V for 320F (2 minutes) and 1.19V for 450F (30 to 60 seconds= ). This voltage is perfectly traceable by any uC with an internal ADC or at least a comparator (as the AVR AT90S2313). For the AVR 2313, I thought to use something simple like that; +5V--o-----o-----------> AVR | | R10k R10k | | .-----------. | | | 2313 | o-----|------|+ | | | |Comparator |----> Relay _V_ o------|- | | | | | _V_ | '-----------' | | | | _V_ o--R1--->|---' | | | | | o--R2--->|------' | | | R2k | | _|_ _|_ GND GND Port pins driving R1 and R2 down change the voltage divider at the Comparator (-) input. The possible 3 states should be very closer to the diodes VDROP at the 3 tracked temperatures. Another easy solution is using a timer to set a PWM value over a port pin= , this feed a capacitor acting as a filter to be compared to the Diodes VDROP; +5V--o----------------> AVR | R10k | .-----------. | | 2313 | o-----------|+ | | |Comparator |----> Relay _V_ .-----|- | | | | | _V_ | '-----------' | | | _V_ o-----R1-----' _ _ | | _| |___| |___ | _|_ PWM Output | --- Cap | | | | _|_ _|_ GND GND The PWM output sets a voltage over the cap, to be compared to the diodes VDROP. Once the Comparator output is "1" (+ input is more positive than the - input) it means the oven is cooler than expected and the uC should turn o= n the relay (heater elements). Remember, the diodes reduce their VDROP wit= h higher temperatures. Of course this is the "poor's man temperature controller", and I wonder i= f diodes from the same lot could have different VDROP profile. If not enough port pins, you could use this simple design; +5V--o----------------> AVR | R10k | .-----------. | | 2313 | o-----------|+ | | |Comparator |----> Relay _V_ .-----|- | | | | |---. _V_ R1 '-----------' | | | |_|< Buzzer _V_ | _ _ | | _| |___| |___ | _|_ PWM Output | --- Cap | | | | _|_ _|_ GND GND The (-) comparator input can be switched to output and feed PWM for at least 10 complete waveforms, then switched to input and used as the comparator input, then switched back to output and so on. The AVR can drive or sink 20mA at any port pin, enough to drive directly any Solid-State Relay 3-32Vdc control, 240V @ 25A unit. All this temp control will cost less than $15 ($10 for the solid-state relay). The only problem is to physically install everything, since the damn oven heats all covers to almost the internal temperature. Probably = it should be an external box that the power cord goes in middle, along with the diodes wires to the oven. The idea is to turn the thermostat to 500F (full power) and feed power to the whole set, the AVR resets and starts the SMD soldering profile, powering off and alarming at the end. This oven has a special thermostat position named BROIL that turns full power to the elements via an extra switch tied to the thermostat shaft. = If things went not so hot into the thermostat chamber, everything could be installed there, and this switch feeds power to the electronics. It means that BROIL position would start the SMD profile soldering, but... it is t= oo hot inside there. VV46NER -- http://www.piclist.com hint: The PICList is archived three different ways. See http://www.piclist.com/#archives for details.