Continuing the series of taking a Fresh Roast SR500 apart leads us to the internal heat / air mechanisms. At this stage you have reached the components critical to any modifications of your roaster. I’m updating this post in December with photos taken back in October when I stripped the roaster down the rest of the way and began building my modifications.
Part 1 we started with the external screws to gain access to the internals. In Part 2 we separated the electronics between the high and low voltage and then lifted the high voltage board out with the heater/fan. This left us here:
Heater / Fan and Power board assembly
We have now reached the point where we unplug the old roaster boards and start looking at attaching alternate controls. At this point you are going to separate the metal connectors from the board (assuming you are replacing any of these parts or modifying it in some way) You will have a black plastic cover from the top, a metal cone underneath that funnels the hot air towards the vented top, and a black plastic pan with a fan sticking to the bottom. The pan will have three screws. You see one of the locations to the left side and another to the right. The third one is not visible in this photo due to it being on the back side of everything you see. If you remove the phillips screws from these locations it will allow the top plastic piece to be separated. The top metal piece is sandwiches between these plastic pieces and held in place with the screws mentioned above. The metal piece is sealed to the heater mechanisms inside the bottom black plastic pan using silicone sealant.
Once you lift the top covers off and break the silicone seal you see this:
and this:
The fan is firmly connected to the bottom and up into the blades. The fan is a straight sheet of metal leading from the middle out to the edges with a flat disc on top. It is not apparently simple to separate and seems quite “stuck” in place. The middle “axle”/hub of the fan does not appear to have an obvious way to disconnect it though I’m sure there is a way to do so.
Looking back at the heating area you will see the bimetallic switch and a temperature sensitive fuse.
In the center there are two bolts/nuts . These anchor the top part of the funnel to the heater coil. You will notice 4 spots that look like staples above. These are how several supports made of the same material that the fuse and switch are riveted to. There is a metal ring holding this all in place with a “washer” made out of the same material again. This material is a high temperature material often used in heat guns, hair dryers, and popcorn poppers. It is designed not to burn and to cool off quickly. There is no point in disconnecting the nuts you see and you are likely to damage something in the process when you try. Immediately under the center part is a small heater coil that connects to the fan. This coil always generates heat whether the system is on or in “cool”. It is used to lower the voltage from 120 volts to somewhere close to around 20 volts DC using resistance and the resulting energy given off by the coil. The remaining electricity leads out from the system to the “black box” rectifier on the bottom of the fan motor. The outer coil is on the opposite side of the slit and continues all the way around providing the majority of the heat. You can see the outer cool quite clearly in the photo below.
In the photo above and below you clearly see the temperature fuse. This device is called a ThermoDisc Microtemp thermal cutoff. There is a PDF that discusses the features of this particular device http://www.thermodisc.com/uploads/PDF-ecombro.pdf. It also has a diagram of the inner mechanisms that make this work.
It is a Thermodisc G4A01216C 216*C Cutoff. Once the red stuff melts (at 216 degrees Celcius) there is a spring inside that is released and it mashes the wire outside of the housing and no longer makes contact to allow the electric to flow. Once it fails the only way to repair it is to bypass it or to put a new one in. Since the mechanisms are anchored with a rivet they are not the easiest to source and replace but it is possible to do.
The last mechanism is a Klixon YS10 46b-s x9ab. This is a Klixon YS10 Series 150*C Beryllium Copper Arm, Standard Length Terminal (31.5). Normally once it triggers at 150 Celcius it then has to cool before it works again. The reset is set to 90 degrees normally and some models has a different offset. YS10 specifies the type, 46 is the temperature (150 degrees) b specifies Beryllium Copper (the bimetalic switch material) -s for standard. In the X9ab position this would possibly be where a different temperature reset amount would be specified. It does not clarify the numbers used there. It appeared in the PDF linked above as (XX) in the part number. This MIGHT be that it needs a 9 degree temperature drop on the switch before it engages again assuming it uses an X as a place holder rather than using a 0. Since it is not in parentheses this might simply be some sort of a plant number or production date rather than a temperature offset.
Modification:
Many modifiers like to disconnect both the fuse and the bi-metallic switch. I do not advise this unless you are absolutely sure what you are doing. If you wish to use the center heat coil separately or with the main coil together this is up to you. You will need to disconnect the white wire that leads to the fan by prying open the brass clip (under the heat shrink) and supplying your own transformer to around 20 volts AC to independently control the fan.
At the moment I’ve been using an Arduino to control a Q4015L5 triac with a MOC3052 and a H11AA1 as a zero crossing detector. There are two triacs each used to separately power the fan and the heat circuits. Each side’s gate is triggered with a MOC3052 opto-isolator by a single pin on the Arduino to a resistor through the moc’s infrared led and on to the ground pin. The H11AA1 works the opposite way triggering a led on the high voltage side and it measures the fall of the 60hz sine wave of US electric. Each fall signals the Arduino on another pin that is connected to a hardware interrupt on the processor. The interrupt sequence compares the fan and heat potentiometers to a map and then uses the lookup value to set the length of a processor timer. The timer then comes back and fires the fan or heat pin that fires the MOC3052 linking the triac and connecting power from the input side.
I disconnected the white wire from the fan and routed it with the primary heater coil so that they both run at the same time. On the fan side I connected it to a transformer from Radioshack that outputs 25VAC. This appears to adequately run the fan and delivers more air flow than normal due to the fan being “overdriven” from it’s normal voltages. While this is not good for long term use this could be useful if properly triggered in the programming for cooling and drying or initial heat up. In other words turn heat on at 70%. Start fan at 120% Gradually drop fan to 100%. Increase heat to 100% while dropping fan to 60%. Etc…
While running the system from a variac and through a watt/amp monitor I found the fan consuming approximately 50 watts at the original 100% air flow on through the new maximum around 70 watts for a 125 to 130% flow rate. Once I turned on the heater I found the original wattage level use set at 80 to 85% heat and what seemed like a normal heat output felt by hand in front of the output. Lately I had been getting around 1520 watts with momentary flutters up to 1580 at high and full fan before bypassing the controls. Now at 100% it was showing upper 1600 to 1750 watts for the brief few seconds before I turned it down.
I’m pretty certain this is not designed to run like this and would likely melt something if left to run this way on it’s own without some sort of “safety” override in the programming. What I would expect to be necessary is to mandate original 100% air flow before the heat can be turned past the usual level and the fan cannot be lowered until a specific number of seconds after the heat is dropped to a normal number. Additionally there should be a limit on the duration of this heat overdrive. This would be used to help drive the roast in a way many home roasters like to use to try reserving some heat until towards the end to drive it to second crack or some other nuance.
This should be thought of as some sort of reserved “afterburners” to a skilled pilot used only when necessary or someone pushing a nitrous injection button on a race car or turning on some super charger.
Once heat has been disabled it clearly cools off much faster than before. As of 10/10/11 I’m waiting to more firmly mount all the controls and switch to the new potentiometers before testing a roast. I also have a few buttons for start/stop and a microSD logger to setup first before I start this because I want to track thermocouple readings vs each of the percent settings etc so I can review it later after my first test. If anyone has a way I can sense the wattage use and feed it to the Arduino too please let me know. I’ve seen a few very LARGE devices intended for whole house sensing but I’m looking for something small….