I just purchased a Kill-A-Watt P4400 to help with our home power consumption, but mostly to use for review purposes and to better understand home coffee roasters, for me and for my reviews. These simple devices have an outlet, a row of buttons, and an LCD display on the front, On teh back is a standard wall plug. Plug the Kill-A-Watt into an outlet and plug a device up to 15 amps into the socket on the front of the Kill-A-Watt, and it displays various parameters. This affordable device (about $20 shipped) will display a number of things including KWH, amps, watts, Hz, and Volts.
One of the topics of discussion is the power level control of the Hottop KN-8828B. Although the meter will not show how the device manages the power levels, it does show what the power consumption is during each level. To measure, I started the roaster without beans and allowed it heat up and then enter the roasting cycle. With the fans off throughout, I cycled through the 11 steps (0% through 100% in 10% increments) and recorded the high and low wattage consumption of the roaster at each step, then graphed the findings.
After I posted the original graph, Ian suggested that it would be interesting to know the temperature level at each of the settings. At first I thought that would be a great idea, but after a bit of testing I found that the information was difficult to gather and of questionable value...
Chamber temperature would be interesting, as Ike suggested, except I found that it's not as easy as it sounds. The problem is that to do the test with beans could take a few pounds to do accurately. I tried it without beans.. It takes quite some time for the roaster to reach thermal equilibrium at each level. Even at five minutes per setting you can only do about 5 steps per test, and even then it is not long enough for the temperature to stabilize- ten minutes or even fifteen is needed. After the preheat cycle, the roaster is ramping up to the point that if you set the heater to 10%, after five minutes roast time it hits 325F. on the display. If you let it run at 10% for about ten or fifteen minutes, and use the fan to "hit the thermal brakes" so to speak, it will settle in at about 257F.
And then it is the other way round at 30%. For the five-minute-per-segment test I had a display reading 267F., but when allowed to run for some time longer at each of 15%, 20%, and then stabilize at 30%, I had a reading of 330F.
So... The only way to get accurate data would be to load 250g. of beans, set the temp to the desired level, and let it run until it held a temperature for a minute or so, then set it to the next temp and repeat. This would take two or maybe three pounds to achieve accurate results, and then, after all that, they would not be terribly meaningful nor useful because no one would roast that way.
Additionally, it takes a long time early on to stabilize, but also, at the end of the roast, at 80 percent, without beans, the temperature rises to the "magical" 228 mark and the beans eject automatically. So, once again, without beans in the roaster, getting a useful temperature reading at 90% or 100% is difficult, if not impossible.
It is sort of like saying, if you hold the gas pedal at half throttle, how fast are you going? or, how far should the throttle be depressed to go 55? It is more important to know what is the zero to 60 time? So, if someone thinks that data more accurate than this is in some way useful, please try to convince me to sacrifice the beans necessary to compile the data, but I do not think it is worth the beans to do the test. Even if done, differences in line voltage, bean density, bean moisture level, fan speed, and even the condition of the main filter will change the data.
OK, so I was an art major, am not a scientist, nor do I have any real knowledge of thermal dynamics, but here is the original graph with some figures added to show the APPROXIMATE temperatures at each heater level:
I will try to explain—
- The temperatures above the graphed line are those I manually recorded using a thermocouple located about midway between the bottom of the Hottop's sensor and the floor of the drum.
- The temperatures below the graph line are those I manually recorded from the Hottop's LCD temperature display.
- The first three temperatures in each set (10%, 20%, and 30%) were recorded using an extended period of time for each to try to achieve stability for meaningful readings. The remainder of the temps (40% and above) were done during a separate roast with about a five minute period of time per level (including teh 10-30% settings, the data from those three levels was not used for this graph).
With the heating element off (just the internal electronics and the drum motor powered), the consumption was 24 watts. With the heating element at 100%, about 756 watts.
At each step there was a high-low range; a fluctuation if you will, but the average deviation from high to low was about 1.2% for any pair of readings, and I would call that insignificant for the purpose of this discussion.
I do not know how this is being controlled by the roaster's electronics. I suspect that they are switching the power to the heating element in a binary mode in "duty cycles" (?) to create the power levels but that is just a guess. Maybe someone with an oscilloscope or frequency counter can measure that.
