JGG Senior Member Joined: 31 Mar 2006 Posts: 547 Location: Kentucky, US Expertise: I like coffee
Espresso: PID Silvia; PID Alexia Grinder: Mazzer Mini E (A), SJ; Rocky... Roaster: Hottop D w/PID; Thermometric...
Posted Sat Jul 26, 2008, 5:38pm Subject: Hottop PID project
After purchasing a Hottop, I found surprisingly little information regarding PID installation. Now that I have successfully completed phase I of my PID project, I thought I would post some information for 2 reasons:
To summarize the information that I was able to find online, and
To solicit input and suggestions regarding the resulting roast profiles.
First, credit to those whose ideas I borrowed:
CafeIKE for his helpful H-B and CG posts, as well as the information posted on his website related to heater wiring and environmental probe positioning (click here)
Randy G for his Hottop reviews and bean mass thermocouple probe placement (click here)
Jim Schulman for his advice on H-B to not allow the environmental temperature to drop during a roast (click here)
The steps involved with my PID project included:
Add a SPST mini switch on the control panel to select external PID control of the heater or standard internal Hottop control of the heater
Connect this switch to a high current DPDT relay that switches both legs of the heater circuit
Install both environmental temperature sensor and bean mass temperature sensor (type K thermocouples)
Run a short, heavy extension cord from the new high current relay to an external PID enclosure
Install two PID controllers in the external enclosure (one for control, one to display bean mass temps and provide high limit protection)
Install SSR in the external enclosure for switching the heater load
Tune the control PID
Heater Circuit Rewiring On most (all?) Hottops, T5 on the main PCB is the high side of the heater circuit, and stays at mains voltage. The basic Hottop turns the heater on/off by switching the low side of the circuit (T6). I followed this convention with my PID installation, and placed the new SSR in the external enclosure on the low (neutral) side of the circuit.
The Hottop heater wires were removed from T5 and T6 and repositioned on the common tabs of the new high current relay. New wires were run from the normally closed (NC) tabs on the new relay to T5 and T6 on the main PCB. And wires from the "umbilical" extension cord were connected to the NO tabs on the new relay. Wiring it this way prevents the PID from energizing the heater if for some reason the Hottop does not have power.
In the photo below, the short blue wires go from the common terminals on the new relay to T5 and T6 on the PCB. The red wires are the original heater wires, and are reconnected to the NC terminals on the new relay. The black and white wires connected to the NO terminals on the new relay are from the short extension umbilical cord going to the external PID enclosure.
The relay is mounted to the steel frame wall of the Hottop. I drilled and tapped a couple of holes to accept #6-32 machine screws. The relay coil is wired on one side directly to line voltage (black wire), and on the other (neutral) side through the new mini SPST switch on the control panel (white wire).
Bean Mass Temperature Sensor I used a SS 1/8" dia x 6" long grounded K thermocouple probe, mounted through the bean chute cover per Randy G. I chose this style (as opposed to a probe protruding through the rear wall of the roasting chamber) in order to minimize stem effects. I will probably further evaluate the protruding probe idea, but for now I am very pleased with the results of the Randy-style probe.
Environmental Sensor I placed a pair of exposed bead thermocouples around 1" above the Hottop's temperature sensor. Then I bent the probe wires up and over a little so that the beads ended up right at the edge of the bean chute, nearly at the top of the drum. I installed a pair of sensors here in order to be able to log temperatures from one probe while using the other probe as input to one of the PID controllers. My probe was built using a short 1/4" SS tube to penetrate the rear wall of the roasting chamber, and then epoxied into place using JB Weld (good to 500F).
I mounted two type K thermocouple jacks into the sloping plastic just below the Hottop's control panel. Labeled ENV1 and ENV2, the new environmental sensors are terminated at these jacks.
The pair of environmental sensors, viewed from inside the roasting chamber. Air temperature is measured at the tips of the probes. The wires were carefully bent up to be as close as possible to the top of the drum.
The Hottop's temperature sensor is seen at the right, immediately below the new probes, and to the right of the drive shaft for the drum. This stock NTC sensor seems to be heavily influenced by the temperature of the rear wall of the roasting chamber, which I think explains why it does not report the same temperatures as either an environmental air sensor or a bean mass sensor.
PID Enclosure Due to some careless purchasing, over the past couple of years I ended up owning 2 Watlow 1/16 DIN PID controllers that required low voltage power. In my PID kit business, I can really only make use of high voltage controllers. So these 2 orphan controllers had been taking up space for many months.
I also had acquired an enclosure along the way that was originally fitted with a couple of different controllers. Luckily, it already had the basic wiring, power switch, terminal blocks, thermocouple jacks, and controller panel cutouts needed for my roaster control project.
After adding a DC power supply from an old cell phone charger, a pass-through 110VAC outlet, and a short extension cord stub, the enclosure was operational. Everything in it came from either surplus, or other items already stocked in my shop. So it felt like it was free ;-)
PID controllers I used 2 PID controllers for a couple of reasons. First, it is very convenient to be able to view both environmental temperatures and bean mass temperatures at all times during the roast. Although this could be accomplished more accurately using a standalone thermometer, I viewed it as a great advantage to only have to fool with one additional thing-a-ma-bob on my roasting table (i.e., the enclosure).
Secondly, since the rewiring scheme used for this project bypasses the standard thermal fuse on the Hottop, I decided it would be a good idea to use the 2nd PID as a high limit controller. It is configured so that if the temperature of the bean mass exceeds a pre-programmed value (say 460F), it prevents the main PID from continuing to energize the heater (in the case of a sensor failure, for instance).
It is possible to choose one of the environmental sensors, or the bean mass sensor, as the control sensor for the main PID. More on that to come in a subsequent post.
Next, controlling roast temperatures with the PID...
Controlling Roast Temperatures with the PID Conventional wisdom says to use a ramp/soak PID controller for roast profiles. This makes perfect sense, but I had had only limited success with this approach for a Poppery control system. It seemed like there was always something non-standard happening that caused me to need to manually adjust the roast.
So for this project, I approached it as a kinda hybrid automatic/manual system. The automatic part is that the PID controller allows you to enter a target temperature and it will handle the power adjustments needed to quickly get the roaster there. The manual part is that I can easily adjust both the timing and temperature of the steps in the profile based on what I am seeing and hearing.
Tuning of the controller and establishing profile points is a work in progress at this writing. But it has become apparent that the roast can be effectively controlled using only 3 setpoints:
Drying phase setpoint
First crack setpoint
End of roast setpoint
Initially, I used the bean mass sensor as input to PID 1 (the roast controller). The setpoints were:
SP1 = 290F. This temperature is maintained until the beans have lost most, or all, of their "green." Only after this has happened, which is at different times for different beans, is the roast permitted to exceed 300F. (I adopted this based on comments I have read at various times by Jim Schulman, and I apologize if I have misrepresented that guidance).
SP2 = 410F. This SP is used to provide the gradual reduction in heat required as the beans become exothermic at first crack.
SP3 = 440F. Once first crack is up and rolling, this final setpoint is used for the ramp up to second crack and the end of the roast.
Because the Hottop does not have a particularly high wattage element, it ramps very smoothly from one setpoint to the next.
This control strategy was very effective at controlling the bean mass temperature profile. But, as the graph below shows, it resulted in wild variations in the environmental temperature. My goal was to have both a satisfactory bean temperature profile and a non-decreasing environmental temperature profile.
Controlling the roast using the bean mass temperature alone did not seem like it could satisfy my second requirement.
An Alternative Control Method Randy G had suggested to me early in this project that I might have better success by controlling the environmental temperature. After the problems I encountered with my first try at controlling using the bean mass temperature, his advice sounded much better.
So I switched the input to PID 1 from bean mass to environmental, re-tuned the P, I, and D parameters, and adopted new setpoints. After a few cycles of trial and adjustment, I have had success with the following setpoints:
SP1 = 360F. This results in a bean mass temp of around 290F, which fits my desire to keep temps below 300F until drying is complete.
SP2 = 450F. This setpoint takes the roast as quickly as the heater will allow to FC, then slowing down at the onset of exothermic behavior.
SP3 = 460F. Around 2 minutes after the first snaps of FC, changing to this final setpoint results in hitting SC approximately 2 minutes later. So total time from FC to SC is around 4 minutes using this profile.
The graph below shows that the bean mass profile is virtually identical to that resulting from controlling using the bean mass sensor. But the environmental temperature, now being directly controlled by the PID, looks much better, and does not have the dips that were evident on the previous graph.
While I reserve the right to discover I am wrong about this, at this writing I believe that using the environmental temperature as the input to PID 1 gives better results than using the bean mass temperature.
Phase II ? I am satisfied with the results of Phase I of this project. I have a single control box that displays both bean mass temperature and environmental temperature. It has a rudimentary high limit protection system. And, by manually adjusting the SP at 2 different times during the roast, a pretty nice-looking set of temperature profiles is easily attainable.
So what's next?
First, I elected to not fool around with overriding the Hottop's fan control. Since the unit I have modified with this project is a D model, I have no choice but to live with the built-in program for fan control. So far, this does not seem to be a fatal flaw, but it would be nice to be able to tweak the fan, too. Since PID 2 in my enclosure has a 0-10VDC (EDIT: turns out is is 4-20mA instead) output, using this to drive the fan through a Darlington pair (similar to CafeIKE's solution) would be straightforward. Worth the trouble? Not sure yet.
Next, I am familiar with some relatively inexpensive 1/16 DIN ramp/soak controllers that have attractive features, including serial interface and free software. Programming and modifying profiles is very straightforward. I am still uncertain, though, whether or not a pre-programmed profile will be a net benefit or hinderance. At a minimum, I will probably give one of these controllers a trial and see how it goes.
Finally, I hope to be able to design an inexpensive bean mass probe that can be mounted through the rear wall of the roasting chamber. I think it will involve a short SS sleeve and an exposed junction. I hope to isolate the t/c wire, thermally, from the SS sleeve and avoid stem effects, while still providing abrasion protection for the fiberglass insulation.
Disclaimer The folks at Hottop have gone to great lengths to make a product that is not an undue fire hazard. They built a lot of safeguards into the firmware.
This modification completely reverses all of their hard work, and may result in a terribly unsafe roaster. I cannot recommend that anyone else follow these steps until and unless you have independently confirmed this design.
It is probably a miracle that I only tripped the circuit breaker once during the development of this project.
I am very, very careful when I use my modified roaster (in the garage, away from all flammable things, and with a fire extinguisher within arm's reach). So please don't attempt a similar project using these posts as a go-by. That was not my intention, and no attempt was made to present this information in a how-to format.
The wiring inside the Hottop is a fairly complicated, and it would be very easy to make a mistake that resulted in fire or personal injury.
Nice job! I use the input air temp to control my PID popper and find it very consistant. I think consistancy is the problem using bean mass temps for the job. Have you posted this at homeroasters.org?
Symbols: = New Posts since your last visit = No New Posts since last visit = Newest post
Forum Rules: No profanity, illegal acts or personal attacks will be tolerated in these discussion boards. No commercial posting of any nature will be tolerated; only private sales by private individuals, in the "Buy and Sell" forum. No cross posting allowed - do not post your topic to more than one forum, nor repost a topic to the same forum. Who Can Read The Forum? Anyone can read posts in these discussion boards. Who Can Post New Topics? Any registered CoffeeGeek member can post new topics. Who Can Post Replies? Any registered CoffeeGeek member can post replies. Can Photos be posted? Anyone can post photos in their new topics or replies. Who can change or delete posts? Any CoffeeGeek member can edit their own posts. Only moderators can delete posts. Probationary Period: If you are a new signup for CoffeeGeek, you cannot promote, endorse, criticise or otherwise post an unsolicited endorsement for any company, product or service in your first five postings.
![JGG: P1000082-small.jpg](javascript:ImagePopup("/images/39924/P1000082-small.jpg", 500, 355, "JGG: P1000082-small.jpg");)
![JGG: P1000087-small.jpg](javascript:ImagePopup("/images/39925/P1000087-small.jpg", 471, 600, "JGG: P1000087-small.jpg");)
![JGG: P1000080-small.jpg](javascript:ImagePopup("/images/39926/P1000080-small.jpg", 500, 281, "JGG: P1000080-small.jpg");)
![JGG: P1000090-small.jpg](javascript:ImagePopup("/images/39927/P1000090-small.jpg", 600, 338, "JGG: P1000090-small.jpg");)
![JGG: PID-HT-trial01-small.jpg](javascript:ImagePopup("/images/39928/PID-HT-trial01-small.jpg", 600, 436, "JGG: PID-HT-trial01-small.jpg");)
![JGG: 20080726-bolivian-env-small.jpg](javascript:ImagePopup("/images/39929/20080726-bolivian-env-small.jpg", 600, 441, "JGG: 20080726-bolivian-env-small.jpg");)
![JGG: P1000088-small.jpg](javascript:ImagePopup("/images/39930/P1000088-small.jpg", 600, 359, "JGG: P1000088-small.jpg");)
![JGG: P1000089-small.jpg](javascript:ImagePopup("/images/39931/P1000089-small.jpg", 600, 468, "JGG: P1000089-small.jpg");)
![JGG: P1000084-small.jpg](javascript:ImagePopup("/images/39932/P1000084-small.jpg", 374, 600, "JGG: P1000084-small.jpg");)
= No New Posts since last visit 