Many users of Watlow temperature controllers can relate to the challenges I’ve encountered. We utilize a thermal chamber designed for temperature stress testing circuit boards, capable of heating up to 85 degrees Celsius, with options for even higher temperatures, and cooling down to -20 degrees Celsius, with higher limits available as well. However, achieving these temperature transitions must occur at specific rates, but that's a separate matter. Recently, we faced a significant issue: we lost all previously established parameters. Although I managed to restart the controller, the accompanying documentation left much to be desired—it was not just poor, it was severely lacking. For instance, I expect a Proportional (P) gain to be expressed as % output per degree of temperature error. However, with the Watlow, it appears to be configured as degrees per 1% output, which is quite confusing. It's also important to note that two sets of PID parameters can exist: one for heating and one for cooling. Currently, it seems like the cooling PID parameters are set to zero. If this is accurate, does that mean I'm using the same PID parameters for both heating and cooling? Can someone confirm my assumption? As for our P gain, it is currently set to 2.7. I interpret this to mean that we achieve 1% output for every 2.7 degrees of temperature error. Is that interpretation correct? Concerning the Integral (I) gain, we have it set to 1, but we’re uncertain whether this is measured in repeats per minute or minutes per repeat. The documentation hints at a setting that allows a switch between these modes, but we couldn't pin it down. Additionally, can someone clarify what is meant by a "repeat"? In my experience with temperature PID calculations for PLCs, the integrator usually relates to a time constant rather than a repeat interval. Lastly, we've considered replacing the Watlow temperature controller with our motion controller. While this would be excessive for our needs, at least we have comprehensive documentation and a full understanding of its operation.
I have experience with Watlow, but I may not be of much assistance. My application involved reading and writing PID parameters via Modbus; however, the PID loops had already been set up by other users. I completely agree that the manual is quite frustrating, and I’ve spent countless hours trying to navigate through it. If you find yourself really struggling, I recommend reaching out to Watlow support. While their response times can be slow, they might provide you with the assistance you need.
Much like Chopin, I utilized similar methods about a decade ago. My primary objective was to expose parameters to the user or the PLC, while the responsibility of fine-tuning them fell to another individual, likely employing the SOTP (Seat-Of-The-Pants) technique for loop tuning. From my experience, I worked with a Generic Ethernet module and spent considerable time consulting the manual to identify my desired parameters. This involved navigating through an array of SINT(3) values to set the Class, Instance, and Attribute needed to link the parameters to one of the 41 input words in the Generic Ethernet I/O table. To facilitate experimentation, I created a sandbox environment that allowed me to perform read or write operations on any combination of Class, Instance, and Attribute. I vaguely recall discovering certain useful parameters that were either omitted or inadequately explained in the manual. Regrettably, I no longer possess any spreadsheets or notes from that time—only the code, which lacks sufficient annotations related to the Watlow device. If you've established communication successfully, you're likely already aware of these nuances. Best of luck!
Peter Nachtwey highlighted an important consideration regarding PID (Proportional-Integral-Derivative) controllers: it's possible to configure two separate sets of PID parameters for heating and cooling applications. It seems that the cooling PID parameters may be inadvertently set to zero. If that's the case, I am presuming that the same PID parameters are applied for both heating and cooling. Can you confirm if my assumption is accurate? Additionally, our proportional gain (P gain) is currently set at 2.7. I interpret this to mean that the controller outputs 1% for every 2.7 degrees of error. Is this understanding correct? We've also configured the integral gain (I gain) to 1. However, there's some ambiguity regarding whether this is measured in repeats per minute or minutes per repeat. We noticed there might be a setting in the documentation that allows switching between the two modes, but we couldn't locate it. What exactly does “repeat” mean in this context? What process is being repeated? Typically, when calculating temperature PID for PLCs, the integrator is considered as a time constant rather than a repeated action. In our discussions about replacing the Watlow controllers with our motion control systems, we acknowledge it would be significant overkill; nevertheless, we appreciate the comprehensive documentation and familiarity with its operation. Based on our experiences with the Watlow 96 controllers, here are some insights into this topic: In dual-output heat/cool PID controllers, one set of tuning parameters is dedicated to each output. The units for PID parameters differ based on whether the controller operates in US or SI mode, which can be adjusted in the Setup/Global Menu under the parameter labeled "Unit." It's worth noting that Watlow uses Proportional Band in a somewhat unconventional manner, expressing it in engineering units rather than percentages (in US settings). This suggests it might represent the inverse of Proportional Gain, potentially as an input value calculated per 1%. Traditionally, "repeat" pertains to the repeated proportional action, which involves the interaction of P, I, and D actions. In relation to integral action, it is often referred to as "reset," especially since it functions similarly to resetting a proportional-only controller. The specified time value indicates how frequently the controller resets automatically, thereby continuously repeating the proportional action rather than being triggered solely at the reset time. For the SI setup, the Watlow 96's integral control indicates minutes per repeat, suggesting a relationship to Integral Gain, but the "per repeat" aspect raises questions about its interaction with proportional action, as it may not operate independently. It’s important to note that many low-cost single-loop controllers suffer from poor documentation and inconsistency in portraying PID computations. Manufacturers often assume that users will rely on built-in auto-tuning features rather than engaging in manual modeling and calculation of gains—and Watlow is no exception to this trend.
It has been six years, but I recall that Watlow offers a comprehensive programming package that can help you achieve your objectives with a single software routine (our controller supports this, but I can't confirm the compatibility with yours). To get started, simply outline the conditions for your test. For instance: 1. Gradually increase the temperature from room temperature to 85°C within 30 minutes. 2. Maintain the temperature at 85°C for one hour. 3. Decrease the temperature to -20°C over the course of 45 minutes. 4. Review your settings and save the program to your hard drive (we also back it up on our network). After that, utilize the software to set your parameters and transfer the program to your controller. Best regards, James.
To clarify, "2.7" represents a 5.4% change in output, while "repeat" indicates the process of soaking.
It sounds like you're facing quite a few frustrations with the Watlow controller—those parameter losses can be a major headache, especially when the documentation doesn’t provide clear guidance. Regarding your question about the PID settings, yes, it does sound like you’re using the same parameters for both heating and cooling if the cooling ones are set to zero, which definitely isn't ideal for system performance. As for the P gain interpretation, you’re correct that a setting of 2.7 means you'd get a 1% output change for every 2.7 degrees of error; that can lead to slower responses than desired. About the I gain, it’s indeed crucial to clarify what "repeat" means in your context—it usually relates to how often the integral action takes place, and knowing whether it's in minute intervals or something else can make a big difference. Your idea of switching to the motion controller for better documentation sounds tempting, but hopefully, you can find a workaround with the Watlow. Have you tried reaching out to their support or user forums? You might find some hidden tips from others who've dealt with similar issues!
It sounds like you're really navigating a tough situation with the Watlow controller! From what you've described, it does indeed seem like you might be unintentionally using the same PID parameters for both heating and cooling if the cooling settings are at zero. As for your interpretation of the P gain, it does align with the typical understanding; a setting of 2.7 indicates that you'd get a 1% output for every 2.7 degrees of error. Just be cautious, as it might mean slower responsiveness. Regarding the Integral gain, it’s definitely a bit ambiguous without clear documentation, but if it refers to repeats that usually points towards tuning frequency rather than a strict time constant. If it's still frustrating, perhaps reaching out to Watlow customer support could shed some light or help you get the documentation you deserve. And I wouldn’t blame you for considering alternatives; clear docs can make all the difference in managing critical testing setups!
✅ Work Order Management
✅ Asset Tracking
✅ Preventive Maintenance
✅ Inspection Report
We have received your information. We will share Schedule Demo details on your Mail Id.
Answer: Answer: In Watlow temperature controllers, the Proportional (P) gain is configured as degrees per 1% output, which can be confusing if you're used to seeing it as % output per degree of temperature error. For example, a P gain set to 2.7 means you achieve 1% output for every 2.7 degrees of temperature error.
Answer: Answer: Yes, Watlow temperature controllers can have separate PID parameters for heating and cooling. If the cooling PID parameters are set to zero, it may indicate that the same PID parameters are being used for both heating and cooling.
Answer: Answer: The Integral (I) gain in Watlow temperature controllers can be measured in repeats per minute or minutes per repeat. The specific mode can be switched, although the documentation might not be clear. A "repeat" generally refers to the time it takes for the integrative action to sum the error to a level that matches the proportional action, but it is often related to a time constant rather than a repeat interval.
Join hundreds of satisfied customers who have transformed their maintenance processes.
Sign up today and start optimizing your workflow.