Greetings Peter,
regarding your statement:
Ron, I am still baffled by the dead time and why there should be a deadtime in the Hot Rod system.
I’m not 100% sure that I’ve ever fully discussed this, but remember that I designed and built the Hotrods (all six of them) as student trainers. It might help to consider that there are TWO thermocouples installed on the Hotrod.
The first thermocouple is mounted right up against the heating element. I call this one the “base” thermocouple. The response of this thermocouple is pretty close to instantaneous. Specifically, once the drive to the heating element is changed, the temperature as sensed by the base thermocouple starts to change on the graph almost instantly.
On the other hand ...
The second thermocouple is mounted about 1 inch away from the heating element - out at the end of a steel screw. I call this one the “tip” thermocouple. The response of this thermocouple presents a fair amount of deadtime. Specifically, once the drive to the heating element is changed, the temperature as sensed by the tip thermocouple doesn’t change at all on the graph for something over 30 seconds. And incidentally, this “tip” model is the one that you've been working with.
In most of the classes I teach using the Hotrod, the students start out by setting up the air flow loop. This one has a fairly fast response so most of the graphs for this exercise are only about 1 minute long. With just a little bit of coaching, most students can successfully tune the air flow loop using just “trial and error” methods. But we go through a systematic tuning approach anyway ... we’re sure to need it later.
Next we tackle the heat loop with the Hotrod set for the “base” thermocouple. Here the graphs normally run about 20 minutes each. This particular loop (due to its reasonably fast response) oscillates at something around 1 minute per cycle. If the student fails to use a systematic approach, tuning this loop can be somewhat challenging. Specifically, if the student makes a wild stab at the tuning parameters ... and gets it wrong ... then it’s going to take about 3 to 5 minutes before the error becomes obvious.
And next we set up the Hotrod to use the “tip” thermocouple. And as you said:
Tuning a system with dead time is MUCH more difficult.
I can assure you that the students would wholeheartedly agree with you on that particular point. This particular loop (due to its inherent deadtime) oscillates at something around 3 to 4 minutes per cycle. And so if the student gets the tuning parameters wrong for this system, then it could take about 15 minutes or longer before the error becomes obvious.
And so the answer to the question “why would the Hotrod have deadtime?” has a simple answer. I specifically designed the deadtime into the system on purpose. For hands-on exercises, the Hotrod makes a nice little benchtop trainer ... which can offer the students some realistic challenges. Everyone agrees that deadtime should always be eliminated (or at least minimized) when designing a piece of equipment. But as we all know, you sometimes run into systems out in the field which DO have a fair amount of deadtime ... deadtime which can NOT be reduced. Naturally it’s a good idea for the students to have some amount of exposure to these types of systems before they run into them out in the real world.
As always, I’m looking forward to what you’ll come up with next. Let me know if I can do anything more at my end.
