residential \"PID\" thermostat?

R. Kalia

Now this is apparently a bad question, because a typical boiler needs on-off control, so P(roprotional) is not possible.

Still, even an on-off control can use D(ifferential) logic to prevent overshoot (a problem with many hydronic systems) by stopping the boiler if the temperature is rising fast, thus reaching and holding setpoint more precisely.

Modulating boilers can use an analog control signal and so full PID logic can be applied. Modulating boilers use internal logic that modulates in order to hold the water temperatue, but it is even smarter to modulate in order to hold the room air temperature, with constant water flow.

Anothert way to use a PID thermostat would be to control a zone valve rather than the boiler. I have seen valves on the web that are continuously variable rather than on/off and are controlled with a 0-10v analog input.

Process controllers (from e.g. Omega) will do what I am looking for, usually outputting both an analog signal and an on-off (relay), but they are meant to mount in DIN enclosures and not on walls. They also don't have setback periods and time displays and so forth.

Is there anything that is made for residential applications? Typical residential electronic thermostats don't do any kind of anticipation (D); although it is claimed that the cycle rate is the same thing as anticipation, this is not really true.

jw

Honeywell

See the T8600 series and my favorite, the PC8900 that will control heat, a/c, ventilation and humidity as well as dehumidify and reheat. Both are on off style. Both have PID logic and are adaptive to the system and stucture.

R. Kalia

nonsense

> Both have PID logic


Nonsense.

Ted_5

Look at HeatLink

I think they have the best zone control system out there!
It's called StatLink. The temp. stays + or - 1/2 of a degree from set point by modulating the flow with the zone valve. Add that together with a Viessmann Vitodens with outdoor reset and full modulation you have close to perfect control! Check it out.

Ted

BillW@honeywell

Adaptive Intelligent Recovery

The T8600 series and the PC8900 have what's called "adaptive intelligent recovery". Their software monitors the temperatures over the past 3 days, and start the heating/cooling system to begin recovery so that the conditioned space is at the comfort temperature when the periods change, for example, from "sleep" into "wake". Old electromechanical clocks and some of the low end electronic stats just start the equipment at a preset time, with no ramp-up or down, so you may be greeted with a cold or hot home when you awake or return home. No residential stat produces a modulating or floating output to drive a valve or damper motor.

If you want a variable response to changing conditions, there are several reset controls for hydronic heat. Ours are the AQ475(varies boiler run time)AQ675 (modulates a 3 way mixing valve) or the AQ775 that operates a 4 way valve or an injection pump. For forced air, try our Networked Zoning System.

The PC8900 also can control a humidifier using the "Humidicalc" software that functions in a similar manner. Both are +- 1 degree accurate.

ALH_3

tekmar

http://www.tekmarcontrols.com/

R. Kalia

Tekmar, Honeywell, Heatlink do not have PID temperature controllers. It is amazing to hear from someone at Honeywell, but Adaptive Intelligent Recovery and outdoor reset are not relevant to my question.

PID is relevant not just to recovery after setback, but to continuous precise temperature control. Honeywell claims that manually setting a cycle rate (based on steam=1, hot water=3 or forced air=6) is the same as anticipation, but of course it is not! Anticipation involves responding to the actual heating rate of the actual system in real time.

Tekmar's thermostats come closest in that, while they do use cycle rate, they have an 'auto' setting that claims to adjust the cycle rate to the response of the system while also minimizing short cycling. I don't know how well that works

The funny thing is, true differential control is better and easy to do. Process control people like Omega do it all the time; it just seems like Honeywell and others have never heard about it. (It does need a better temperature sensor than the cheap thermocouples generally used, because you need a more accurate temperature reading to get an useful measurement of the rate of change of the temperature with time.)

Ted_5

Did you even look a Heatlink?

or do you just assume it does not work. It may not be true PID but it is amazing how accurate it is. It does base some of the logic on cycles but modulates the valve on how the system is responding. The pulsing action of the HeatLink stats gives the acual room temp time to react by:

A] allowing the energy from the heated floor mass to be absorbed into the room in-between on/off cycles

B]monitoring room temps in-between on/off cycles,and then

C]initiating a corresponding cycle time based on veriance between actual room temp to desired set-point temp

because of the slow acting zone valves the position varies from total open or close to give it a modulating apperance.


Ted

jw

oops!

>No residential stat produces a modulating or floating output to drive a valve<

NONSENSE!

See the Ruud/ Rheem Contour controls!!!! Fully modulating input based on room temperature. (but not mfg by honeywell) Stat has a "v" terminal for variable input.

BillW@honeywell

We do \"know\" about PID...

as a matter of fact, our industrial process controls, aircraft controls, and some commercial building controls use this technology all the time. It just hasn't been adapted for residential use, and the controls I described are the closest I can come in a residential setting. I'm curious as to why you need or want such sophisticated controls for a residence. The cycle rate setting is a software-driven replacement for the electromechanical heat anticipators in the older generation of controls.

BillW@honeywell

A clarification...

Our resdential line doesn't have a PID thermostat. I didn't mean to imply that others did not.

ALH_3

tekmar

tekmar zone and house controls claim to have PID logic. (367, 368, 369, 370, 371)

R. Kalia

true

You are right, I was talking about their thermostats.

ALH_3

i misunderstood

The tekmar 507-512 thermostats do work very well though; much better than a traditional on-off thermostat. The display could be a little more user friendly.

R. Kalia

Yes, I noticed that unfriendly design from the documentation. To change the temperature temporarily you have to go through 4 steps, whereas with any normal thermostat you just push the up-arrow or down-arrow.

But as far as I can tell, it IS a traditional on-off thermostat, except for that auto-cycle-rate feature. It does not even have an usage counter, but it costs $125 or so.

R. Kalia

> The cycle rate setting is a software-driven

> replacement for the electromechanical heat

> anticipators in the older generation of controls.


But it is not the same. In old-fashioned anticipators, the longer the heat has been on, the hotter the resistor gets, so the thermostat stops calling for heat earlier. The Honeywells and all other electronic thermostats don't actually look at current conditions, they want the user to set the number of cycles per hour, based on the type of heat. This is a fixed parameter, not an adaptive behavior. Indeed, having multiple on/off cycles slows down the process of heating and reduces overshoot, but it is not as good as a resistive anticipator, because it is not responding to current conditions.

During recovery my Honeywell 3600 briefly stops calling for heat 2F before the setpoint---an intentional, preprogrammed behavior. That could be a good thing in preventing overshoot, but unfortunately it calls for heat again after 2 minutes and continues until the temp crosses the set point. (The hydronic system is of course too massive to level off within 2 minutes---that brief pause would work fine with forced air but is not useful for hydronic.)

PS Sorry about my testy comment, of course Honeywell knows about PID and I should not have suggested otherwise!

BillW@honeywell

The CT3600

is a retail model thermostat. They are quite different from the trade models despite a superficial cosmetic resemblence. They have a different and older version of recovery, and fewer features than the trade models like the T8600 & PC8900. Why were you suprised to hear from someone from Honeywell? Check out our website at http://hbctechlit.honeywell.com, and you can download any tech sheet you want, no charge & no account number required. For "consumer" type info try www.honeywell.com/yourhome.

R. Kalia

> is a retail model thermostat. They are quite

> different from the trade models despite a

> superficial cosmetic resemblence. They have a

> different and older version of recovery, and

> fewer features than the trade models like the

> T8600 & PC8900. Why were you suprised to hear

> from someone from Honeywell? Check out our

> website at http://hbctechlit.honeywell.com, and

> you can download any tech sheet you want, no

> charge & no account number required. For

> "consumer" type info try

> www.honeywell.com/yourhome.

bb

Recovery

Bill;

Sorry, but adaptive recovery has nothing to do with PI or PID control. It only brings on the equipment to bring the room to the required temperature at the required time.

bb

BillW@honeywell

I understand that...

I was trying to offer as close a product as I have to help this poster. P+I is nothing new, it's been used in commercial electronic and pneumatic controls for years. PID is more suited for critical applications in process controls, avionics and commercial building controls.