VisionPro Thermostat Problem

Nick_37

Based on the feedback on this site I installed a new VisionPro thermostat to replace my old mercury thermostat so that we could use setback features. The issue is that it won't shut off the call for heat once it has reached the operating temperature.



I have been through all settings and double checked that it is set to 1CPH for my steam system. I set it up for heat only. I can turn the temperature up and down. If I manually turn it down it will shut off the furnace. If I turn it up it then it just keeps calling for heat until all the radiators max out and the furnace eventually shuts off . It does this regardless of the higher temperature that I set it to i.e. it doesn't care if I turn it up 1 degree of 5 degrees it will just keep heating until the radiators are heated all the way across and the valves are hissing like mad.



I spoke to Honeywell and they think it is probably defective but I wanted to check here first before I send it back to pexsupply.

jpf321

check out this thread

Check out the thread here... perhaps you guys can compare notes ..

http://www.heatinghelp.com/forum-thread/128750/to-hot-when-it-gets-really-cold



What is the exact model # of your t-stat .. sometimes the model # is beneath the batteries .. perhaps the manual is online here if it is a VisionPro 8000 or IAQ:

http://www.forwardthinking.honeywell.com/products/thermostats/thermostat_products.html



It helps the folks here to know exactly which t-stat you are referring to.

lutorm

1CPH?

Correct me if I'm wrong, but if you set it to 1CPH, then won't it at most turn on and off the heat once per hour. That would indeed lead to larger overshoots in temp, wouldn't it? Are you sure it will keep the heat on for >1h even after reaching the set point?



I was thinking of getting this thermostat, too, and I remember reading something on the faq about how it doesn't just look at air temp but also at the radiation temp, so if the air is warm but the stuff in the room hasn't heated up yet then it would keep going for a while.



Here it is:



Why does the thermostat temperature not match my temperature thermometer in my home?



Answer: Honeywell digital thermostats are designed to display the room temperature in a way similar to how people sense temperature. This means the thermostat

takes into account not just the ambient air temperature but also the

radiant temperature of objects in the room (i.e., wall and furniture).

Room thermometers often only read ambient air temperature so they will

not match the thermostat reading. Also, Honeywell thermostats do round

in the display to the nearest whole number (half number in Celsius) and

to the temperature setting. For example, if you have the thermostat set

to 72 degrees the system will turn on and off but the inside reading on

the thermostat never changes from 72. The actual temperature did fall

to 71 or up to 73 and that is what turned on the heating or cooling but

the thermostat display will stay at 70 to avoid jumping up and down

constantly.

Mike Kusiak_2

Not necessarily

Take a look at this thread:



http://www.heatinghelp.com/forum-thread/127201/How-does-a-digital-thermostat-really-work-inside



The last few postings go into a lot of detail on how the cycle rate program operates on a digital thermostat.



Basically, the thermostat measures the room temp and compares it to the setpoint. Assume the room temp drops to the lower threshold ( setpoint minus differential ) for a call for heat and fires the boiler. When the room reaches some predetermined temp approaching setpoint, it initiates the timing cycle. For 1 CPH the timing interval would be 30 minutes on, 30 minutes off. So the thermostat stops the call for heat and then checks again in 30 minutes to see if setpoint has been reached. If it has, it waits to fire the boiler again until the low temp call for heat threshold is reached. If setpoint has not been reached, it fires the boiler again and repeats the cycle. By stopping firing before setpoint is reached, and then waiting half the cycle time before again checking, the thermostat prevents overshoot in a similar manner as the anticipator did on analog thermostats.

FJL

But . . .

There are mornings when my boiler runs for between 45 and 60 mins to recover from a setback.  Isn't that contrary to what you just wrote?

David Nadle

Not quite...

The VisionPro on 1CPH will certainly heat for more than 30 minutes to reach a distant setpoint. I can tell you first hand it will heat for more than an hour if it has to. I can also tell you first hand that if you have good venting and radiators that fill quickly you can set 2 CPH with no issues and possibly better control of temperature. My understanding of the CPH setting is it's the maximum number of "on" events per hour but does not limit the run time.



The VisionPro uses a proportional + integral (P+I) feedback algorithm. Part of the control signal is proportional to the current temperature error (difference), and part is proportional to the integral of the error. Using the integral of the error helps the thermostat come on earlier, avoiding the droop common to old style thermostats, and shut off earlier, avoiding overshoot.

jpf321

everything there is to know...

a great article about CPH and anticipators.. http://www.carolfey.com/articles/anticipation.html



i believe that CPH is a lockout period .. and i do not think that CPH cancels an existing call for heat



. for example .. you come home from a long weekend and your house is setback to 50dF .. you crank up the stat to 70dF .. and the stat continues the call for heat until it reaches a balmy 70dF... at which point the CPH says to the stat .. i don't care how many windows and doors they just opened it maybe 32dF in here now but .. i'm not going to ask that poor tired boiler for more heat until X minutes have passed... and sure enough after Xmins .. and a 32dF temp inside and the CPH says .. OK .. i think the boiler has had enough of a break .. let's go ask him to warm us up again... (and I really wish they would close some windows .. what are they thinking there's a spanish flu pandemic going on?)



anyway .. that's what it looks like in my head.

FJL

1 CPH

Ok.  I guess I never fully understood what 1 CPH meant.  The t-stat calls for heat and the boiler will turn on and run as long as it needs to satisfy the t-stat.  After that, the temp can fall as fast as it wants, but the t-stat will wait 30 minutes before calling for heat. 

jpf321

yep i believe so ...

the Xmin clock only starts ticking after setpoint is achieved.

there is never a clock that sees how long i've been asking for heat,

he might never stop asking and therefor the CPH clock never starts ticking.



according to right side of pg 3 in this document: http://customer.honeywell.com/Techlit/Pdf/69-0000s/69-1800.pdf ..



2CPH = 30min lockout - steam/gravity

3CPH = 20min lockout - HW/high-efficency furnace

4CPH = 15min lockout - gas/oil (presumably furnace forced air)

5CPH = 12min lockout - gas/oil alternate setting

6CPH = 10min lockout - electric



it has to do with how quickly a particular system type is expected to deliver and distribute it's heat.

Mike Kusiak_2

No

If you are coming out of setback, the room temperature never gets close enough to setpoint to start the timing cycle. If it takes an hour or more to reach the cycling threshold, only then will it initiate the cycling. What I was trying to say was that the thermostat senses when you are almost at setpoint and then ends tha call for heat slightly early, to compensate fot the thermal mass of the system. Then it checks again in 30 minutes to see if setpoint was indeed achieved.

FJL

Thanks . . .

Now I understand better. 

Mike Kusiak_2

Overshoot

Going by your example, if the thermostat indeed ran the boiler until the 70 degree setpoint was reached, the mass of the system would continue adding heat even after the thermostat was satisfied and overshoot the setpoint. But if you ended the call for heat at say 68 degrees, and then checked 30 minutes later to see if you really achieved 70 then you could overcome the overshoot problem. Thats what I was really trying to say in my previous posting. In other words, the cycling would only start when you got to 68, just a little shy of the setpoint of 70. Until you achieved 68, the boiler would run continuously without cycling.

jpf321

i see your point ...

I see your point .. I'm not really sure how overshoot is compensated for (if at all) .. but I'm pretty sure that CPH does not cancel or interrupt a call for heat. If it were me and that kept occurring (or more specifically, I kept getting too hot on calls for heat, I would readjust my stat to a lower temp or my differential perhaps. Maybe that's why you need a "steam rated" stat so that you can set it for steam and it can programmatically adjust for anticipation .. there is also Honeywell's Intelligent Recovery tech that learns the way your system outputs heat.

David Nadle

CPH is not CPH

I'm not actually sure that 1 CPH really means one cycle per hour by contract. I believe that its an "ideal" number. When you change the CPH setting on a VisionPro I think you change the coefficients of the PI algorithm, and the end result is more or less frequent cycles.

FJL

1 CPH

Until reading this thread, I assumed that 1 CPH meant that the t-stat won't call for heat more than once within "the hour."  And that has been my experience with my VisionPro t-stat.  It never calls for heat more than once within "the hour."  This means that if the t-stat calls for heat at 7:05 a.m., the earliest it will call for heat again is 8:00 a.m.  And I've noticed many times, on very cold days, that when the boiler is running a lot to maintain the temperature, the t-stat will call for heat at the top of each hour.  

jpf321

generally how long

generally, how long is your heating cycle (from call to satisfied stat)? about 30mins? what CPH setting are you using? 

jpf321

more than anyone should know.....

here is an engineering document by Honeywell that goes into 518pages of details regarding their automatic controls. this is certainly more than anyone really should know about thermostats http://customer.honeywell.com/TechLit/pdf/77-0000s/77-E1100.pdf

especially have a look PDF pg#'s --26-31



I also found this chart which may be useful .. but I don't translate French very well ..

http://customer.honeywell.com/techlit/images/gif/MF4414.gif

David Nadle

That graph...

...is in the back of almost every Honeywell thermostat user guide. It's not much better in English. That controls book however, is a real gem. Where did you find that? As you can see, the section on PI control makes reference to "reset time." The intergral feedback coefficient is Ki/T, so changing the reset time changes the coefficient as I said above. I think CPH is the reciprocal of reset time.



You seem to be a master Googler. I have seen a Honeywell white paper explaining the principle behind CPH vs swing setting but of course now I cannot find the link anywhere.





 

FJL

1 CPH

My t-stat is set for 1 CPH.  Needless to say, the heating cycle varies depending on how cold and windy it is.  But for the past few days, the boiler has been running approximately 25-30 minutes in the late afternoon and evening to maintain a 68 degree temperature in the apt with the sensor.  As I've said, I've never known the t-stat to call for heat more than once within the hour.  

Mike Kusiak_2

Very Interesting

The explanation of two position control beginning on page 27 really describes the situation in a home heating system where the only mode of control is the on-off function of the burner.



The paragraphs on time proportioning seem to describe the mode of operation which Easily Found observes, where with the 1CPH cycle rate the burner runs only once per hour, with a variable duty cycle dependent on heat load. I am a little confused by the term "heat load" and how it is determined in the home heating application. Fig.25 seems to indicate heat load to be the difference in temp between indoors and outdoors as would be the case in an outdoor reset situation with an outdoor sensor. Without the outdoor sensor, the heat load seems a little harder to determine. The only thing I can think of is that heat load is calculated from the variation in temperature between the setpoint and actual room temp. If there is a large difference, as in a setback recovery, then the on duty cycle is 100%, while if there is only a small variation, the duty cycle is reduced to predictably decrease the firing time to prevent overshoot.



I am not quite sure how PI control can be applied to on-off control, as by definition, the control is non-proportional in nature having only two states. Perhaps the use of duty cycle modulation as described above can provide some approximation of proportionality to to the control algorithm.



Thanks for posting the Honeywell paper as it seem to answer a lot of questions!

jpf321

Honeywell Patent c1995

Perhaps having a read through one of their patent apps will help to further clarify things .. http://www.freepatentsonline.com/5467920.pdf especially beginning around pg 4, col 2 line 28

jpf321

this seems pretty interesting ...

Here is a letter from Honeywell to EPA regarding the EPA proposed requirements for EnergyStar programmable thermostats .. Please Note that Firefox Browser didn't like taking me to this site .. but it is a legitimate site and a very interesting letter expressing "Honeywell's Point of View" https://energystarproduct.com/ia/partners/prod_development/revisions/downloads/thermostats/Honeywell-Comments.pdf



Here is the full list of documents at the EPA site regarding EnergyStar Programmable T-Stats .. including their original Draft and the comments from all other "partners"

http://www.energystar.gov/index.cfm?c=revisions.thermostats_spec



And as an added value .. we now have an email address for John T., Honeywell's Vice President & General Manager, Americas Residential Home Products & Systems in case we have any issues or questions that need to be escalated :-) .. Like crappy Vaporstats...hmmm.

jpf321

NEMA Standard

The Oct 2009 NEMA Standard for Residential Controls- Electrical Wall-Mounted Room Thermostats document can be downloaded for FREE here .. but you need to create a user account with a valid email address prior to downloading: http://www.nema.org/stds/dc3.cfm#download



In that document they have a few charts discussing droop and cycles.

lutorm

PI and on-off

You can use proportional controllers on on-off equipment, the proportional signal just drives a PWM (pulse width modulation) output. For this to work, the pulse frequency of course has to be shorter than the response time of the system and the measuring equipment.



For example, this strategy is used to control most temperature-controlled computer fans, with a switching frequency of tens of Hertz. Effectively, that makes the fans turn at a constant speed.



You could imagine doing this on a boiler by switching the fuel valve faster than the response time of the burner, so that it would effectively be controlled proportionally. Of course, you'd need something that could be switched that fast, something like a car fuel injector, to regulate the fuel flow.

lutorm

Interesting

That was an interesting read. In general, their comments seemed reasonable, but I thought it was disingenous to say "electric heating converts 100% of the energy to heat". That's true, but to make the electricity, you need to use at least 1/3 more energy, so electricity is not very energy efficient overall.



Unless of course the electricity is used to run a heat pump with a good reservoir (like a ground source heat pump).

oscondoles

back to the original question

It sounds like there is an issue with the boiler too.  Even if the thermostat keeps calling for heat due to a defect or other reason, the boiler should cycle off on pressure instead of running to the point of mass hissing out of the vents.  Also sounds like the vents should never get to the point of loud hissing - perhaps there is also a main line venting issue.  Finally, I believe the vents should close when steam hits them.