Honeywell 8000 series setback being stupid

ChrisJ

Ok I'm really getting annoyed at my VP8110 firing up the boiler and then just as it starts steaming shutting it off due to my setback.  Now yea, I know why its doing it but its not difficult to program something to see if its going to be doing a setback in the next 15-20 minutes and if so either not turning on, or extending the setback time some. 

The amount of fuel this wastes every time it does it can't possibly qualify is being "green"



Now I know I won't be changing the thermostats firmware, or Honeywell's minds.   But what I would like to know is, is there a programmable T-stat out there, that would work the way I want?  Does the Nest have enough intelligence to not do something this dumb?



I do a 2 degree setback at night because I like to sleep in a cool house, also because I like the radiators to be cookin when I wake up.  My system will even do a 9 or 10 degree recovery without tripping on pressure most of the time but thats just too cold for me.



On a seperate note, the 8110's intelligent recovery, does this only effect turn on time, or does it also effect turn off time to help with overshoot?  If its only turn on time I want to disable it because I'm tired of waking up to a hot room 15-30 minutes before my alarm goes off.



Thank you for your time and help!

Mark N

Thermostat

How can a tstat be stupid? Does it start the setback at the programmed time? Is the house back up to temp at the programmed time? Does it maintain the programmed temp when not it setback? I doubt that the boiler is running every time the tstat is programmed to start setback. If it is maybe try changing your setback time to earlier or slightly later. Same with the recovery if you find the intelligent recovery starts to early try starting 15 minutes later and see if that is better. Maybe you can try what I do, don't use setback and leave the tstat set to one temp and don't worry about it all winter.

nicholas bonham-carter

Disappointment with setback

Yes your setback may be costing more fuel than keeping a constant temperature. These systems were designed to supply a constant heat, and not a "cooking heat".

There may be other reasons for the seeming short-cycling you are experiencing.

Try shorting the terminals on the boiler which are connected to the thermostat to simulate a Clall for heat, and see if the same firing behaviour continues. If it does, try to increase your main (not rad) venting. In addition to that, check the pressure with a good low-pressure gauge, and see if you can keep it down below 8 ounces, for comfort and economy.--NBC

ChrisJ

pressure and stupid

Mark, a thermostat can be stupid by not taking things like doing a setback before the system will produce heat and wasting 5-10 minutes worth of burner time for no reason.  There is no reason a computer cannot take these things into consideration if told to do so.





Nicholas,  system is doing a recovery now and is about to shutoff.  I went down to check my pressure and my 3.0PSI gauge is on 0 bouncing a tad, only noticable if you look at it just right in the light. This is why I mentioned being able to do a 9-10 degree recovery without tripping on pressure.     The system just shutoff and after doing the 2 degree recovery it didn't even produce enough steam to make a single Hoffman 1A click, steam never completely filled the rads.



You are right, these systems are designed to maintain temperature and I understand that.  I am happy with the system, real happy, just not impressed with the $100+ thermostat I bought for it, I think they can do better. 

Jean-David Beyer

these systems are designed to maintain temperature and I understand that.

I do not think fancy setback thermostats are the answer to everything.

I have a really fancy Honeywell, now discontinued, that can have four different temperatures each day, and no two days need have the same temperatures, so I can have 28 different temperatures each week. It also has a so-called smart recovery where it can see that I need a certain amount of recovery, so it can start a heat call up to an hour or two in advance, so the temperature will be right at the desired time. I can also set it to cycle once an hour up to perhaps 10 times an hour.



However, it is useless, and a $20 simple thermostat would be just as good. And this even without worrying about whether it takes more money to recover from a setback than is saved during the setback (I do not happen to believe that, but it seems to be a controversial point. It may be true for steam.)



The reason it is useless for me is that I have radiant heat in a slab. So when the setback starts, it takes, perhaps, 8 hours to get a little setback, and 24 hours to get more. I involuntarily got 11 degrees of setback when the power was off for about a week due to Sandy.  Since I have very tight reset curves, providing only enough heat to maintain temperature, I guess it would have taken quite a few days to recover. As it was, I defeated my setbacks and the temperature came back up in about two days.



When I got that thermostat, I experimented with it. To get the house cool at night, I had to start the setback around noon. Even with the intelligent recovery, it would never start recovery more than two hours in advance, so I had to start recovery about 10 PM if I wanted it warm again in the morning. The thermostat would never figure it out to give more than two hours of recovery, and usually, it gave only about one hour.



Now if I dialed it up to do the setback during the day and the recovery during the night, it would be a little better, but that only if each day was pretty much like the other, and that is not the case around here. I now have the thermostat set to do no setback unless I go for vacation for a week or more. Then I set it back about 7F and set it to turn on 24 hours before I get back.



If my boiler allowed me to connect my computer to it, and run the thermostat input and to diddle the reset curves (and the manufacturer wisely does not allow this), I could have a good time fooling around. But it would take several years to figure out decent algorithms, and even then, they might not be stable. An unstable control system is not a good thing.



So with my system, I do no setback in the radiant slab zone, and only one degree setback in a baseboard zone. And going fom a $100 thermostat to a $3000 one would make little difference, other than dry-cleaning my wallet.

Jamie Hall

Computers are stupid

inherently stupid.  They will do exactly what they are programmed to do -- that is, what the person setting them up tells them.  Nothing else.



This can sometimes be very annoying...



In this situation, there are two very different sets of routines operating.  The first, and primary, one is that the thermostat is trying to hold whatever temperature it is set to.  Thus, if the temperature drops slightly, it will fire up the system to bring it back to the set point.  On the other hand, that set point can be adjusted to any of several values atuomatically, based on a clock time.



What is not programmed in is the ability to "look ahead" and determine, within certain limits which would also have to be programmed in, that the set point will change down so soon that there is no point in firing up the system right now.  It is, in fact, something of a wonder that the thermostat is able to look ahead and see that the set point will increase at a certain future time, and to reach that future set point it is necessary to start the boiler now -- and adjust that anticipation time based on previous attempts.  The ability to look ahead at a dropping set point and determine that there is no need to fire up the system could be programmed in in the same way -- but that would just change the conditions under which you would get a short fire; it wouldn't eliminate them.



Programming control systems -- even a simple thermostat -- to be able to anticipate future conditions and be stable is very difficult (and expensive).



I might add that with these learning thermostats, a sure way to really foul them up is to change something such as the amount of set back.  It will take as much as several weeks of steady outdoor conditions before it will really figure out how much anticipation it needs -- and how often do you get that?

BobC

They don't live with them

I understand what you mean by finding the boiler making steam just before it's about to go into setback. Luckily I sit right over the boiler and can hear the gas valve clunk so I lumber over to the thermostat to kill that cycle. i also agree that it would be easy to put something in the code that could look at the system intelligently and stop this from happening. The problem is these things are designed for forced hot air systems and then they add a few things to allow changes to the CPH setting. I'll bet the designers don't live with high mass slow response heating systems so it just doesn't occur to them that they should make allowances.



For your other problem, you might try a staged recovery from setback. I come up 3 degrees initially and then come up the other 3 degrees later on and I find that works for me. My thermostat is the RTH6450 from Honeywell.



Bob

Mark N

Thermostat

Chris just because the tstat doesn't work the way you think it should doesn't make it stupid

it makes it irritating. This can be looked at 2 ways. Either going into setback wasted the 5 to 10 minutes worth of fuel, or it saved the other 10 to 15 minutes worth that it would have used to complete the cycle that you didn't want because you chose that time to go into setback. It can get expensive experimenting with different types of thermostats. Why not get the Nest and let us know if it operates the way you think it should.

ChrisJ

what is stupid

Mark, with all due respect running a burner for 10 minutes to do absolutely nothing is stupid and wasteful.  Running it an additional 5-10 minutes and actually getting heat from it is not.  You actually get something for your money rather than simply burning it. Sure  MAYBE some of the heat leaks up from the basement but not much.  You also left out, or did not see that I said the thermostat should just leave the burner off rather than firing it up in the first place.  Then, when it fires it up to maintain whatever the setback temp is at a later time, it would not short cycle.

This is not a matter of opinion.  Others have said IF Honeywell wanted to program the device to operate this way they could have.  It looks into the future to do recoveries, it could do the same with setbacks.  Either way, how the thermostat operates is stupid, why?  Because wasting fuel when a simple modification to the firmware would stop it, is stupid. 



I would buy a NEST and try it, however I don't have money to waste on a $200+ thermostat to "test" it only to find out, that like this one its designed around forced hot air.  I would rather spend the money building my own except I have yet to learn how to program more than basic stuff.  I started learning C# a few months ago but ran into some problems which of course the programing got thrown off to the side.

ChrisJ

building a tstat

Just spoke with some of the engineers I work with and I will be attempting to learn how to program in C.  This will make it so I can build my own thermostat using a PIC24 processor.



This is my plan, but who knows if it will happen.  I'm going to try.

Mark N

Good Luck

Now your thinking. Build a tstat that works the way you want it to. Honeywell builds their stats to work with a wide variety of heating systems. They try to compensate for this with the cph feature. I looked over the Nest website and it looks like their tstat would work best with forced hot air. Probably the best stat ever designed for steam was the old Honeywell T87. Once you had the anticipator set up correctly you were good to go. I had 1 of these for years. The only problem with it was every once in a while someone would bump into it a jack the temp up or down.

ChrisJ

T87

I gave my mercury T87 to Crash.  

I don't regret getting rid of it, I like the 8110 and even my neighbors 5000 series as a normal thermostat.  They both maintain a pretty good temperature.

The T87 was pretty cool when I had it adjusted so my radiators never cooled off completely even on a mild day, I just don't know what it did for my gas bill.

The system wouldn't heat the first cycle, but by the 2nd  cycle it would start heating and would maintain a very, constant temperature.  Yea, I know, it was "stupid" to have it set that way.



 I guess in a few months we'll see if I can get time to learn C.  It seems like it would be really cheap to build what I want hardware wise, just no idea how many hours, or days, or months worth of programming it takes.

conversiontime

vote not intelligent

I turned off the intelligent option as I found it pointless. E.g. I program the boiler to fire at 545AM as we run hot water from boiler coil in winter season and from 545 to 630 want plenty of hot water for showers, kitchen etc. Yet the "intelligent" honeywell started firing the boiler at 515 so it would have it nice and toasty for me, the **** who told it 545 when obviously I really meant 515 right? So at 545 house is warm but boiler is off, showers start and boiler starts firing up and down to feed the coil--exactly what I wanted to avoid.



The algorithm for 1+- maintain temp is annoying as well. If you want a therm that projects false reads based on where it "thinks" the room temp will/should be in the future, then honeywell is your choice. In fact I generally assume whatever temp it shows for room temp is simply a projection off by 2-3 degrees.  I just installed mine this fall and even at discounted price I would not buy it again but for now I am using it.

ChrisJ

Wrong temp

Same exact thing I have noticed, it only displays the real room temperature when its convenient, otherwise it often lies to you.  Say set it to 70F from 68, it will claim the room is 70F long before it is according to all other thermometers in the area.  In fact, I have watched it jump up 3 degrees before I even produced steam.

conversiontime , do you know of a good thermostat that does what I want?

Mark N

FAQ

Have either of you read the FAQ on the Honeywell web site for the Visionpro? In the explain how the Visionpro displays the temp and how the adaptive recovery works.

ChrisJ

Recovery

Yes I have,  and I still am not happy with it.

conversiontime

honeywell algorithm

I do understand the idea behind the "intelligence" and have read all documentation but to me if fails. Basically honeywell substitutes accurate current temps in a swing format for "fuzzy" baseline temp that is kept deliberately static and inaccurate during fire/setback so one can claim the temp stays within a single degree. All packaged in some geekspeak "algorithm" that basically masks swing moves with false readings. just my 2

Jamie Hall

Gentlemen, gentlemen...

May I refer back to my original comment?  Computers will do exactly and precisely what they are programmed to do.  Every time.  That may not be, however, what you might want them to do -- in which case the answer is, as our intrepid ChrisJ is suggesting, to program your own system.  The answer is not to disparage the device; it's doing what it's meant to do (one might compare that to being disappointed that yoiur Chevy pickup didn't win at Daytona this year and saying that the pickup is not good because of that).



For ChrisJ, however... I do hope that you will keep us all posted on your efforts to create your own temperature control device (it won't be a thermostat: the purpose of a thermostat is, as the name says, to maintain a constant temperature with a very small, but defined, dead band).  May I suggest that the first step should be to define exactly what it is you want it to do, and second to define equally exactly what inputs you have available to make it do it?  After that, the next step will be to define -- again, very exactly -- the complete behaviour of the system you are trying to control (the combination of your boiler, radiation, house, and weather environment).  Then carry on...