complex requirements!

Those hall control requirements are a bit complex to fill… and it’s become clear with just how fast the hall now heats up, we have to think about user controls that keep it from overheating, whereas before that simply wasn’t often a problem!    I’m not quite sure what we’ll do this time, but I’m working on it.

Here’s a rough list of requirements. It’s a bit hard to decide what’s “must” and what’s “should”, so that’s a bit fluid.

The controls must:

  • allow the user to turn the heating up or down temporarily, preferably until the next program instruction (so that as long as it’s programmed for every room hire separately, no group affects any other)
  • allow the user to turn the heating on from a “cold” start (hall at, say, 10C), preferably for a short time (say, an hour) .  It’ll be annoying to hit the switch every hour if a group was forgotten in the heating schedule, but the system is expensive enough to run that annoyance is warranted; it doesn’t happen often.
  • not permit the user to change the program or make any changes that last longer than a few hours/until the next program element
  • not allow the user to turn up the heating beyond around 22C (REASON:  (a) it’s expensive! and (b) lots of users just dial up to max thinking that will make the space heat faster)
  • have six or more  on/off set points – eight is even better
  • have some form of optimisation that will mean the system comes on for less time in warm weather

The last two are for economy – the system is fast enough heating up the space that they will definitely save money the more they can keep it off.

The  controls should:

  • allow programs that  specify different temperatures for different times
  • not  require batteries
  • present a simple and obvious  interface  – for instance, for the end user, maybe up and down buttons or a dial for variation, with any length of time for the variation very obvious, and push to turn on from off (if necessary).
  • look like heating controls (and not, say, light switches or doorbells)
  • be designed for accessibility (large fonts, tactile buttons)
  • show whether the heating is on or off and what the target temperature is (so that they don’t fiddle with it if it’s going to reach temperature anyway – useful for when someone has left a door open)
  • do summer/winter adjustment automatically
  • allow the user to heat for a very short time if they wish (for cleaners and other people passing through at random times, because otherwise the temptation to heat for a full hour at a sample cost of £1.70 will be too great).
  • preferably, be programmed from some space management can get to even when the space has been hired

So, there are four kinds of “bits” in the pre-internet world, all of which have some properties that fit some of the specification.  They are programmers; room thermostats; programmable room thermostats;  and countdown timers.

Programmers and room thermostats work together – the programmer says what times the heating should go on and off, and the room thermostat says what temperature it should aim to maintain when the heating is on.  Some room stats come with top and bottom limits that can be set at installation – in configurable  firmware or old-fashioned break-off tabs.  Any changes made to the temperature setting on a room stat are “permanent”, until the next time someone changes it.  Programmers usually allow the user to turn the heating on or off temporarily, usually by a one hour “boost” setting or by just advancing  to the next program time early.

Programmable room thermostats combine these two things, but with a few control finesses.  Many of them let the programmer say, for instance, “make the space 18C by 9 am, 15C by noon, and 19C by 5 pm, but you can decide when to actually start”.   It can be hard to know exactly how some of them decide.  Sometimes there are configurable settings to specify some kind of heating curve (with or without an external temperature sensor to know what the weather’s like), and sometimes it learns from experience.   They often let the user adjust the target temperature up or down temporarily, at its simplest, using up and down arrows.  Sometimes this is until the next program step and sometimes it’s timed.  A few models with complex user interfaces have both.

Countdown timers let something be brought on for a while.  Sometimes it’s just a push button for a set time, sometimes it’s configurable using switches on the back, and sometimes the push button cycles through options visibly on the facade.  Occasionally the amount of time is set using a dial.

Locking can be done either by the device – using a pincode or a special combination of keys that most people don’t know about – or by using a lockable box mounted around it, usually called a “thermostat guard”.

So, looking at the requirements,  wanting different temperatures at different times suggests a programmable room thermostat.   The problem is, then we can’t lock up the device completely, as the end user controls are on the same facade as what the property managers use.  When they do lock, it’s all or nothing.  The Warmworld Dataterm, which this site used to use, has three levels of lock specifically to keep user and manager functions separate.  That’s because it’s aimed at the social housing market.  There used to be another partially lockable model – I think it was a Honeywell – but I can’t find one now.  The only thing I’ve seen that comes close is a Graesslin model (“Feeling”) that has a sliding cover to at least hide the advanced features, with users able to satisfy their needs (in this case, to turn the heating on, up or down until the next program setting)  without tempting them to slide the cover.  It has settable top and bottom temperature limits, while  the Dataterm’s high limit is constant at 26C, where most older buildings would top out through boiler fatigue anyway.  Unfortunately, since the management have said they find the Dataterm a bit complicated to  unlock and get to the program, I don’t think they’ll like the Graesslin either.  It is, shall we say, feature-rich :-).  On top of that,  for most design decisions they’ve made, they’ve taken a different path from the crowd.  That can be a good thing for something where the lock is psychological – it discourages fiddling – but it also makes it harder to learn.

On the other hand,  unless we relax the requirement for different programmed temperatures at different times, we have to use a programmable room stat.  A countdown timer could be wired to let a user bring the system on when it’s off, but when there’s already heat demand, there’s no way to add an external device that will override the temperature setting on the programmable room stat.  So what if we remove the requirement for different temperatures at different program times?  Then we can use a programmer instead, with a countdown timer and a room stat  – but if a user changes the temperature on the room stat “in the normal way” (barring some advanced feature that is sure to be less obvious),  it’ll be changed for everyone. And then there’s no way to turn the system off for the duration of their group, just down as far as we let them.

Constraint satisfaction.  It’s a bear.

That might push us into the age of the internet, which also means wireless, of no particular benefit in this case.     I’m not sure I’m a fan.  I mean, it’s nice to set the heating from a nice warm office somewhere – even easier for the manager than the entrance hall  – but there are more things that could go wrong and then there’s still the question of what exactly the user can do at the facade.  The architect’s original spec was for an internet model (the Myson Touch Wifi, presumably now model 2).  You don’t have to use the internet with that one – in a rare choice, it’s programmable at the facade – but  it has a strange twist on the top temperature limit. Instead of being “don’t let the user set the system lower than X degrees or higher than Y degrees”, it has “don’t let them change the temperature by more than Z degrees”.  Subtle, but that’s enough to raise questions for us, since any range wide enough to make the user comfortable if we forgot them will be too wide to be effective for groups where we did.

Well, at least I’m prepared to explain their choices now.

Leave a Reply

Your email address will not be published. Required fields are marked *