T87 thermostat

Jamie Hall

mattmia2 said:

In theory it is supposed to be set to the current of the control it is operating, right?

That's the starting point. You go from there up or down as needed. It actually is a lot more responsive to varying conditions than @ChrisJ suggests. It's a little hard to visualise, perhaps -- never mind explain -- but it does not create a fixed time interval or cutoff at all, but the system responds (as it should) to both the magnitude of the deviation from set point and to the rate at which the set point is being approached.

ChrisJ

@Jamie Hallno matter how much you like old stuff it doesn't change how it actually performs.


The T87 is not as good performance wise as a Vision Pro and it never will be.

For longevity and style i agree it was better but it's not a better thermostat.  There's a reason the Nest copied it's style.   Honestly I don't know why Honeywell didn't before them.  A digital t87 style modern tstat with a color screen would've been really cool.

It is what it is.

109A_5

ChrisJ said:

That's the rules they came out with but it depends on the system.

All it is is a heating element that warms the bimetal spring a little to shut the system off early.  It works very well under stable conditions.

The adjustment changes how much it heats it and since it uses the current from the appliance that's there that comes in.

The standard setting for steam is 1.2 which is memory serves is technically (off).   

For my purposes I will not give up my T87, it holds the temperature so constant at its location I can't see a difference.

If you understand the design of the heat anticipator, it is a resistor that needs to generate a specific amount of heat to accurately bias the mercury switch and the bi-metal spring. The anticipator resistor is variable to electrically accommodate as many different heating systems as possible. The variable resistor's indexing is in Amps, so when set correctly it would generate the same about of heat regardless of the system. This variable resistor also apparently has the side effect of being able to affect the heating systems cycle rate (for better or worse and / or to just simply fine tune the systems operation to minimize temperature overshoot). The heat from the resistor warms the air around it and may also help draft air through the thermostat's housing, allowing better sensing of the room's temperature.

Without further study I think an additional benefit of the heat anticipator is it narrows the dead band, hysteresis or differential of the thermostat's switching set point.

" The standard setting for steam is 1.2 which is memory serves is technically (off). "

My understanding of this setting (1.2) is that it is based on older systems that their Gas Valves that were not manufactured by Honeywell and / or other equipment that drew a lot more current as compared to a newer Honeywell Gas valves like on my boiler that is over 45 years old. A heat anticipator resistor for these higher current systems may not have fit nicely in the T87's housing. With this seriously overly generic 'Steam' setting, I believe it gives the heat anticipator a bad name, since in some cases it may be set wrong and as such does not operate as originally intended.

My gas valve draws 0.260 Amps and the heat anticipator is set as such, and works great. In my case if it were set at 'Steam' or "1.2", I bet the temperature here would seriously overshoot every cycle.

rick in Alaska

I will take a T87 any day, but you can't give me one of the new electronic ones. I have had more than one that wouldn't work at all.
Rick

ChrisJ

I think the point I'm trying to make that @Jamie Hall and @109A_5 is missing is the fact that amount of anticipation needed changes with the weather.  The amount needed when it's 10f out is very different from when it's 40f out.

Not that you can't adjust it manually.

Jamie Hall

You don't have to. The anticipation changes in line -- quite reliably -- with the RATE at which the set point is being reached, which is exactly what you want it to do. It's not fixed by any means.

109A_5

ChrisJ said:

I think the point I'm trying to make that @Jamie Hall and @109A_5 is missing is the fact that amount of anticipation needed changes with the weather.  The amount needed when it's 10f out is very different from when it's 40f out.

Not that you can't adjust it manually.

You may want to re-think that. My experience here (most radiators) near the perimeter of the house thermostat near the center of the first floor. Is regardless if it is 0 degrees Fahrenheit outside or 60 degrees when I walk by the thermostat and look at it the thermometer matches the set point.

Now that being said if I do a 8 degree setback the temperature will overshoot. So I would bum it back up in 3 or 4 degree steps over the course of an hour or two.

For the cost and the age of the technology I think it works incredibly well. However if you want to deal with the high tech issues and spend lots of money to do that it is fine with me, just leave me out of that feedback loop.

ChrisJ

Ok.

I'm open to changing my mind.

How isn't it a fixed amount? You're adding the same amount of heat to the thermostat no matter what. No?

There's times my system shuts off 4 minutes before the thermostat and there's times it shuts off 15 minutes before the thermostat.


How can a fixed heater in the thermostat perform the same function?


Now, my system is an extreme example but the new Honeywell thermostats with the CPH do attempt to do this and adjust it vs overshoot and undershoot.

Jamie Hall

The secret is that that poor little heating element has to heat the air moving through the thermostat. If that air is cooler or moving faster, it can't heat the bimetal as effectively, which will decrease the anticipation time (it will shut off later). It's rough -- you can indeed do better with a thermostat which senses both error and rate of change of error and bases the prediction on that plus the known or "learned" constants for the space, but that takes more computing power than low line digitals can manage. High line ones can do it, of course -- part of what you pay for.

I'll happily grant that it isn't perfect, and getting it set right is a bit fiddley (the current draw figures are only an approximation). But it's simple and it works.

EdTheHeaterMan

Children! Children! BEHAVE! @ChrisJ has answered @DanHolohan question with a NO. and @Jamie Hall, @109A_5, and @rick in Alaska all say Yes to Dan's Question. That will settle it.

There will be no discussion on the topic or you will be sent to your rooms without your porridge.

Now shake hands and say you're sorry.

No more bickering.

Now don't you feel better, now that this is over.

Sometimes I think we need a den mother to handle these things. Where is @Erin Holohan Haskell when you need her?

ChrisJ

@EdTheHeaterMan

I don't feel we were bickering.
I very often have such discussions with @Jamie Hall and I feel we always keep them respectful.

EdTheHeaterMan

I was just making light of the conversation. I enjoy adding a bit of levity @ChrisJ

But I am in between you and Jamie in age so I need to put my vote in. I like the analog because it is easier to understand. But if I was born in the 1990s I might be writing code and boasting about digital superiority.

In college, I worked on an IBM 360 that took up the 5th floor a downtown office building. I think my laptop has more computing power that that whole series of metal boxes had in the 1970s

ChrisJ

What.
Born in the 90s?

So now the insults start flying eh!   

EdTheHeaterMan

1890s ?

DanHolohan

I love you all. 😂

109A_5

Since Jamie got to it first, I would like to add that the heat that the resistor generates is not enough to overcome the dead band, only narrow it a bit. Also the placement of the thermostat is usually centered in the middle of the heat sources (radiators in my case) and the heat sources are placed where the most dissipation is of the heated environment. So for the heat to radiate to the thermostat as the walls, windows ceiling cool down it takes longer to warm the thermostat. Since the heat anticipator narrows the dead band the heat source is shut down early. Depending on the heat source type the heat anticipator can be fine tuned. No need for constant tuning. Also by comparison a slight undershoot is better than a large overshoot.

Also keep in mind when the thermostat shuts off the call for heat the resistor cools down. The mechanical hysteresis of the bi-metal spring and the bulb with mercury in it and a touch of heat is amazing. Probably complicated to emulate in software or even other mechanical means.

The heat anticipator does not make the evenness of the heat in a room more even, other than by just very constant temperature at one fixed point.

ch4man

T834 setback thermostat. still looking for a subbase

Jamie Hall

I'll bet with some ingenuity you could MacGyver one starting with a T87 subbase and some careful tinkering.

Jamie Hall

I always respect @ChrisJ -- and he always makes me think hard about the topic at hand, which is a very good thing.

Erin Holohan Haskell

DanHolohan said:

I love you all. 😂

Same here!

Mistro

Hey Dan, I don't think I've seen you since we and Jim O. shared a street March with protestors in Frankfurt last century!
Anyway, we installed a few thousand T-87s and loved them, then and now. They were simple and sophisticated, reliable and understandable. Would you believe the instruction booklet was 8 pages long? T-87's were a mastery of engineering prowess. Yet so simple, one dial for desired temp, another for temp. What's not to love?
We have a lot of stories, about spiders and flies and geckos and brats, and yes, marital temperature wars. For setback, X-10 made a little timer heater to be installed under the stat to make it think it was warmer in the room than it was, or not. We used them as master-slaves and freeze protection and deceptive control and more. They were fun!
We never had to train our clients, never got calls to send someone to program the damn thing, the batteries never died, and they went on working forever. We still have hundreds in service and sometimes we get requests to remove the frustrating so-called smart stat and reinstall T-87s.
A favorite story is when an especially important client went on a junket to Israel, came back home in the middle of winter and half of his zones would not heat. He was 7 hours away. I drove all night to get there, only to find little Jewish prayer flags hanging on the thermostats, keeping the mercury bulb from moving. He just laughed; I did too, but did not think it so humorous. I wondered if Bhuddist prayer flags would have the same effect.
Anyway, "there is no treasure in this life, richer than a memory". Thanks for the reminder.
Warmly,
Michael Luttrell
(still hiding out on the left coast)
Warm Floors, now Warm Corp West.

RayWohlfarth

@EdTheHeaterMan I like the way you think LOL

ch4man

T834 set back thermostat. Im still h

Jamie Hall said:

I'll bet with some ingenuity you could MacGyver one starting with a T87 subbase and some careful tinkering.

its about 3/8" larger diameter. yeah I tried

EdTheHeaterMan

ch4man said:

T834 set back thermostat. Im still h

Jamie Hall said:

I'll bet with some ingenuity you could MacGyver one starting with a T87 subbase and some careful tinkering.

its about 3/8" larger diameter. yeah I tried

@ch4man
Go back to page 1 of this thread. My first post is about a customer who's father must have been a close friend of MacGyver.

Tinman

Just saw this on Twitter as a gag Father’s Day gift.