can a tekmar get smarter?

CC.Rob_2

My tekmar 260 was programmed about 8 weeks ago. Have not changed any setting since. I have been keeping a log of the outdoor temperature and the corresponding supply target temperature. (Trot down to the boiler in the AM, note the outdoor temp and the supply target on a piece of paper.) Since it refuses to get cold here in southern New England, there have been a lot of "repeat" outdoor temperatures during my morning visit (e.g., the 39F outdoor this morning has been common for weeks). I noticed last week that the supply target temp for a given outdoor temp is now a couple degrees cooler than it was back in October. For example, in late October, 39F outdoor would have been 134F supply target. This morning at 39F outdoor, the supply target was 131F. This appears to be true across the (narrow) range of outdoor temps we've seen the past month or two. Target temps are now being computed 2-3F lower than they were back when the system was initially installed.

It is also possible that the autodifferential has been stretching out (was ~10F, now more like 15F), but I have not been noting that information. Casual observations only.

I know little to nothing about the details of PID logic. Only thing I can figure is that there is some feedback between the differential and whatever logging of supply target temp (or rates of temp change at the sensor over some time interval) that allows the control to do a little "fine-tuning" of the reset curve to maximize the differential relative to the target temps.

There is no indoor feedback. Just the outdoor sensor and the boiler supply sensor. Two plain-Jane thermostats (LUX TX500) running two heating zones. The house stays solid, temperature-wise.

Anyone know what goes on in the mind of the 260?

Thanks.

Brad White_164

PID Control

Proportional-Integral-Derivative or PID control is self-educating. The 260 is uses PID logic but I am unfamiliar with the particular internal logic details of the 260.

Suffice it to say though that PID logic uses experience with gains relative to setpoints over time to make adjustments to system output. If it determines that a lower supply water temperature or an earlier start time for example, it will act accordingly to a prescribed priority (time versus temperature for example).

Now if the control starts to call itself the "HAL 9000" and starts calling you "Dave", I would get another control and soon.

CC.Rob_2

educating

Thanks Brad. Nice to know it can in fact get "smarter." Unexpected benefit in a way that increases efficiency. Neat! I had looked at the reset curve as something rather static, defined only by the initial input variables and whatever equation it uses for the curve. Apparently not the case. Ongoing optimization is good.

It does behave itself, so long as I respect the force field now in place around the boiler and controls....

Brad White_164

You are welcome, Rob

Keep in mind that if you make certain adjustments such as the curve or maximum boiler settings, design outdoor temperature, etc., you will essentially wipe out your control's developed internal database. It will then have new parameters and will have to "re-educate" itself to the new criteria.

I am not sure if the control re-sets the curve or if it uses "parallel shift" or even "skewed shifting" whereby the curve is offset to achieve the ideal indoor target temperature. For example, sometimes milder days require warmer water than the primary curve suggests. On some colder days a cooler water may be needed to "maintain" an established indoor temperature. General description of curve shifting, not necessarily specific to the Tekmar 260. Perhaps the folks at Tekmar will chime in with the real story?

Brad

CC.Rob_2

fiddling

I have been resisting the urge to tweak a few things for just that reason. Wouldn't want to flush its hard-earned data.

Just a couple days at design would answer the question "how good was my heat loss calc?" and allow any final tweaks. I probably overestimated the required supply temp at design and want to see real-world performance before using the number I got from the heat loss software.

Mike T., Swampeast MO

Almost sounds like the reduction in target is half of the increase in differential. Hmmm....

Don't forget that your thermostats do provide feedback to the controller in the form of, "How long is a heat call in zone x at y temp outside?" and, even more importantly, "How long between heat calls in zone x at y temp outside?".

Presuming this is a non-modulating boiler, it sounds to me like the controller is attempting to lengthen heat calls by firing the boiler less frequently and/or for longer periods of time. Such could very easily be related to the relatively mild weather.

If you believe that your curve is significantly higher than necessary, I would reduce it.

It seems easier to start with a curve that's too low and "sneak up" than to start with a curve that's too high...

I'd suggest a slope at the bottom end of what is recommended for your type of emitters.

Dave_4

Tekmar

This makes me wonder if using a timer to enable the unoccupied building at night would be a far better choice than using setback thermostats?

Rob_30

my thinking precisely

Mike: Was thinking about this on the ride home. Length of stat calls at various temps, etc. probably all play into this. Thanks

It is a non-modulating boiler (Revolution).

I was pondering the merits of starting too low vs. too high, and decided to err on the high side so as not to get a call "hey, it's not warm enough!" from those stuck at home while I'm off somewhere (usually warmer...) for work. Hence the need for a few design days. To see how it does overall.

As for a bump in the curve at the low end, it does appear to have one (or at least it did a few weeks ago) for this terminal unit (fin-tube). Check out
http://forums.invision.net/Thread.cfm?CFApp=2&&Message_ID=287702 for a missive on that.

Uncle Pete: was thinking about the timer, but opted to keep the stats as they are for now. We're on the verge of ditching even the modest 2-3F setback anyway.

Al Roethlisberger

How do the thermostats and controls like the 260 interact?


I posted this question earlier, but didn't see a response.

So, the outdoor temp resets like the 256 and 260 obviously employ their logic to control the water temperature, and how often the boiler cycles.

But, of course, they do this by interpreting data such as the call from the thermostat.

However, many programmable thermostats(including ones like the Tekmar 500 series) allow the thermostat to learn or statically program cycle lengths typical for a particular heat source.

So, my question is this: With some thermostats trying to be "intelligent" and statistically learn ideal cycle and lead times for the boiler, and then devices like the 260 using the calls from that thermostat to then learn its own data to manage temperatures and cycles.... will they work cooperatively, or will the two never quite settle into an ideal efficiency as they both try to adjust against each other?

Simply put, does a simple "dumb" thermostat work better with devices like the 260, or do they compliment each other?

Thanks
Al

CC.Rob_2

good question

I had the same concerns. Maybe I didn't give it enough time, but I ended up replacing an old Chronotherm III with a simpler stat about two weeks after the 260 was installed. During that two weeks, that zone was all over the map, cycle-wise. I think the tekmar was trying to do one thing, and the CT3 another. Not sure if they would have come to an understanding or not.

The stats I have now are very basic and as I interpret the manual (and what I observe in their operation), base their heat calls around temperature swing and do not attempt to control the cycle rate. If the temp is below target, the stat calls until it's satisfied. That probably gives the 260 a very reliable data stream on call duration, etc. With the ODR, the stat calls are very long (hours) as heat input seems to be pretty close to the heat loss.

Bottom line so far is that the system seems to be doing just what it ought to be doing. Temps stay very solid, recovery from modest setbacks is acceptable, etc.