Tekmar 361 with indoor feedback sensor overshoots

Tim Potter

I have the Tekmar 361 installed with outdoor reset & the indoor feedback sensor.
When I am away (for long stretches of time, this is a vacation house) I can remotely switch on the unoccupied contacts & have the tekmar programmed for 55*. I leave my thermostat set to 68*, (my occupied temp). I have a Bumble Bee so constant circulation only runs in the 15 watt range. Through the magic of the internet I can monitor both supply & return temps as well as the room temps. What I am finding is the room temp overshoots to 60 or better upon a call for heat. Then Depending on the weather, It takes 24 to 36 hours or more to drop back to 55. The heating is in-floor pex in gyp upper floor & pex in concrete lower level (walkout) bedrooms.

When we are at the house (occupied setting 68*) it overshoots, sometimes to 72 or so.

Its always wanted to overshoot, so last summer (2013) I added the indoor feedback hoping that would close the loop so the 361 would know the indoor temps.

I was hoping that the control just needed to "learn" my system, but I've been able to monitor it remotely for about 3 weeks now, doesn't seem to be any change

I was thinking I have something set wrong in the 361 programming, but for the life of me I cant figure out what it might be. I have been over the programming guide (D 361) and don't see anything that appears to be out of whack.

I have the unit setup for outdoor reset
terminal unit set to 1 = High Mass radiant
room occupied 68
room unoccupied 55
outdoor design -20* (its damn cold in Grand County at times)
I lowered the mix design down to 110*
mix max 120
mix min = off

Thoughts?

Thank You:

Tim

Gordy

Is there solar influence?

SWEI

Is this a single zone system, or are there zone pumps/valves?

Where is the OAT sensor mounted?

Gordy

Another thing is floor covering choices over the gyp. And slab. Carpet of higher r value can expand the flywheel effect.

Techman

How did you add the indoor sensor( from the mechanical room to the sensor location) after the system has been installed/running for X amount of time. How did you determine where to locate the indoor sensor? And where is the sensor located ? How many zones?

Tim Potter

Gordy, yes i get solar pickup, & there is carpet on 1/2 the upstairs, tile on the other 1/2 & all carpet on lower level.

SWEI / Techman single pump with 6 zone valves controlled by their own t-stat. T-stats are set at 68*, & used as high limits. OAT sensor is on North side. Indoor sensor mounted in living room, south facing, It has lots of glass, its the one with the solar gain, & the room we are in the most.

Techman, I'm not sure what your 1st part of your question is asking

SWEI

The overshoot comes from a fundamental limitation of zone valves: They only have two positions (100% and 0%.) When you combine this with a tight ODR curve, the delay between when the thermostat calls for heat and the system delivers it can be quite long.

The first thing I would suggest you try is relocating the ODR sensor to the northeast corner of the structure in a location where morning sunshine hits it. This will send a message to the system that it is warming up out there long before the indoor sensor sees anything. You might also consider running two different reset curves: One for the rooms that have solar gain and another for the rooms that do not. This would require a motorized mixing valve and a bit more control work. Another option would be to replace one or more zone valves with remote bulb TRVs. This changes your on/off control to proportional control and will vastly improve temperature regulation.

Tim Potter

Interesting idea with moving the ODR sensor. I'll have to look at the temp logs from cloudy/snowy days to see if that could help.

The zone valve limitation makes sense, but wouldn't it be more in play during the occupied setting? The system is in
un-occupied mode now & all the zones are calling for heat constantly.

Techman

Tim, I was trying to figure out how you ran the sensor/wire to its location in the living room. Is there a Variable speed pump that feeds the single pump for the zone valves?
With all of the zones calling for heat, does the "path of least resistance" come into play here? Is it possible the L room has longer loops than the other zones?
SWEI, and others, Please be easy on me and my questions, I have several radiant customers that have had zero problems so the learning curve is low. I recently picked up a customer w/radiant problems w/ a tekmar 361 and I have been studying & reading and studying.This learning curve is steep.Thanks.

Tim Potter

Techman, I ran a 2 conductor t-stat cable from the Tekmar to the LR & spliced it to the indoor sensor.

Yes, the vs (007) pump pulls from the Boiler Buddy & is controlled by the 361 (injection system). The secondary loop (house side) has the Bumble Bee, in constant circulation as long as at least 1 zone is calling. Boiler Loop has a 007 wired to the control.

Gordy

Moving that outdoor sensor can be huge in dealing with solar gains, and overshoot.

Hopefully your carpeting selection was radiant friendly low r pad etc.

The carpet can delay room warming to the point where the slab will get warmed up to much then delay the response over time. Couple that with solar gain.

This all can be ironed out just takes some experimenting with the reset curve, and sensors.

TRVs can be an excellent benefit. Proportional control always trumphs bang bang technology.

Tim Potter

Gordy, how would you change the reset curve to help the overshoot? lower the mix design??

I probably wont be able to change the sensor location till next summer. Next time I'm out there, hopefully we will be up to our neck in snow, but we will see.

Techman, I'm not the original builder/owner, but the loops seem correct, ie: 2000' house, / 8 loops = 250'/loop

I think I would install TRV's if I were to build another house, they seem like the way to go.