Thermostat rapid cycling?

Camperch

Thermostat cycles within seconds when calling for heat???

I have a 2 zone hydronic system with a TACO 3 zone controller (ZVC403). Honeywell CT87K for 1st floor (replaced)
Honeywell mechanical CT30A1005 for the second floor (simple mechanical heat only stats)
System has (1) 1" TACO zone control valve (571-3) and (1) TACO 3/4" zone control valve (571-2) (heat operated type)
I manually lower the 1st floor setting at night so its cooler for sleeping (66F). When I wake up I raise the 1st floor to 72F for the rest of the day.
When the weather is cold enough to require heat towards the early morning, from my bed I hear the boiler cycling on and off between 2 seconds and 20 seconds. I observed the thermostat contact opening and closing within seconds, the boiler controls are fine. If I jump out the stat the boiler runs fine. Also checked thermostat wiring by jumping out at stat all OK.
I blamed the 1st floor electronic stat and replaced it with same stat as on the second floor a Honeywell simple mechanical stat. Same result cycle every few seconds, which I found was happening with both mechanical stats.
This seems to happen when the house is cold, and the boiler is cold. I tried all different heat anticipator setting, I keep them at the lowest setting. On the highest setting you can actually feel and uncomfortable amount of heat on your finger from the anticipator heating resistor. I tried the 1000hm resister trick for power stealing stats which doesn't apply here anyway (same results).
I am ****-U-Ming the current through the thermostat is from the control board relay coil and not the actual heating element in the TACO valves?
HELP getting tired of getting up 4am running up and down stairs to see what's happening with the system.
Thanks
Matt

Jamie Hall

Somewhere in that system -- a little hard to say where -- something is running way too much current through those thermostats.

First off, be sure they are getting 24 volts, not 120 volts. If that's OK, can you sketch up a wiring diagram for us?

Camperch

It operates off 24 volts which I confirmed. The amperage draw through the thermostat is .99 amps @ 24V and jumps to 1amp @24V when the valve end switch makes. The system is wired as per the diagram for the TACO controller, two wire thermostats to R and W and two zone valve outputs 24V going to valves with one wire triggering the boiler relay once the end switch of either valve closes.

Jamie Hall

The anticipators have an adjustment range from 0.10 to 1.2 amps, so a 1 amp draw should be OK. Should be. If, however, they are getting hot to the touch, they are getting more current than that -- and with both of them misbehaving, I'm more inclined to look for circuit problems than thermostat problems.

Camperch

OK, thanks, with the anticipator turned to 1.2 the resistor does not "feel" hot to the touch, although I am sure its producing some heat. I can't figure any other way to wire the circuit. Seems to be more stable with the cover off. Tonight I am going to modify one of the stats and eliminate the anticipator entirely and see what happens in the morning. I don't mined get new thermostats, but I don't think that's the answer. Thanks for you replies, I will report back.

109A_5

Hello @Camperch,
I would think when wired correctly the thermostat current would be a lot less than the Zone Valve current, in this case.
What is the point of running an 0.99 amps through a thermostat just to see if it is closed (a Call for Heat)? If your answer is heat anticipator functionality, that will work with a lot less current.



EDIT: OK I got a better look at a ZVC404-4, appears there is no current buffering, the box is just a pretty wire nut(s) with a transformer, a Pump relay and pretty indicator lights and priority if you need it.

Camperch

I jumped out the anticipator resistor, prevented the second floor from calling to avoid confusion, and around 3 am my wife heard the boiler start, woke me, I checked the radiators they were still cold so I know it just started, the boiler stayed on and did not turn on and off through the cycle. I am going to contact TACO and asked about a compatible thermostat for the controller and see what they say. I don't know if the thermostat triggers a TRIAC, or a relay to supply power to the valve, but that is where the load would be if the power to the valve is isolated from the stat. I will report back. Oh--- is Saturday, may have to wait till Monday for TACO, I thought I saw a spec sheet from TACO indication the thermostat draw, I am going to look for that in the mean time. Thanks Matt

Camperch

Experiment phase two- I reconnected the 2nd floor stat to observe the operation while I'm painting the room. I can see the contacts constantly making and breaking every few minutes with a small arc across the contacts(cover removed) its worst with the cover in place. TACO FAQ states 1 amp is normal draw at 24V. I just realized, I installed the same exact set-up in my previous house using the TACO controller and valves and did not experience this. Going to invest in a different stat, like I had in the other house and see what happens.

109A_5

Hello @Camperch,
I guess my first question should be, did the existing equipment work correctly (Historically) before any device was changed ? If it did work correctly what changed ? Some non-Mercury round Honeywell thermostats have a lithium coil cell buried in them. Others may rob a small portion of the 24 VAC sine wave to power themselves (a Burden Voltage). The small Voltage drop across a heat anticipator resistor may be part of that strategy. Just a theory, I have not had the opportunity to reverse engineer your type of thermostat.

The TACO zone valves with the wax motor technology are rated at 0.9 Amp draw. Once they warm up that 0.9 Amps draw has a duty cycle to keep the wax warm but not burn up the unit. So seeing 0.9 Amps at the thermostat does not surprise me since there appears to be no current buffering in the ZVC403.

Things I would consider.
A poor connection somewhere in the system so the Burden Voltage of the thermostat actually becomes too low (during a call for heat) for its proper operation, so it cycles fast. This could include low 120 Volts to the "TACO 3 zone controller (ZVC403)" during a call for heat. Does it get worse (or change) with both zones calling ? Does the 24 VAC on the ZVC403 circuit board remain intact during a call ?
Low battery Voltage in the thermostat (if it has them).
The heat anticipator setting may be very critical in the case of your system. I would try a setting of maybe 0.8 just as an experiment. Too high (less resistance), the burden Voltage of the thermostat is maybe too low for proper operation, too low, the heat anticipator makes too much heat and the call for heat terminates too early (but not in just seconds).

Personally I like the old Mercury thermostats. Way less drama, except from the environmentalists, and they don't live here.

Camperch

The entire control system was added this fall, except for the electronic Honeywell yo-yo type I had for the house. I made the second floor a separate zone, so I landed up with two zones installing the TACO controller and TACO zone valves. I purchased a temporary plain Jane heat only bi-metal thermostat for the second floor just to keep it at 66F while working up there. The system operated to maintain temperature on both floors. I had trouble with the gas water heater next to the boiler, and had to relight it's pilot. When the house is quiet like at night, I can hear the hot water heater or boiler starting up. I heard what a thought was a failing hot water heater (spare thermo-couple still taped to the side of the unit just in-case). Turned out to be the boiler cutting in and out all that time. I replaced the Honeywell yo-yo stat (electronic) with the same bi-metal stat used on the second floor, but it kept cycling usually around 4am as the house called for heat. Sometimes just as the valve opened and the flue damper on the boiler would begin opening it would cut out, some times it would run for a few seconds.I jumped out the thermostat wires and the boiler runs faithfully so I know the wiring and controls are OK from the stat down. I can actually see the contacts opening and closing on both stats. I tried all the anticipator settings, and tried the advised resistor for "power robing" stats no good. I have never experienced anything like this before, I am a fan of the bi-metallic stat. I ordered an electronic programmable stat and see if it works better. In the meantime the stats are working fine with the anticipators shunted. Thanks for you input it is much appreciated Matt

hot_rod

Could be the vent damper end switch causing erratic burner operation
Does it have a switch or lever to force it open, manual override?

Camperch

The system including the damper end switch function faithfully when the stats are jumped,
Thanks

Jamie Hall

I presume that by now you have managed to cure the symptom, blaming it on the Honeywell thermostat. However, I doubt very much that you have either found -- or corrected -- the problem, and I wonder when it is going to arise to bite...

Camperch

The TACO controller has Green indicator lights showing a call for heat and Red indicator lights showing if the valve end switch is closed. Not that I am relying solely on the indicator lights, but it is easy to see when the stat makes and breaks at the boiler also. I will monitor the 24Vs under load to be sure the transformer can keep up, although if it could not, the system would drop out regardless of the stat. ???? I will report back. Once again your sharing of knowledge is appreciated. I'm beginning to think the cheap bi-metal stat is just poorly engineered.

Camperch

To Jamie, still troubleshooting, not cured yet, not blaming stat just yet. Honeywell has making controls for many years. I will report back with what I find.

HomerJSmith

I believe you have a wiring problem. Try this. Move the offending thermostat to the priority position on the Taco control board, move the three wire 571 to the priority position, turn the priority switch to off and see what happens.

That 571 has to be wired to the transformer on the control board properly. Use the following diagrams to the Taco control board and 3 wire ZV.

Camperch

Hello HomerJSmith
I have a TACO ZVC403-4 Zone control. I have two pumps and two zone valves. The first floor stat goes to Zone 1 thermostat connections R and W. The first floor Zone valve is wired as you have shown 1,2,3. The second floor is wired the same way except to zone 3 with the priority setting off so that the two pumps run independent of each other. Why two pumps? Please don't ask, I can't type that much tonight. I was painting in the room on the 2nd floor where the stat is in, with the cover off, the contact are fluttering on-and off constantly. The system can not get the radiators hot until the room temp drops low enough for the magnet to grab the larger contact and hold it in for a, a heating cycle will warm the room up to temperature then the cycle begins again with the fluttering. The stat maintains perfect room temperature, but at the expense of cycling the boiler and pumps may 30 or 40 times per hour! (don't let it do that unnecessarily) I noticed the stat has a larger contact that gets drawn in when close enough to the magnet. BUT inside the contact and isolated from it is a small contact ahead of the main contact which when it makes first, before the main contact, it powers the heat anticipator through a fine wire, as small as the heat antcipator wire itself. I am thinking, when the small inside contact makes before the main contact gets drawn in by the magnet, it supposed to only power the heat anticipator and not begin the heating sequence until the main contact gets drawn in. The small contact is triggering the controller to start a heating sequence, then dropping out as it warms, which happens quickly. I wonder if this stat should not be used with this type of controller, or zone valves? although all of this equipment is so common. And yes the stat is rated for 24V. I am going to try to contact Honeywell and TACO Monday and see what they have to say. Thanks for your interest and advice, I will look over all the advice I have been given here and add it to by troubleshooting. I will of course report back any findings and/or resolution.

kevink1955

Something is wrong and it is not the stat. You said "The amperage draw through the thermostat is .99 amps @ 24V and jumps to 1amp @24V when the valve end switch makes"

Something is wrong there, the current should be the same whenever the stat contact is closed. What is controlled by the end switch should not draw current thru the stat contacts.

Most times the load on the end switch is powered independently of the stats supply and should have no affect on stat current. If that's not how the taco controller is doing it it's a poor design.

Waiting to hear what Taco has to say when you talk to them

109A_5

A few things;

The difference between .99 amps and 1 Amp is 10 milliamps or 0.010 Amps. In this type of circuit I would not even worry about that, it could just be the way that meter displays an even smaller variance in the current. Do you have a high quality expensive multimeter, with 1 milliamp or less resolution on the scale you are using ?

Another way of saying it, should we worry about a 1 % change in current, even if it is actually true ? I don't think the Wax motor in the Zone Valve cares that much. The Zone Valve is rated at 0.9 Amp draw anyway.

The symptoms you are experiencing with the CT30A1005 have shown up on Amazon reviews, so I believe it is a thermostat issue.

If I am looking at the correct picture of the internal workings of the CT30A1005. Since the stationary contact appears to be a set screw adjustment. Probably varies the magnet's influence on the contact that moves with the bi-metal spring. Or the hysteresis of the magnet's influence. I think if you like experimenting I would turn the screw so it moves to the right so the magnet can grab the moving contact better, keeping the heating system engaged longer. It would then take more force of the bi-metal spring (room heat and anticipator heat) to shut down the heating system. The set screw adjustment may not need much change.

The one thing I not real clear with is the original Honeywell CT87K that appeared to have the same symptoms ? And maybe it did, but for other reasons. Or maybe you are now fighting multiple issues.

Camperch

Hello, I am using a quality GreenLee meter, I'm not too concerned about the .1 amps but I can retest and also take a reading with the zone valve load out the circuit to see if the load is through the valve and see is the voltage fluctuates. The fact that I first suspected the electronic stat also make me hesitant to blame the stats, unless it was the upstairs stat causing the cycling all along, I ran from the basement to the second floor many times at 4am tracking this, but that's before coffee maybe I screwed up. Right now I'm having coffee after the first floor cycled this morning and did not cut out once (except on boiler water temp.) with the anticipator jumper on the stat. Even without the anticipator it maintains a good room temperature differential. What I did not try is setting the Honeywell electronic stat to a longer cycle/hr setting with the dip switches set to steam, I should have tried that, even so it should not cause this rapid cycling regardless of any setting. I don't think I ever changed an anticipator setting before this. I think the anticipator is more for a steam system where you have residual heat from large cast iron radiators long after the room temperature is satisfied. I'll looked at that set screw, and thought it was there to calibrate the actual temperature to match the dial indicator. Thanks Matt

clammy

Personally I just get rid of those taco heater type zone valves they are slow to open and usually need a secondary transformer . I wonder does your boiler have a automatic vent damper ? I ve seen issue on single transformer system where between the damper and valve motor exceed the va of the transformer . Go over the installation manual for the taco zone controller and be sure it’s wired correctly .i de also say get rid of the yo yo pos and get ab1/2 way decent thermostat like a Honeywell t 4 pretty simple . it would appear you have issue w your wiring and end switches on the zone valves . How old are the zone valve bodies ,and how long has this been a issue ? Personally I use taco sentry valves and tacos zone valve controllers for years and have really never had a issue . Start over and jump out your thermostat wires and if your still having a issue then it’s not you stat look elsewhere or possibly call in some one familiar w low voltage zone valve wiring some guys just suck at wiring anything and if it a birds nest of wires that ain’t going to help any one .the new non mercury t 87 are garbage . Obviously after multi attempts to correct it’s time to call some one who does this for a living because as easy as it seems u less your do it for a living you will never know the in and out and the should and should not be of the business and lack of familiarity w what your looking at . Sorry not everything is a you tube video or internet answer in minutes some times you gotta pay some one to fix some ones else mess . Peace and good luck clammy

109A_5

Hello @Camperch,
The heat anticipator settings is and has been and interesting and confusing subject. And then add the 'Steam' setting and it gets worse. Often the 'Steam' setting effectively disables the heating of the anticipator resistor. Historically I believe the old gas valves drew current in excess of 1.2 Amps. So disabling it would prevent it from burning up and before it burned up it would create excessive heat causing thermostat performance issues. Once it burned up there would be no heat. Manufactures could have put in higher wattage resistor but that may have had other downsides.

I have steam heat and the much newer gas valve that draws about 0.300 Amps and that is about where the anticipator is set and it works great.

It is my opinion the industry has not bothered to better define the 'Steam' anticipator setting due to the variety of equipment possibly still out there. However they do recommend measuring the current and use that as a starting point. Digital thermostats that do not have a heat anticipator needed some method of simulating the heat anticipator behavior it so they went with the Cycles per Hour method since that could be done with software timers. Older mercury thermostats had no correlation to time.

Yes I believe the set screw may influence the temperature calibration. It would not surprise me if the set screw would be calibrated first for proper hysteresis / dead band calibration and that there another calibration method for the temperature done second.

Or maybe you are just having issues with poor anticipator engineering.

Camperch

I'm confident with the tinkering I have done through out the years I can handle this. I reached out to this site due to vast experience as a group all of you have and appreciate your input. I know a lot of you have domestic repair experience. I am not going to assume anything and continue troubleshooting. As strange as it may seem even though I am retired, I am very busy with home project deadlines, and visit this as much as I can. Thanks again

Camperch

Found this post interesting on The Wall Feb 2005 Short Cycling Boiler... last post - It's working good now.

Thanks again for all the responses everyone. I went out and got 4 CT500 stats and installed them. The CT500 was the only one they had, so I went with it. From what I can tell at Honeywell's site, it's basically the same stat as the others that were suggested (all digital manual). I couldn't see any difference in them (although I'm sure there are some very minor differences). I installed them Friday morning and now, Saturday afternoon, I've yet to see a single zone cause short cycling. The T87s that I removed were all reading 68 when I pulled them out. The CT500s all seemed to read 70 when first plugged in (I assume this is where the default is set from the factory), but quickly changed to 65. I also picked up a wireless remote thermometer so I could have something else to check the temps with to see how accurate the new stats are. All of my readings are within 1 degree of each other. So the new stats seems to be pretty accurate. Seems as though all the T87s were off by 3 degrees or so like someone mentioned above. I sett all the heat cycle rates at 3. The temp control default at c1. The house is as comfortable as can be now. Most of the zones run for around 3 minutes or so before shutting off.

I don't want to mock the T87s as a whole, but it sure seems like the ones I had were worthless. Now I can finally gloat about how much I like my new heating system

Not to say this is it, but it does sound familiar, Matt

Joe Mattiello

571 zone valves are very reliable. They do take approximately 1.5 min to open because of its inherent design. If you have heat anticipatory type thermostat, turn it all the way up.
if that doesn’t help.
might be faulty thermostat. It happens 
call Taco Tech support at 401-942-8000 if you still have issues 

Camperch

Thanks Joe, I sent a message to TACO via email this morning. I will try calling tomorrow. The valves controls and boiler are functioning normally, just find it hard to believe a Honeywell stat could possible cause problems unless it was damaged out of the box? Thanks for responding.

109A_5

Hello @Camperch,
If you read other post on other Honeywell products like the Pressuretrol and Vaporstat. Newer versions of those products suffer from poor performance and sloppy calibration issues. My opinion as with other products as time has gone on their quality has diminished. Also to me it seems Honeywell is into marketing so many different things now they have lost sight of what they are commonly known for by the general public.
https://www.honeywell.com/us/en

BTW The thread mentioned above;
https://forum.heatinghelp.com/discussion/87130/short-cycling-boiler

Camperch

Going to tell you my responses from TACO and Honeywell, then go back to painting. TACO has not heard of any type of thermostat that caused issues with the ZVC403 controller operating 571 zone valves.
Once I mentioned to Honeywell the thermostat was connected to a TACO controller they to me to call a professional. I still think I am quite qualified to figure this out. Don't want to seem stubborn, but I can't help referring back the thermostat as the problem. I have seen old bi-metal stats of all different brands operate in some instances for over 35 years under the harshest of conditions where the equipment would fail and the old stat remain and never give a problem unless they were physically damaged or mangled by someone. I will of course continue to let you know the outcome. Thanks everyone for your support.

109A_5

Hello @Camperch,
If I understand correctly you bypassed the heat anticipator resistor and the rest of the thermostat works normally. So the heat anticipator which should enhance the thermostat operation is what is causing the problem.

So, to me, either the anticipator is the issue (it does not function correctly, probably making too much heat) or the hysteresis of the switch (bi-metal and magnet) with the influence of the anticipator is not calibrated correctly.

Have you tried the CT87K with the other zone totally disabled ?

I'm guessing a contractor would just install their favorite (currently available) thermostat and it would all be done. However the cost to you would be a lot more than what you are doing now.

Camperch

At this point, its easier for me to just get two new stats, I already installed a LUX $50.00 stat on the 1st floor. I like it so much, I ordered and second on for the 2nd, should arrive today. Already the 1st floor is performing perfectly, and I manually operate the second floor until the stat comes. None of this is urgent, I have heat, its just so unusual, I reached out to the Wall to see if anyone else experienced this. Yes, at some point I need to go back and recheck the CK87, since I don't believe both style stats can fail like this. I know for sure the second new stat installed on the second floor is NG just watching it flutter.I will report back, Thanks

Jamie Hall

There really are only tthree basic ways in which a thermostat can provide control and, at the same time, enough hysteresis to avoid "fluttering".

The original, wonderful, bulletproof mercury thermostats -- of which the original Honeywell T87 was an example, but not the only example, did it in the simplest way possible: there was a bimetal coil on top of which was a mercury switch. As the coil cooled, it uncurled, and that tipped the mercury switch, and the blob of mercury slid to the other end and made contact. This also, however, unbalanced the bimetal, which then had to heat up some to tip the mercury switch the other way. Simple. Dead reliable. To provide anticipation, there was the anticipator -- a very small heating coil -- which also heated the coil (a little) when the circuit was closed. It was adjustable.

Non-thermostats with a bimetal almost always had open contacts (I think there were a few later ones where the bimetal operated a microswitch), and they relied on a magnet; as the bimetal bent, it would get close enough to the magnet to be pulled the rest of the way to the contact. Since the magnetic pull varies with distance, again the strip had to cool below the trip point before it would pull away. These could have an anticipator, too. Again, simple, but not quite as bulletproof, as the contacts could be damaged or get dirty, resulting in unreliable operation. Also, if anything happened to the magnet or the steel on the bimetal to which it was attracted, these could flutter very badly -- if they worked at all. The type with the microswitch relied on the difference between the "on" trip force and 'off" trip force to accomplish the hysteresis, in a very uncontrolled way.

The last is, of course, the newer digital thermostats, which use a variety of electronic schemes to convert a voltage or resistance signal (usually resistance) to a temperature and compare that to a set temperature and then trigger either a relay or a triac. They use various electronic (sometimes digital, sometimes analogue) schemes to control the hysteresis.

Which type was the type you are having trouble with?

Camperch

If I'm not mistaken, the 1st floor electronic stat (Honeywell) CT87K "seemed" to be cycling on and off on an early morning call for heat. This stat which I understand has no heated anticipator, actually states CPH with the dip switch settings Steam, Gravity being 1 CPH up to 9 CPH for electric. I tried different setting, but cycling in less than a minute should have nothing to do with the CPH. So I took the bi-metal stat from the second flr. replaced the electronic stat with it, and the problem continued regardless of anticipator settings. Finally, I shunted out the anticipator and the stat stopped cycling so often, BUT the new stat (bi-metal) which I had ordered in the meantime to replace the second floor stat is much worst, the contact flutters via a small tab turning the boiler on and off even before the main contact is drawn in by the magnet. I am going up right now and re-install the electronic yoyo to see what happens. I have a new LUX electronic stat working fine on the 1st floor now and another LUX coming today. I don't know who makes LUX, but its what I had in my previous house and did not have this problem. Thanks for your extensive explanation about the thermostat operation.

Camperch

Just put the Honeywell CT87K in the second floor, I believe it might be a bimetal type, OH NO - Just heard the boiler cut in then back off, I just got back downstairs, let me try to catch it in the act.

Camperch

So much for painting! Glad I didn't wet the roller yet!
Heard the boiler on when I came down after installing the stat and thought I heard it cut out, check it was running OK. As I was saying, the yo yo Honeywell is probably a bi-metal with and electronic CPH circuit since I heard a distinct click as I raised the set-point. Let see what happens.

Camperch

Found this interesting, God Bless Amazon, went to check on the thermostat I ordered yesterday, now when you view the tracking, a Google map comes up showing a dot by my address and a dot by where the delivery truck is, and how many stops it has till it gets to my door! ARE WE SPOILED!!!!!

Camperch

LUX is made by Johnson Controls. Familiar with Johnson Controls, operated and work with there controls for 37 years.

109A_5

Camperch said:

Found this interesting, God Bless Amazon, went to check on the thermostat I ordered yesterday, now when you view the tracking, a Google map comes up showing a dot by my address and a dot by where the delivery truck is, and how many stops it has till it gets to my door! ARE WE SPOILED!!!!!

May help reduce porch pirate activity.

Anyway this heat anticipator style may not work as well at the higher currents like with Wax motor Zone Valves.

dko

Those CT30A's have extremely high return rates that I don't even bother selling it anymore. It's like the 5PSI low pressure gauges we use for gas tests, one knock and it stops working. Delicate, fragile innards.

Are the Lux thermostats that are working properly battery powered?

T87K board photo below, electronic
Should be the same as the CT87K

hot_rod

Doesn't the thermostat just see the relay in the Taco box, not the actual zone valve load? I thought the relay isolated the load from the stat?

This sure sounds like more than a T-stat issue.

Jamie Hall

Ah. I thought we were talking about the Honeywell round thermostat -- the T87series. Um. I guess not?

Camperch

Back from Shop and Stop ha (of mice and men), I asked TACO what current the thermostat sees, is it only the relay or TRIAC on the circuit board? he said yes, so I said so it will always be 1 amp correct, he said it depends on the zone valve, so I said so the load goes through the valve and he said yes, so I don't know which. I am going to lift the leg off of one of the zone valves while measuring the draw and I'll let you know. To 109A_5, Yes it's the same stat as the picture you sent. If you look at the rectangular loop that contacts the magnet, there is a smaller inner contact that makes first with the screw and has a fine wire leading to the bi-metal of 2 ohms I don't know what it does, but when the magnet draws it in it gets shunted, I'll try to upload a picture of what I am talking about.