I'm trying to figure out the logic for an American Standard system I had installed back in Oct 2018. This is in central Texas.
The air handler is the variable speed TEM6A0C48 and paired with the 2 stage Gold 17 4A6H7048 outdoor unit with a line set length of 30ft or so.
Since it was winter when it was originally installed, the tech did not add more refrigerant, said it looked good enough on the gauges and said to call him in the summer to double check but that didn't happen. Now in Nov 2019 (one year later), the unit ran low on refrigerant. We eventually discovered a crack in the line set. A 12" section on the suction line was cut out and replaced. It was winter again so hard to use gauges to adjust the charge so the unit was evacuated and the factory charge weight was added along with "a bit extra" (.5lb?) for the line set length. I give this back story as after we got the unit up and running again in Nov 2019, it seemed to just work "better". I have a fluke IR meter and I noticed it seemed in cooling it was blowing 54 degrees versus 55 or so, and in heating it previously would not get higher than 88 degrees when it was in the 30's outside but it would quickly get to 88 and then slowly climb to 90. So it seemed perhaps it was low on refrigerant for the first year with the factory charge. It certainly performed better in the second year at least.
And then in July 2020, this past summer, lightning stuck near my house. Everything was fine other than the 824 thermostat on the unit was off. Tech came out and verified there was no power going to the thermostat so he replaced the control board but the thermostat still would not power up. Put in a new thermostat and we are up and running again. I can't remember if he checked the charge or not, I know I wanted him to since it's never been checked in the summer, but can't remember.
The short version of this long back story is I currently don't think my defrost board is working right but I don't really have anything to compare it too. I've never owned a heat pump and during the first winter I believe the charge was low. For the second winter, it was extremely mild with temps rarely going below 40 degrees during the day time.
Which brings us to the current blizzard sitting over Texas right now. This winter had (had!) been quite mild until the last few days, but it seemed like the unit was going into defrost more but I could ignore it. Now that temps are staying below 40 degrees for days on end, I think something is wrong.
As the temps got into the 40's on the 10th, I noticed it was defrosting more than I remembered in previous years. I started to pay attention to them and time them and in most cases, it was going into defrost once per hour. Very little steam would come off of the unit. Temps were still between 35 and 40, no visible frost was ever on the unit. BUT, every now and then, the unit would go 2 or three hours with no defrost. So being that the CNT05010 defrost board is supposed to "learn", I figured perhaps it had just "learned wrong" during the previous months of mild weather. So I switched off all the breakers to the indoor and outdoor unit to reset any learned parameters. I did this around 11pm so it would "learn" in the worst conditions. I restored power and bumped the thermostat up so it would run for at least 30 minutes, and like clock work, after 30 minutes of run time, it went into defrost. Ran a few more minutes after that before I lowered the thermostat some and went to bed thinking I'd solved all my problems. I checked the history once I woke, there were 7 defrost cycles so I knew nothing changed.
I have a couple of servers in my home office that I can graph room temperatures from.
Here is the chart from that night. The red lines are the defrost calls. You can see after resetting the indoor and outdoor units, everything was mostly normal for about 4 hours and then it starts defrosting at every 30 minutes of total run time. Pulling some weather data, the dew point and temp were mostly the same all night. It makes no sense why it would suddenly have to start defrosting continually.
So back to the drawing board. I noticed the defrost cycles were longer than I remembered. Usually in the previous years, they'd be 3 or 4 minutes. These were running 5 - 8. Where the unit is at in my yard, the north wind blows right through it, and it had been really windy for days. Figured maybe the extra defrost were due to the wind making defrost cycles take longer and thus making the control board think it was due to ice and then calling for more cycles. I got some rubberized pipe insulation and wrapped and tied it around the defrost coil sensor to see if that would change anything, but I don't think it did.
So back to the drawing board again. The only thing left is to figure out how the board is supposed to work and see if reality lines up with that. I think I've read just about every post on the board that has the word "defrost" in it. From what I gathered from the fine folks on here, the board measures the "coil" temp against the ambient temp and as ice accumulates, the "coil" temp goes down as ice insulates against heat absorption. This makes sense, until I realized that the "coil" temp sensor was not mounted on the coil, but on one of the tubes after the EEV manifold. It is measuring refrigerant temp after the EEV which of course will be lower than ambient. Maybe as the unit ices up, the EEV closes more to produce a bigger drop and that is what it is looking for? Of course in defrost mode, the "coil" senor would actually be measuring in the other direction and it would be the last place to actually warm up. So maybe the defrost board is basing more logic on how long a "defrost" takes and just continually runs defrosts until it finds the right interval to get a 3 minute defrost cycle? But that sounds awful and would obviously defrost way too much as temps get lower or outdoor wind speeds increase. I thought that was my original issue, hence me trying to insulate the coil temp sensor. But no go.
Digging more, I finally found an original document for the specs on the defrost board: Trane_Heat_Pump_Defrost_Control_Book_34-4101-07.pdf
So that basically states that it is looking for the "coil" temp to continually fall below ambient. And a defrost is needed once this 'delta' goes beyond a combination of pre-programmed and learned logic. So seems easy enough to measure what my unit is doing and see if that is the case.
I had 'reset' everything again yesterday evening and started plotting numbers. The unit was basically running for 15 minutes on, and 15 minutes off. So every second run, I'd hit the magic 30 minutes total runtime and would get a defrost cycle. So basically I knew when a defrost was about to happen. So I ventured out into the cold right before I knew the unit would come on. Ambient was 31. Unit came on, I watched with my fluke IR meter as the tubes coming out of the EEV quickly went down to about 19 degrees. Over about a minute, they slowly rose to 22 degrees. A couple more minutes and it was at 24 degrees.For the next 6 minutes or so, the temps meandered between 26.5 and 29 degrees before a defrost cycle kicked in (right on time!). The whole time, the defrost board had the green LED flashing once per second, I've only seen it ever flash once per second, even when I had both temp sensors unplugged to measure the ohms (previously measured 31k on both while it was 33 degrees out, while unit was running was reading 31k and 28k-29k).
So according to the "theory" of how it is supposed to work (calling for defrost as the delta increases beyond a calculated value), I'd say it is not working right. But then it dawned on me that I've only been "testing" things in 1st stage heat. The unit rarely ever goes into 2nd stage, even in the middle of summer, but everyone figured a 4 ton unit was right for a 2200 sqft house. Perhaps, I'm oversized. So perhaps the defrost board needs to "learn" with some 2nd stage heating calls? So with this newly thought of knowledge, I set the thermostat to a 1st stage heating lockout when temps are below 40 degrees. And with that, off I went to bed knowing that the unit would run in 2nd stage only. I also set the thermostat to 67, had been leaving it at 70. I wanted to really give it a workout in the morning.
So I awaken this morning. Check the history. There were 17 calls for heating, and 5 calls to defrost. More than I'd want, especially since it was super dry out with the north wind, but at least an improvement. I bumped the temperature back up to 70 from 67. It was currently 29 outside. 2nd stage heat came on and blew at 93 degrees from the vents for about an hour. Impressive! After about 15 minutes, it ran again for about 30 minutes. Still no sign of a defrost. Looking good I think. Then after another 15 minutes, it cycles on and a few minutes later, a defrost. I still don't see the point, but for nearly 2 hours of run time, I guess it deserved it. It was a longer defrost, about 5 or 6 minutes, once the compressor fan kicked on again, I got a small puff of steam but that was it. Disappointing. On the next cycle, you guessed it, another defrost. I let it run with stage 1 heat locked out for most of the morning just to make sure. It eventually made it back onto it's 30 minutes of runtime before a defrost schedule again. Now the funny thing is, I noticed while running with stage 1 locked out, the defrosts were really cooling the house down more. I measured the vents blowing at 65 degrees. Strange, since 72 is normally the lowest I would normally see in a defrost. Looking at the thermostat, it showed a call for AC2 + ID1. Normally it is AC1 + ID1. No joke, if it enters defrost in stage 1 heat, it runs the compressor in stage 1! What bone head engineer thought that was a good idea? Or is this perhaps a relic of times past when a 2 stage unit actually had 2 compressors in it? This new bit of information I find really depressing. I've since taken off the stage 1 lock out, and am mostly back where I started. Only now out of ideas.
At this point, I'm not sure what to do. Is this normal behavior? Or is my unit perhaps undercharged or overcharged? Seems to be running really well, stage 1 heat puts out at 88 degrees when it is 30 outside. Puts out at 90 when it is 32. Can't really complain about that, but these defrosts are killing me. Extra noise from the fan running on high, cooling the house down, and then the constant dimming of the lights as the compressor flips back and forth. Or is the board actually bad? It was struck by lightning after all, but has no faults and I tried shorting the test pins and forcing a defrost and that worked fine.
And then the irony is, while I spent 2 hours typing this up... The dry weather went away and it started to drizzle outside, freezing drizzle. Suddenly the defrost cycle goes away. The unit runs for over 90 minutes of runtime without a defrost (lost the exact count). Eventually it goes into defrost, really long, like 8 minutes. And then makes a huge steam show. This was also the case a few days ago at the start of this cold front. It was 35 outside and raining for most of the day. I could not see ice on the coils, but there was ice build up on the fan grill on top. It was several hours of runtime before it went into defrost.
It seems to be, at below 38 degrees outside (and dry out), once the house temp equalizes and the system falls into a 15 minute on, 15 minute off cycle, that I run into this defrost "issue". This can't be normal?
The air handler is the variable speed TEM6A0C48 and paired with the 2 stage Gold 17 4A6H7048 outdoor unit with a line set length of 30ft or so.
Since it was winter when it was originally installed, the tech did not add more refrigerant, said it looked good enough on the gauges and said to call him in the summer to double check but that didn't happen. Now in Nov 2019 (one year later), the unit ran low on refrigerant. We eventually discovered a crack in the line set. A 12" section on the suction line was cut out and replaced. It was winter again so hard to use gauges to adjust the charge so the unit was evacuated and the factory charge weight was added along with "a bit extra" (.5lb?) for the line set length. I give this back story as after we got the unit up and running again in Nov 2019, it seemed to just work "better". I have a fluke IR meter and I noticed it seemed in cooling it was blowing 54 degrees versus 55 or so, and in heating it previously would not get higher than 88 degrees when it was in the 30's outside but it would quickly get to 88 and then slowly climb to 90. So it seemed perhaps it was low on refrigerant for the first year with the factory charge. It certainly performed better in the second year at least.
And then in July 2020, this past summer, lightning stuck near my house. Everything was fine other than the 824 thermostat on the unit was off. Tech came out and verified there was no power going to the thermostat so he replaced the control board but the thermostat still would not power up. Put in a new thermostat and we are up and running again. I can't remember if he checked the charge or not, I know I wanted him to since it's never been checked in the summer, but can't remember.
The short version of this long back story is I currently don't think my defrost board is working right but I don't really have anything to compare it too. I've never owned a heat pump and during the first winter I believe the charge was low. For the second winter, it was extremely mild with temps rarely going below 40 degrees during the day time.
Which brings us to the current blizzard sitting over Texas right now. This winter had (had!) been quite mild until the last few days, but it seemed like the unit was going into defrost more but I could ignore it. Now that temps are staying below 40 degrees for days on end, I think something is wrong.
As the temps got into the 40's on the 10th, I noticed it was defrosting more than I remembered in previous years. I started to pay attention to them and time them and in most cases, it was going into defrost once per hour. Very little steam would come off of the unit. Temps were still between 35 and 40, no visible frost was ever on the unit. BUT, every now and then, the unit would go 2 or three hours with no defrost. So being that the CNT05010 defrost board is supposed to "learn", I figured perhaps it had just "learned wrong" during the previous months of mild weather. So I switched off all the breakers to the indoor and outdoor unit to reset any learned parameters. I did this around 11pm so it would "learn" in the worst conditions. I restored power and bumped the thermostat up so it would run for at least 30 minutes, and like clock work, after 30 minutes of run time, it went into defrost. Ran a few more minutes after that before I lowered the thermostat some and went to bed thinking I'd solved all my problems. I checked the history once I woke, there were 7 defrost cycles so I knew nothing changed.
I have a couple of servers in my home office that I can graph room temperatures from.
Here is the chart from that night. The red lines are the defrost calls. You can see after resetting the indoor and outdoor units, everything was mostly normal for about 4 hours and then it starts defrosting at every 30 minutes of total run time. Pulling some weather data, the dew point and temp were mostly the same all night. It makes no sense why it would suddenly have to start defrosting continually.
So back to the drawing board. I noticed the defrost cycles were longer than I remembered. Usually in the previous years, they'd be 3 or 4 minutes. These were running 5 - 8. Where the unit is at in my yard, the north wind blows right through it, and it had been really windy for days. Figured maybe the extra defrost were due to the wind making defrost cycles take longer and thus making the control board think it was due to ice and then calling for more cycles. I got some rubberized pipe insulation and wrapped and tied it around the defrost coil sensor to see if that would change anything, but I don't think it did.
So back to the drawing board again. The only thing left is to figure out how the board is supposed to work and see if reality lines up with that. I think I've read just about every post on the board that has the word "defrost" in it. From what I gathered from the fine folks on here, the board measures the "coil" temp against the ambient temp and as ice accumulates, the "coil" temp goes down as ice insulates against heat absorption. This makes sense, until I realized that the "coil" temp sensor was not mounted on the coil, but on one of the tubes after the EEV manifold. It is measuring refrigerant temp after the EEV which of course will be lower than ambient. Maybe as the unit ices up, the EEV closes more to produce a bigger drop and that is what it is looking for? Of course in defrost mode, the "coil" senor would actually be measuring in the other direction and it would be the last place to actually warm up. So maybe the defrost board is basing more logic on how long a "defrost" takes and just continually runs defrosts until it finds the right interval to get a 3 minute defrost cycle? But that sounds awful and would obviously defrost way too much as temps get lower or outdoor wind speeds increase. I thought that was my original issue, hence me trying to insulate the coil temp sensor. But no go.
Digging more, I finally found an original document for the specs on the defrost board: Trane_Heat_Pump_Defrost_Control_Book_34-4101-07.pdf
So that basically states that it is looking for the "coil" temp to continually fall below ambient. And a defrost is needed once this 'delta' goes beyond a combination of pre-programmed and learned logic. So seems easy enough to measure what my unit is doing and see if that is the case.
I had 'reset' everything again yesterday evening and started plotting numbers. The unit was basically running for 15 minutes on, and 15 minutes off. So every second run, I'd hit the magic 30 minutes total runtime and would get a defrost cycle. So basically I knew when a defrost was about to happen. So I ventured out into the cold right before I knew the unit would come on. Ambient was 31. Unit came on, I watched with my fluke IR meter as the tubes coming out of the EEV quickly went down to about 19 degrees. Over about a minute, they slowly rose to 22 degrees. A couple more minutes and it was at 24 degrees.For the next 6 minutes or so, the temps meandered between 26.5 and 29 degrees before a defrost cycle kicked in (right on time!). The whole time, the defrost board had the green LED flashing once per second, I've only seen it ever flash once per second, even when I had both temp sensors unplugged to measure the ohms (previously measured 31k on both while it was 33 degrees out, while unit was running was reading 31k and 28k-29k).
So according to the "theory" of how it is supposed to work (calling for defrost as the delta increases beyond a calculated value), I'd say it is not working right. But then it dawned on me that I've only been "testing" things in 1st stage heat. The unit rarely ever goes into 2nd stage, even in the middle of summer, but everyone figured a 4 ton unit was right for a 2200 sqft house. Perhaps, I'm oversized. So perhaps the defrost board needs to "learn" with some 2nd stage heating calls? So with this newly thought of knowledge, I set the thermostat to a 1st stage heating lockout when temps are below 40 degrees. And with that, off I went to bed knowing that the unit would run in 2nd stage only. I also set the thermostat to 67, had been leaving it at 70. I wanted to really give it a workout in the morning.
So I awaken this morning. Check the history. There were 17 calls for heating, and 5 calls to defrost. More than I'd want, especially since it was super dry out with the north wind, but at least an improvement. I bumped the temperature back up to 70 from 67. It was currently 29 outside. 2nd stage heat came on and blew at 93 degrees from the vents for about an hour. Impressive! After about 15 minutes, it ran again for about 30 minutes. Still no sign of a defrost. Looking good I think. Then after another 15 minutes, it cycles on and a few minutes later, a defrost. I still don't see the point, but for nearly 2 hours of run time, I guess it deserved it. It was a longer defrost, about 5 or 6 minutes, once the compressor fan kicked on again, I got a small puff of steam but that was it. Disappointing. On the next cycle, you guessed it, another defrost. I let it run with stage 1 heat locked out for most of the morning just to make sure. It eventually made it back onto it's 30 minutes of runtime before a defrost schedule again. Now the funny thing is, I noticed while running with stage 1 locked out, the defrosts were really cooling the house down more. I measured the vents blowing at 65 degrees. Strange, since 72 is normally the lowest I would normally see in a defrost. Looking at the thermostat, it showed a call for AC2 + ID1. Normally it is AC1 + ID1. No joke, if it enters defrost in stage 1 heat, it runs the compressor in stage 1! What bone head engineer thought that was a good idea? Or is this perhaps a relic of times past when a 2 stage unit actually had 2 compressors in it? This new bit of information I find really depressing. I've since taken off the stage 1 lock out, and am mostly back where I started. Only now out of ideas.
At this point, I'm not sure what to do. Is this normal behavior? Or is my unit perhaps undercharged or overcharged? Seems to be running really well, stage 1 heat puts out at 88 degrees when it is 30 outside. Puts out at 90 when it is 32. Can't really complain about that, but these defrosts are killing me. Extra noise from the fan running on high, cooling the house down, and then the constant dimming of the lights as the compressor flips back and forth. Or is the board actually bad? It was struck by lightning after all, but has no faults and I tried shorting the test pins and forcing a defrost and that worked fine.
And then the irony is, while I spent 2 hours typing this up... The dry weather went away and it started to drizzle outside, freezing drizzle. Suddenly the defrost cycle goes away. The unit runs for over 90 minutes of runtime without a defrost (lost the exact count). Eventually it goes into defrost, really long, like 8 minutes. And then makes a huge steam show. This was also the case a few days ago at the start of this cold front. It was 35 outside and raining for most of the day. I could not see ice on the coils, but there was ice build up on the fan grill on top. It was several hours of runtime before it went into defrost.
It seems to be, at below 38 degrees outside (and dry out), once the house temp equalizes and the system falls into a 15 minute on, 15 minute off cycle, that I run into this defrost "issue". This can't be normal?
source https://hvac-talk.com/vbb/showthread.php?2226613-Trane-Demand-Defrost-Logic&goto=newpost
No comments:
Post a Comment