PDA

View Full Version : New split system install fluctuating high side







MrTube
06-06-2017, 04:34 PM
Hi all, having a bit of an issue and decided to step back for a while and ask for help. Never seen this before and really don't know what to think.

Here's what happened.


2 stage Copeland Ultratech (unloads part of scroll for 67% capacity in low stage) 3 ton
variable speed air handler with TXV. R410A. Flowed N2 while brazing and pressure tested at 250 PSI. Pulled 100 micron vacuum which stabilized at 120 microns before charging with 2 pounds of virgin refrigerant before opening service valves on condensing unit.


Air handler up in attic approx 20 feet above condensing unit.


Problem : high side pressure fluctuates, lowside steady.




In low stage as per MFG charging instructions :
Suction pressure steady at 145 PSI.
Liquid pressure fluctuating 250-275 PSI. Compressor gets louder as pressure drops. No bubbles in sight glass 24" before TXV.


Attempted to charge system by subcooling, but didn't realize highside gauge was sticking which hid the pressure changes at first.
Highside pressure slowly climbs to 275 PSI and then rapdily drops to 250 PSI causing compressor to get louder as it drops. Subcooling around 9 degrees depending on pressure.

Assumed TXV is starving, causing hunting, added 1 pound 6 ounces of refrigerant slowly, resulting in overcharge, pressure now fluctuates between 250 PSI and 350 PSI. Ambient temp : 70F. Subcooling very high at this point, overall. Suction line temp at evaporator fairly steady 45-46 degrees F with overcharge.




Youtube video using lowside gauge as highside gauge, with machine running in HIGH stage before overcharge. This is the HIGHSIDE with a blue gauge. - Waiting on a replacement gauge as we speak. The behavior in high stage is the same as in low, just faster and a bigger pressure fluctuation.




https://youtu.be/T1GtxTniIyE


I obviously need to pull refrigerant out, will most likely recover the entire chargeand start over. But what's causing the highside fluctuations!?

FaultCode
06-06-2017, 05:58 PM
Is TXV matched to condensing unit sizing?

What control, if any, is there on the condenser fan?

What model/make is the condensing unit?

Suction temperature seems a bit high for AC.

Just a note - your pressure test should have been at around 575psig not 250.

MrTube
06-06-2017, 06:04 PM
Is TXV matched to condensing unit sizing?

What control, if any, is there on the condenser fan?

What model/make is the condensing unit?

Suction temperature seems a bit high for AC.

Just a note - your pressure test should have been at around 575psig not 250.

Condenser fan is 2 stage, controlled by outdoor unit.
TXV came with air handler and bulb is in correct location. Air handler is matched to outdoor unit.

Both units are Amana/Goodman from local supply house.

Indoor : AVPTC37D
Outdoor : ASXC16036

MFG said to pressure test at 150 PSI which I felt was a bit low, but I didn't want to go completely crazy on it.

FaultCode
06-06-2017, 08:45 PM
Okay, so were the condenser fans cycling when you were monitoring the pressures?

Not a case of going crazy at those pressures, simply a correct test pressure for that refrigerant.

RANGER1
06-06-2017, 08:49 PM
Be interesting to know what pressures unloader load & unloads at, also condenser fan.
If you can stabilize things you may be able to isolate problem.
If you could also trend liquid subcooling & superheat to see if linked to discharge pressure fluctuations.
Normally TX valve will hunt if discharge pressure changes all the time.

Does it have a liquid receiver, or is it critically charged?

MrTube
06-06-2017, 09:11 PM
Thank you both for taking the time to respond.


Condenser fan doesn't have ability to cycle, simply runs on low in 1st stage with compressor partially unloaded, and in high in 2nd stage.

No receiver sadly.


Right now, it's assumed the system either some how has non condensibles in it, or is overcharged so we're going to do a recovery and re-weigh in the charge and start over. Seems like the best idea right now.

Has anyone ever seen a TXV cause high side fluctuations like this yet have a stable suction pressure? Everything I've seen seems to indicate a hunting TXV causes low side swings, not high side.

RANGER1
06-06-2017, 09:23 PM
Thank you both for taking the time to respond.


Condenser fan doesn't have ability to cycle, simply runs on low in 1st stage with compressor partially unloaded, and in high in 2nd stage.

No receiver sadly.


Right now, it's assumed the system either some how has non condensibles in it, or is overcharged so we're going to do a recovery and re-weigh in the charge and start over. Seems like the best idea right now.

Has anyone ever seen a TXV cause high side fluctuations like this yet have a stable suction pressure? Everything I've seen seems to indicate a hunting TXV causes low side swings, not high side.

Can only suggest liquid banking up in condenser.

MrTube
06-06-2017, 09:32 PM
Can only suggest liquid banking up in condenser.


I'm curious,
Is this the same reason non-condensibles can cause fluctuations? Because they can restrict flow in the condenser?

It would make sense, also explains why the compressor got louder as the highside pressure dropped. Seemed backwards at the time, but I have no idea what the actual head pressure was doing, likely climbing.

FaultCode
06-06-2017, 10:21 PM
Been reading Goodmans install manuals fir similar systems as can't find an Amana one.

You said then after vaccing system you added 2 pounds of refrigerant, was this the required amount for the extra length of lineset?

Goodmans state that their unit is pre-charged for a 15ft lineset. If we allow say 25ft (unit 20ft above) then at the additional charge rate of 0.6ozs/ft then only an extra 15 ozs is needed.

Just a thought - wonder whether service valves let by N2 during pressure test and contaminated the charge.

Grizzly
06-06-2017, 10:42 PM
Hi Mr Tube.
I may be way off.
But I have seen something similar happen when the Liquid line filter is Blocked.
Just a thought!
Grizzly

Glenn Moore
07-06-2017, 01:58 AM
Hi Mr T
There is probably nothing wrong with your plant, just the Copeland scroll loading and unloading which it is designed to do. Unlike normal Scrolls the scroll has an unloader controlled by a small solenoid which energises and de energises to load and unload to keep a steady suction condition to control both temperature and humidity. The simple act of the unloader is to reduce the capacity of the the machine by approx 33% by simply exposing a portion of the suction intake of the scroll back to the compressor shell. So the suction condition is stable but the sudden reduction in compressed gas at this time leads to a fall in the discharge pressure. The discharge pressure goes up and down as the unloader is pulsed by the control stat. So Your txv and your system is working as designed you just have to learn the new compressor technology, also the compressor gets noisier when it unloads due to the unloader mechanism, it's just a different version of the digital scroll. So let it run , set the thermostat and have a nice cuppa and a bacon sarnie all the best

Grizzly
07-06-2017, 05:36 AM
We all live and Learn Glenn!:D

Grizzly

MrTube
07-06-2017, 06:16 PM
Been reading Goodmans install manuals fir similar systems as can't find an Amana one.

You said then after vaccing system you added 2 pounds of refrigerant, was this the required amount for the extra length of lineset?

Goodmans state that their unit is pre-charged for a 15ft lineset. If we allow say 25ft (unit 20ft above) then at the additional charge rate of 0.6ozs/ft then only an extra 15 ozs is needed.

Just a thought - wonder whether service valves let by N2 during pressure test and contaminated the charge.


Amana and Goodman are essentially the same thing, so what you found is likely correct.
The extra was for the length of the lineset, 20 feet is only the height difference.

I've wondered about N2 leaking passed the service valves as well, though I seem to get a lot of different opinions on the subject.

Is it typical for non condensibles to cause fluctuations on the highside such as this?

MrTube
07-06-2017, 06:28 PM
Hi Mr T
There is probably nothing wrong with your plant, just the Copeland scroll loading and unloading which it is designed to do. Unlike normal Scrolls the scroll has an unloader controlled by a small solenoid which energises and de energises to load and unload to keep a steady suction condition to control both temperature and humidity. The simple act of the unloader is to reduce the capacity of the the machine by approx 33% by simply exposing a portion of the suction intake of the scroll back to the compressor shell. So the suction condition is stable but the sudden reduction in compressed gas at this time leads to a fall in the discharge pressure. The discharge pressure goes up and down as the unloader is pulsed by the control stat. So Your txv and your system is working as designed you just have to learn the new compressor technology, also the compressor gets noisier when it unloads due to the unloader mechanism, it's just a different version of the digital scroll. So let it run , set the thermostat and have a nice cuppa and a bacon sarnie all the best


Hi Glenn,

Thank you for taking the time to read my question and reply. I appreciate it!

I think you may be mistaken though. As far as I am aware, this style of Copeland either runs in low stage, or high stage depending on what the thermostat in the building calls for. I think you're thinking of Copeland's digital scroll which actually controls the unloader to change the duty cycle of the system. A really interesting sounding setup, but way above what I could afford to use. This system does have a high and low pressure switch, but they're there just to protect the compressor and shouldn't do anything under normal conditions.

I was tempted to grab my meter and watch to see if the solenoid was getting power or not, but according to the control board it claimed it was running C1 which is Cool 1 and stayed there. As you said, it almost seemed like it was going between high and low stage for some strange reason.

This is my first time installing a 410A split system. Most of my work is on antique refrigerators running a low pressure refrigerant similar to R11 and R123. That said, I do understand the importance of working clean and dehydrating a system well and keeping NCG out. The antique machines don't even have the luxury of a filter drier, just a strainer.

Which is why I was shocked when a few guys suggested I may have a NCG problem. I've heard of guys that feel N2 leaked passed the service valves while pressure testing, while others feel the system came with a contaminated charge from the factory. I spent hours pulling a 100 micron vacuum and then ensuring it stabilized at 120 microns before charging.

I guess anything is possible.

Right now, it seems like either the system has NCG, which my understanding is likely stuck at the top of the condenser. Or, is overcharged which can behave similar?

I'm sorry for asking so many questions.
I have no doubt we'll get the system running fine this weekend, but I want to understand what happened and how to prevent it in the future as well as I just like learning. The NCG issue is interesting as even the 1930s manuals say it gets trapped in the condenser. I guess it's similar to air in a hot water heating system. It gets stuck in the first high spot, and never leaves.


We'll be using a friend's recovery machine to pull the charge. Pump the system down and then weigh in virgin R410A. This should solve any concerns of NCG, as well as an overcharge.

I've considered changing the filter drier as well, but the problem is it's mounted in the outdoor unit and we'd have to cut the system open and likely relocate the new one before the TXV. To me, it's safer to avoid opening the system to atmosphere at all costs.

Grizzly
07-06-2017, 07:26 PM
Hi Mr Tube.
You have not mentioned carrying out a deep vacuum.
Which boils off any moisture and removes non condensable's Prior to charging with refrigerant.
A recovery rig is not suitable for anything other than recovery and recharging the system.
If you use a vacuum pump as you should, any opening to atmosphere is negated.

If I am reading your post correctly and no vacuum pump has been used.

Then your system is compromised.
Also all the above becomes irrelevant.
Grizzly

MrTube
07-06-2017, 07:46 PM
Hi Mr Tube.
You have not mentioned carrying out a deep vacuum.
Which boils off any moisture and removes non condensable's Prior to charging with refrigerant.
A recovery rig is not suitable for anything other than recovery and recharging the system.
If you use a vacuum pump as you should, any opening to atmosphere is negated.

If I am reading your post correctly and no vacuum pump has been used.

Then your system is compromised.
Also all the above becomes irrelevant.
Grizzly


Hi Grizzly.

I used a 6 CFM 2 stage Robinair pump after testing it alone and confirming it could pull down to 60 microns.
I pulled the system down to 100 microns using that pump, shut all valves including the blanking valve on the pump and turned the pump off. The system raised to 120 microns and stabilized. I then waited another 10 minutes to make sure the system did not climb any higher, and it did not.

The micron gauge is connected to my core removal tool which keeps my connection as close to the system being evacuated as possible. It also allows me to isolate it from my manifold and hoses.


You'll notice in my photo my charge line is connected to my N2 tank while a larger hose is going to the vacuum pump. The smaller hose used for purging, and pressure testing was shut off during this. I then bled it of N2 and once it was full of 410A proceeded to charge.

I didn't even remove the micron gauge before the system was under pressure as I didn't want to risk sucking air in.

Sorry for the confusion.

We will be using a vacuum pump after recovering the charge.

RANGER1
07-06-2017, 08:31 PM
MrTube,
Don't worry about asking questions, that's what it's all about.
If nitrogen in system it would normally end up in condenser & stay there.
If liquid seal is broken it can recirculate, so not sure in your case, thought evaporator superheat would see change.
Also not sure if there is a internal relief that let's go at "X" pressure
or unloader is forced on with high pressure.
Have you checked superheat readings to see variations etc?

Surely it has a refrigerant charge that you can weigh in, if not charge until sight glass stops flashing, or few bubbles instead of subcooling, as that means it's banking up in condenser.

Grizzly
07-06-2017, 10:26 PM
Yeah.
I agree with Ranger MrTube don't apologise for asking.
I was just thinking the easy answer to your issues.
Given what you have now confirmed, the answer is not so simple.
When running could you measure for a temperature difference across the liquid line filter.
what you measure will answer whether there is an issue there?
Good Luck and keep us all posted.
Grizzly

Glenn Moore
07-06-2017, 11:40 PM
Hi MrT
From everything you've told us I doubt if NCGs are present as your install and commissioning seems ok although the pressure test did sound a little on the low side for R410a.
The control stat will switch between the 2 stages around the set point to prevent short cycling by reducing th capacity. the internal solenoid is 24 volt dc . From what I've found out the difference in current draw between the 2 stages is around th 20 o 30% , so I would use a clamp meter to see if the current is jumping up and down which maybe caused if the diff on the control stat is to small. Try setting up stat to its lowest setting in o see if the pressure hunt stops.
Ps if the Tev is being starved of liquid then the Tev won't hunt as it will be driven wide open trying to get the superheat down and under control.
What type of stat is fitted?! What is the set point?
I'll discus with an old colleague of mine to see if he has seen this as he is now aCopeland man

FaultCode
08-06-2017, 12:11 AM
Sorry to keep on about it but what is the diameter and equivalent length of the suction line?

Is there an inverted refrigerant loop at the evaporator?

The subcooling should be measured at the evaporator according to Goodmans.

Airsteve
08-06-2017, 12:54 PM
2 stage Copeland Ultratech (unloads part of scroll for 67% capacity in low stage) 3 ton
variable speed air handler with TXV. R410A. Flowed N2 while brazing and pressure tested at 250 PSI. Pulled 100 micron vacuum which stabilized at 120 microns before charging with 2 pounds of virgin refrigerant before opening service valves on condensing unit.


Mr T
Are you charging the R410a as a liquid?

MrTube
08-06-2017, 03:41 PM
Hi all,

Thermostat is a Honeywell VisionPro 8000 series.

Airsteve : Yessir. Bottle upside down.

Faultcode, I think you're correct being concerned about the lineset size. The total length is only 34 feet.
I noticed this morning while looking at the information from Goodman my suction pressure is too high. They're calling for something closer to 105-107 PSI under the conditions I had. The lowest I saw was 140 PSI and this of course went up when I added some refrigerant.

I think I made the common mistake of trusting the sight glass. The system wasn't low when I started adding refrigerant in the first place.

Ranger1 is probably right. Liquid banking up in the condenser. I bet the system is grossly overcharged and the only reason my highside pressure wasn't crazy is because it was cool inside and outside.


Is it completely safe to remove R410A from a system while it's running using the highside into a recovery tank, throttling the flow to keep it liquid of course? My understanding is the "proper" way is to recover the entire charge and start over, but I'm curious if removing it as a liquid from the highside is acceptable.

Also curious which method will result in the least amount of oil loss.

FaultCode
08-06-2017, 08:45 PM
The only way to get liquid out of a running system is to connect to a liquid line. This should be your 3/8" line to the TXV.

MrTube
11-06-2017, 03:39 PM
So,

Recovered the entire charge, approximately 8 pounds. Factory charge was 6 pounds, I needed 12 ounces for lineset. That said, after the initial charge I added 1 pound 6 ounces more to see what reaction it would have so it's not surprising.

We added the factory charge back in plus 12 ounces. We closed the TXV 1/2 a turn and were able to get subcooling and superheat within factory specs, though the highside is still fluctuating around 25 PSI.

My friend said, to him, it looks like the TXV is slow to respond, and then ends up overfeeding so it shuts too much.


Is there any chance, this behavior is from the design and layout of the evaporator?
The setup comes setup to be vertical, but they allow you to lay it horizontal in two different directions. Mine, is horizontal and in the less used (blow right) direction. This blows into the A coil in the opposite direction of pretty much every other setup I've seen.


I've attached a picture of the evaporator layout and airflow direction. This is before the condensate drain and freeze thermostat was attached etc.

I've also attached a picture of the head of the TXV. This is on a 3 ton system.

RANGER1
11-06-2017, 09:03 PM
MrTube,
Can you tell us TX valve bulb location, pipe size, what position on suction line & is it in air flow.

MrTube
11-06-2017, 10:01 PM
MrTube,
Can you tell us TX valve bulb location, pipe size, what position on suction line & is it in air flow.

Yessir.
The bulb is on the suction line just before it leaves the cabinet. 7/8" OD tubing, the bulb is mounted in the 1 o'clock position directly over the equalizer line.

This picture is before I wrapped a bunch of 1/8" thick foam tape around it to insulate it, and the tube around it better. There's probably a good 3 or 4 layers around the bulb now. It's clamped tight using the supplied hardware which I believe is a copper strap and 2 bolts which I also insulated.

RANGER1
12-06-2017, 01:47 AM
7/8" tube in in middle of suggested recommended position.
If me I would mount it at 4 o'clock, also note position after equaliser.

http://www.achrnews.com/articles/102420-the-sensing-bulb


http://www.emersonclimate.com/Documents/FlowControls/pdf/Contractor%20Tip%20Cards/Tip-Card-08.pdf.

MrTube
21-06-2017, 06:27 PM
7/8" tube in in middle of suggested recommended position.
If me I would mount it at 4 o'clock, also note position after equaliser.

http://www.achrnews.com/articles/102420-the-sensing-bulb


http://www.emersonclimate.com/Documents/FlowControls/pdf/Contractor%20Tip%20Cards/Tip-Card-08.pdf.


Sorry it took so long to respond.

Here's a video showing the actual behavior of both low and highsides during normal operation. It is still fluctuating, though slowly and not a huge amount. Both sides are moving, I didn't notice that before due to my broken gauge.

https://youtu.be/DzocuMNVdlw

Oddly enough, I see everyone except apparently Goodman says to mount the bulb at 4 or 8 oclock? Goodman says 10 and 2.

I'm thinking of pulling the bulb out of the cabinet and mounting it further down on the suction line, as well as at the 4 oclock position. Thoughts?

Is there a chance this lower section of suction line could be acting as a trap, and occasionally causing liquid to work it's way up near the bulb in it's current position? In the standard vertical position, this "trap" wouldn't exist, it would be downhill rather than up. Picture rotating the entire evaporator 90 degrees clockwise. That's how they are typically installed.

MrTube
21-06-2017, 06:48 PM
Can anyone weigh in on this?

Which is it? Upstream or downstream? :)

FaultCode
21-06-2017, 08:10 PM
Sporlan 'recommend' that the sensing bulb be fitted up stream of the pressure sensing connection.

They also state their preferred bulb position.

They do however allow that manufacturers may have different installation requirements and that they should be followed.

Sporlan literature 10-9 and 10-10.

MrTube
21-06-2017, 08:16 PM
Sporlan 'recommend' that the sensing bulb be fitted up stream of the pressure sensing connection.

They also state their preferred bulb position.

They do however allow that manufacturers may have different installation requirements and that they should be followed.

Sporlan literature 10-9 and 10-10.


So in my case, my bulb is mounted exactly where Goodman wants it, but my TXV is apparently hunting.
What do I do?

It's not the charge, it's not NCG. The TXV is proper for the unit as per Goodman.

RANGER1
21-06-2017, 09:08 PM
So in my case, my bulb is mounted exactly where Goodman wants it, but my TXV is apparently hunting.
What do I do?

It's not the charge, it's not NCG. The TXV is proper for the unit as per Goodman.

Instructions are a bit confusing & contradict each other.
I would believe Danfoss than any other, but see below.
External equaliser on compressor side of TX bulb & mount bulb at 4 o'clock, also see in instructions about liquid traps in suction line around bulb area, as it must be avoided.
Manufacturers are not alway right either on everything, some designs are rubbish.

http://www.hvaceducationaustralia.com/Resources/Danfoss%20Resources/Fitters%20notes%20tx%20valves.pdf

http://www.emersonclimate.com/Documents/FlowControls/pdf/2004FC-141-R7.pdf

see pages 4 & 7