If the pressures are truly normal and the system has the correct charge it should frost evenly. There are many issues that can cause the frosting issue you are having, including a out of calibration defrost sensor.

I have a suspicion that the valve was not leaking as 1.5 lb in a month is a lot of refrigerant. If you are looking for a leak you will notice that much of a leak after the hoses are disconnected.

I don't know if the same tech has been there every time but I think you need a better set of eyes on this unit.

 

Last time the tech did check the defrost sensor, and it was working. Also, I don't understand how that could cause it to have no frost on the top half, but a very thick coating right below that.

There was definitely a leak, as he saw the UV dye inside the service valve, and it was bubbling when he put soapy water on it. He said it was a small leak, but I don't know if that means much over a period of a month. I'm guessing 400psi will cause a fairly good loss of refrigerant over a month?

I should also say they cleared the system with high pressure nitrogen when installing the new air handler, but they did say it may not clear any restriction if the coils were clogged up.

Is the outdoor coil one big continuous coil from top to bottom? I see inside there are 4 pipes coming out of the TXV, so does that mean the coil is 4 separate sections? I'm just trying to figure out what the frost pattern would look like if there was indeed a restriction in the coil.

 

That unit should be off the ground several inches

 

Yes, you can't see in that photo, but it is sitting on top of a concrete pad, which is itself sitting on top of 4 concrete round pads, so it's at least 4 in off the ground.

 

When the coil defrosts all the water runs to the bottom. If the coil does not get hot enough before the unit shift out of defrost the water will freeze at the bottom which in turn causes the bottom to freeze faster which then develops a pattern where the ice at the bottom keeps getting worse. This can also happen with a low charge or a coil that is not feed equally, a bad TXV, or a control issue to name a few. It is hard to tell from the pictures but it looks like the unit could be 'banding" where the coil is not feed properly and frost bands form on the coil.

 

Yes, it is banding at the top. The bottom just completely frosts up.

The defrost doesn't last long...generally there is the 30sec delay when the compressor and fan turn off, then the compressor comes on for about 10 seconds (with outdoor temps around 45F), then the 30 sec delay. That is with it set to 30 mins. Not sure if that's normal, but even just turning off the indoor thermostat will cause the outdoor coil to start defrosting, so I'm guessing it's normal.

 

There’s ice between the pad and the drain holes.

Bet there’s a huge block of ice inside.

 

Just checked, and it's clear of ice inside, but it certainly has become completely clogged with ice inside previously. The drain holes under the coils are a few inches above the concrete pad and they're training ok. I'm guessing there are other drain holes near the centre of the unit, and that bit of the unit is touching the concrete pad.

 

Indoor coil and outdoor unit compatibility Are CRITICAL on a heat pump system.

Post the model and serial numbers here so we can first make sure that what you have will play nice together.

The initial symptoms you mentioned: "After installation there was some concern . . . as the high pressure was too high and low pressure was too low." concern me for component match.

And the lower-only frosting could be caused by a lack of refrigerant charge. My concern would be less for a leak and more to an intentional under-change as a way of disguising the "high pressure is too high" issue.

So let's start by learning the model and serial number of the inside And outside units.

PHM
---------

We've been having problems with our 1.5 ton Goodman heat pump icing up. It tends to frost up badly on the bottom, but the top 60% of the fins are completely clear.

Originally this heat pump was mated to an R22 air handler. I think the previous homeowner decided to replace the outdoor unit without realising he should have replaced the air handler as well. We replaced the air handler last year. After installation there was some concern that there might be a restriction in the coil due to the R22 oil, as the high pressure was too high and low pressure was too low.

This winter we noticed it icing up again. The HVAC company found the refrigerant was low, and added 2lbs along with some leak sealer. Pressures were ok after, so they don't think there is a restriction. A month later it was still icing up at the bottom again, and they again found it about 1.5lb low. They topped it up again, and found the leak in a service valve, which is now repaired.

The problem is, it's still frosting up at the bottom just a few days later, so I suspect the frosting is unrelated to the leak.

I just want to get some opinions before spending more money on this.

Could there be a restriction in the coil where it's starting to ice up? Would the pressures look normal if the restriction is partial and enough time has passed since starting the compressor for the pressures to equalise before and after the restriction?

The system is set up for 30 min defrost with the delay setting, and defrost seems to be working fine on the top half of the coil. Entire coil does completely defrost when I switch to A/C for 20 mins.

View attachment 814177

 

Air handler is AVPTC25B14 (2 ton goodman variable speed). Outdoor unit is 1.5 ton 13 seer, HSPF 7.7 with time/temp defrost. Not sure where to find the model number.

Here is the comment from when the new air handler was installed:

"appears there was a leak on vapour service valve at one point but was not leaking at this time. Recharged system with 7.5 lbs. Pressures and temperatures don't correlate directly with chart given. Found head pressure and subcooling slightly high and suction pressure slightly low. May need to look at further if customer notices any performance issues."

 

I am willing to bet you have low refrigerant again due to leaks in the evaporator coil.

 

Possibly, although they did just fix the leak and recharge it last week, and it's being doing the same thing pretty much immediately afterwards so I'm guessing there might be another issue. I'll definitely be getting them to come out again and look at it, I just wanted to get some opinions here first before spending more money on it. I don't think there are leaks in the coil as far as I know...the tech didn't spot any apart from the service valve, and I couldn't see any with a UV flashlight (although it's hard to see around the back of it).

 

Just a thought, but I had a Ruud H/P that did the same thing right out of the box. Turns out that the distributor tube to the section freezing up was restricted from the factory! My wholesaler provided a new unit and worked fine. First and only time I have seen this but it is a thought.

 

I suspect a restriction in our case is a definite possibility, due to the previous R22 air handler.

 

Didn’t they remove the factory drier inside the OD unit, in all the times this system was open?

Did you register the OD unit initially and also the Air Handler, manufacturers warranty decreases significantly if never registered.

If your positive there is no leak, may want to clear OD coil completely of ice, then run and check the OD metering device if they feel it’s OK, recover charge, remove factory drier ( if still in there ) blow nitrogen all which way while system is open, if possible elevate OD unit off pad. Install new drier by Air Handler, evac. and weight in virgin refrigerant, ( Goodman recommended charging method in winter ) taking into consideration refrigerant charge for every feet over a certain amount. Says 109 ounces charged from factory, adding a small amount per foot if lineset over a predetermined amount.

They don’t make the GSZ13 anymore but did find a older flyer for the GSZ13 and it showed the AVPTC24 as a match. Yours is a AVPTC25

https://www.ecomfort.com/manuals/GSZ13Specs.pdf Dated 2014

Page 20 shows the match for the GSZ13018 and AVPTC24 gives 14 SEER, 12 EER, and 7.70 HSPF. Yours is the AVPTC25 thou.

 

Didn’t they remove the factory drier inside the OD unit, in all the times this system was open?

Yes, they replaced the filter drier in the outside unit when they installed the new air handler last year.

 

After finally being able to manually defrost the coils now that the weather has warmed up, here are pics of how it's frosting up. The first image is before the automatic 30 min defrost, and the second is after. You can see the frost is appearing in bands, and the defrost isn't fully getting rid of it at the bottom.

I think it only gets to the state of the entire bottom half being frosted up if it's been left for a few days without me manually getting rid of the accumulated frost.

Is this banding normal, or does it usually indicate either low refrigerant or some distribution problem?

 

It could be low on refrigerant, or it could be a flow/distribution issue.

 

I will second this, seen this a lot when charge is low.

 

Defrost sensor could be in the wrong location.

 

I always put brick pavers between the unit and the pad. This helps keep the brain holes from becomeing closed up. When cleaning the outdoor unit this also lets the water run out better.

 

Ok, thanks for all the info. Tech is coming out next week to look into it further.

I did notice something odd today when monitoring the heat register temps. After icing up for half a day, I was getting a 10 degree F rise 15 mins after auto defrost, and after defrosting the coils manually I was then getting a 20 degree rise (outdoor temp 35F). I assumed the coils had lost 50% efficiency due to the frost, but then after the next auto defrost the temp rise was back down to 10F after 15 mins.

So I'm wondering if a partially stuck reversing valve is causing the problem, and it got unstuck when I used hot water to defrost it inside. We were told a year ago that we had a sticking RV, but they didn't replace it (different company)...they just cycled it a number of times which they said fixed it. Since then it has been cleaned (when new air handler installed).

It's the kind of thing that might be difficult to track down, if it only happens during cold weather and it's warm when they come to diagnose the problem, so I'm just wondering if this is a possible explanation. Do faulty RVs sometimes stick in cold weather?

 

So when you completely defrosted the complete OD coil 100% with hot water, inside and outside of the coil did you notice if the distributor tubes just after the OD metering device start frosting immediately or soon after?

This thread keeps going and going around in circles. There are ways to somewhat check temperature difference on the RV to see if it’s shifting 100%.

Again look at the OD metering device for immediate frosting ( heat mode ) with a 100% clear coil and the distributor tubes.

IMO, recover total charge and weight it. Then compare to what the weight in method that Goodman wants in winter mode to charge your system by. Replace drier again, add virgin refrigerant. Looks to be 109 ounces, which includes a matching ID coil and 15’ of properly sized lineset ( should also say the amount on the OD unit electrical data plate ) for the 1.5 ton OD unit, and for every feet above 15’ of 3/8” OD liquid line you add a certain amount of refrigerant. Elevate the OD unit if possible. Then take it from there. While system is open blow nitrogen all which way but loose, ( providing OD metering device checks out OK worst case they can remove OD metering device while blowing nitrogen all which way ) Then observe the before ( your unit operation now ) and after, the distributor tubes downstream of the metering device.

They can simulate to start a defrost and let the machine terminate by itself to make sure the defrost is working, typically you’ll see the water mist vapor elevating off the unit while in defrost if it’s wet. Also you can check the supply air temperature coming out of the closet supply vent nearest the Air Handler ( if no electric heat coming on ) the air temperature should get quite cold, while in defrost.

Goodman wants the weight method for charging in winter time, not add a little refrigerant, remove a little refrigerant ,check pressures, check superheat, subcooling or whatever. It says that in the install guide.

Anyway that’s my:.02: worth.

 

With the coils partially frosted up after the auto defrost, the capillaries from the TXV frost up within about 5 mins. Will check with a clean coil.

No steam comes off after the auto defrost, but that did happen last year. It's set to 30 mins with delay, and defrost doesn't last long...30 sec delay with compressor off, then compressor comes on for 10 or 20 secs, then off for 30 secs delay. It is doing something, as it completely removes all frost from the top half of the coil. I'm guessing it doesn't take long for the coil to reach the 75F termination temp. Temp sensor was apparently checked recently and found ok.

Last year I think defrost was set to 60 min with no delay, and the defrost lasted about 5 mins with steam coming off at the end.

Pouring hot water on the RV without cleaning the coils didn't seem to have any effect on heat output. I'll just mention it to the tech and he can check it properly. It was definitely sticking a year ago, and then some dirt cleaned out of it when the new air handler was installed.

 

Gave you the factory charge for a 1.5 ton OD unit as you mentioned you had a 1.5 ton OD unit in post 13 . In post 17 says you have a 2 ton OD unit, which shows to have a factory charge of 99 ounces. Can verify at the OD unit data plate.

You sure they reinstalled the defrost sensor in the exact position as the factory, as I’m guessing it was removed to avoid damage when they removed the factory drier, which is in the same area no?

Maybe try going back to no delay ( non smart shift ) for test purposes, and even the 60 minutes once you figure out your issues. I only ever seen once a restricted OD coil, and that was due to ( R22 unit ) the rookie tech soldered ( soft soldered ) the lineset with so much solder that beads or small pellets of solder got lodged into the distribution tubes. Defrost of OD coil was always incomplete, even in regular heat the air temperatures where a little cooler than normal. Had Technical Rep. come out and after some temperature and pressure readings they condemned the coil. They gave the person a new OD unit, once installed system worked fine.

When the OD coil is completely clear do you get good air temperatures out of your supply registers on a day like 35+ degree OD temperature? I suppose if the system been like this from day one that may be a hard question to answer as you don’t have anything to compare it to.

 

Defrost should normally take 5-10 minutes unless it is above 30 ODA. Your post about a 30 sec delay then the compressor on for 30, then another delay does not sound like a proper defrost sequence. As said it may have to do with sensor location.

As for the Rev. Valve, if it is only a problem shifting when it is cold there is a simple fix which can help it so it shifts better. It could also be that the RV is a symptom and not an issue. Being low on refrigerant can cause shift issues with RV's, not enough pressure to get it to move.

A knowledgeable tech should be able to diagnose this for you fairly quickly.

 

Yes, I see we have a 2 ton unit. Our home inspection said it was 1.5 ton, but I guess he was just guessing. That's good to know...I was thinking 1.5 ton was undersized for our home and it should be 2 ton.

After I manually defrost the coils I get a 20F rise out of the registers (over indoor temp) when it's about 35F outside. After a few hours that drops to about 10F rise (measuring 20 mins after auto defrost so compressor has been running for 20 mins). System works great when it doesn't have to defrost.

I'm thinking that Goodman wouldn't design SmartShift so that it doesn't properly defrost, so there must be some other problem. A more experienced tech (the one who installed the new air handler) is coming to look at it, so I'll let him figure it out. I think he turned the SmartShift to reduce noise when he installed the compressor blanket for us, so he'll have a better idea if the sensor is in the right position.

The capillaries from the metering device frost up within 10 mins of starting after I completely defrost the coils, and there is also light frost banding on the coils after 10 mins. Not sure if that has any bearing. This is with 97% humidity and 38F outdoor temp.

Also, it does seem to be working fine (normal heat register temps) after the first auto defrost...I think it's just gradually losing performance when it frosts up more and more. So main issue is just not defrosting (which may be due to a restriction or being low on charge).

 

At 38 I wouldn't think it would have to defrost much. Also at 10 minutes run time it should be pretty well stabilized which would minimize any banding with correct charge/flow.

 

Well there isn't much frost after 10 mins...just a very thin layer (< 1mm), enough for the capillaries and the bands on the coils to have a white sheen. Seems to take about 20 mins for temperature at heat registers to reach peak after a defrost cycle.

 

Looking inside, I can see that the banding is starting from the capilliary tubes coming out of the metering device. The capillaries are all heavily iced up and the defrost cycle doesn't remove the frost. The feeder tube (where I think the defrost termination temperature sensor is located) only frosts up a little bit and the defrost cycle removes that frost.

So, am I correct in thinking that if everything is working normally it shouldn't be icing up from the capillaries? Given that the temp sensor is on the opposite end of the coil, I'm guessing that the lowest temperature should normally be where the vapor exits the evaporator coil, and if it's icing up at the capilliaries it means the pressure is too low going into the coil due to low charge or restriction. Does that sound about right?

 

That’s the beginning g of absorbing heat so the beginning of frosting!

 

https://www.google.com/url?sa=t&rct...n+the+A_GSZ13024+and+A_GSZ13030+Heat+Pumps.pdf&usg=AOvVaw1Vm6HlO15MQTt8xxTcLmce

Dated 2010 thou, seen another Technical Bullitin for the GSZ14 dated 2019. Does say to elevate unit. Doubt this applies to you unless your OD unit was dated somewhere around 2010.

Maybe have the company call Technical Service to see what they say, granted your unit could just need the correct refrigerant charge.

 

The model I have (which is later), the drain holes under the coils are elevated a few inches above the bottom of the unit, so even if it's sitting on a slab the drain holes will be a few inches above. Installer doesn't think it is an issue, but I'd probably look into elevating it if I installed a new unit.

 

Nothing on top of the filter drier. Underneath there is liquid, but I can see it's coming from frost melting off the capilliaries, and it just glows purple under UV light.

 

Nothing on top of the filter drier. Underneath there is liquid, but I can see it's coming from frost melting off the capilliaries, and it just glows purple under UV light.

You have a UV light?

Purple?

 

I had a suspicion that it was low on refrigerant.

I think I would have them isolate the outside unit for the inside unit and pressurize both with N2, then come back in 24 hours to see which one lost pressure.

 

You might want to think about changing the outside unit if there is nothing wrong with it. Find that leak and repair it. Put in the right amount of frig and start over. i would blow out the line set both ways if there is a blockage.

 

Yeah, I understand, but it's really a bet that it isn't in the coil, and right now that's looking like the most plausible location. If we replace the coil we may as well just spend some more money to replace the entire unit and get the 10 year warranty. Other benefits are much reduced noise if we get one with a 2 stage compressor.

There has been extreme ice buildup over the coils in the past (about 6 inches thick and 24 inches high) surrounding the coils and fins, which likely isn't good. Right now I just don't trust the unit, and am inclined not to spend the money on the pressure test and recharge.