Catastrope! ...And Complete Evac & charge procedure Q's for Mini-Split

[whatismouse]

Prior to my mini-split install I watched numerous videos and read many threads to familiarize myself with the tools and the process. Install was successful (seemingly) and had been running for about 24 hours. I thought to check my four lineset connections one last time before buttoning up things. While slightly moving the 1/4" line at the connection point near the indoor unit I heard a leak. Mind you the whole system had held a vacuum prior to releasing any refrigerant, and I had leak tested all connections with soap bubble solution while on heating and cooling mode. (I did not do a nitrogen pressure test) But now there was a definite leak, that I could hear more or less depending on how I wiggled the line. Everything I had read had addressed how to deal with leaks prior to refrigerant being in the system, or perhaps how to correct slow leaks (where you could pump-down system, fix leak, add back proper amt of refrigerant). I did not know what to do with a significant leak on a pressurized system. I decided to torque it down, thinking that perhaps it was just under-torqued, increasing torque from 11 ft-lbs to 13 didnt seem to work, and in fact the leak was getting worse from the manipulation. (I had turned system off upon first noticing the leak). And then *POP*, the flare broke off completely releasing all refrigerant. At this point, all i could think to do was the close the high and low side valves at the outdoor unit, cap off and seal both ends of the open copper, leave the toxic vicinity, and rack my brain as to what went wrong.

My conclusion is that I must have either forgotten to, or applied an inadequate amount of nylog on the outside of the flare. Then when torquing it down to spec, the nut was grabbing the flare and twisting it. My attempt to tighten the connection to stop the obvious leak was the final straw that broke the flare clean off. That or the factory copper lineset and/or flare was faulty/poor quality.

It's quite likely and worth noting that even though the issue passed my vacuum test and initial leak detecter tests, It most likely would have failed and been noticed by a proper nitrogen pressure test, avoiding this disaster.

My initial response was to throw my hands in the air, and hire an HVAC professional to re-flare, re-connect, pressure test, vacuum, and re-charge.

My secondary response was to do more research, and figure out how to fix this problem myself. I am generally pretty precise about things and don't make mistakes like this often. I have no reason to believe that I'm just totally inept and shouldn't touch my system any more that I already have. Or maybe I'm just stubborn and want to fix what I mucked up.

In any case I've begun to familiarize myself with the process of adding refrigerant, and weighing it in. (after re-doing the flare, and evac-ing the system again)

My little questions regard the valves of the outdoor unit, and whether they should both by opened now when I draw a vacuum again. (previously the valves were closed when vacuuming because the ODU was full of refrigerant and I was only to vacuum the IDU & lineset. But now there is no refrigerant in the system, so I imagine it is proper to draw a vacuum on the ODU as well, by having both high and low valves open. I just want to make sure that there is no issue drawing a vacuum on the ODU, and that this is proper procedure for an empty system.

The second question is, while adding refrigerant, would I leave both high and low valves open, or perhaps close the low side so no liquid r410a enters the outdoor unit? (when I say high and low valves I'm referring to the outdoor unit valves, and not the high and low pressure guages on the AC guage set) Most youtubes concerning this step are in reference to a system where the outdoor unit is factory charged, and thus valves remain closed, extra refrigerant is added to the lineset to account for longer lines, and then valves are opened to release factory refrigerant. In my case the factory refrigerant has all been released to the atmosphere, and thus I'm charging the entire system with external r410a.

I much appreciate any expertise and advice!

 

[chrispyny]

First, i am not an experienced a/c tech. But i did stay at a holiday inn expressonce. You need a precise amount of refrigerant in your minisplit. Not knowing how much is in the unit, you need to evacuate EVERYTHING. Linesets and condenser. Then add the appropriate amount freon by weight.

2nd question. Even though you are adding refridgerant by weight, Your low and high hoses from gauges will be hooked up to both high and low ports on the minisplit condenser. As such, everything should,be open.
You are forgetting one step. Since you shouldn’t add liquid freon to a split, you need the inline adaptor (kwik charge, https://www.trutechtools.com/535-C?gclid=EAIaIQobChMIlv_Jt7ja6wIVQ77ACh13KguhEAQYAyABEgLrVvD_BwE ) which gasses out the liquid freon prior to entering the system. Adding liquid to the system,will kill the compressor. Compressors compress gas, not liquids.

This help?

 

[chrispyny]

P.s. when adding freon(410a) you quickly and for a very short time ‘crack’ the tank valve, adding freon in small bursts to allow the freon to gas,off in the qkwik charge.

You will likely need to run the compressor to pull in all the required freon. That is when this is done.

 

[whatismouse]

OK so no issues with drawing a vacuum on the entire system, with both valves open.

As for adding refrigerant, there is only one service port on the low side. If I were to close the low side valve (to the condenser), then refrigerant would travel through the low side line to the air handler unit, through the high side line, and ultimately back to the condenser. I imagine it would be a gas well before it even reached the air handler, maybe even before it gets through the ac guages to the the service port? In which case there may be no need for a vaporizing attachment? Especially if I charge slowly and give time for equalizing/distributing as you say? Of course I have no idea if this is how its done, or even if it would be harmful to have liquid refrigerant enter the low side lines...

And as you mention, there is the next question of how much refrigerant can be charged with the unit off. I need 30oz total. I imagine I can charge as much as will be accepted, and then I can follow standard procedure for how to add additional refrigerant while system is in cooling mode. (obviously with both valves open at this point). Already have a scale that will work for weighing in total amount.

 

[brihvac]

You cracked the flare over tightening. The right charge is important but you can be a little off. If you are not going to weigh the charge in and add for the extra length in line I like to set it by indoor coil temp. I do HVAC and the proper way is to weigh the charge in but if I do not do that I have a digital gauge set with probes and look to have a 46-48 degree indoor coil and that is usually around 130-140 suction pressure. I just did a Mr. Cool in my garage this weekend and that was actually the factory charge that was in it for a 16' line set. 48 degree coil and pressure was 139 on a 75 degree day

 

[BillK]

Catastrophe ????? I was expecting to hear that the entire unit exploded in a ball of flames ! Just a minor oops. Fix it and get on with life.

 

[Jim greengo]

Never trust a vacuam pump for a leak test,always test with nitrogen.

 

[LS6 Tommy]

You cracked the flare over tightening.

X2. The nylog acted as a lube and threw off the torque readings, allowing you to over torque the fitting.

Tommy

 

[whatismouse]

X2. The nylog acted as a lube and threw off the torque readings, allowing you to over torque the fitting.

Tommy

Wouldn't this be the case if I were to use nylog on the actual threads? Which I did not do. I don't see how nylog on the flare itself would throw off the torque wrench?

 

[jjrbus]

The videos are a blessing and a curse. Some people can make big screen quality videos but not have a clue about the subject. Many of the videos are from R 12 days when lower psi was much more forgiving.

I had a Daikin install kit which contained a flair gauge, several of the pre made line sets, the flairs failed the go no go gauge! The nuts on the premade sets are too light duty for 600 psi.

 

[Terry D]

X2. The nylog acted as a lube and threw off the torque readings, allowing you to over torque the fitting.

Tommy

I agree.

 

[Terry D]

Wouldn't this be the case if I were to use nylog on the actual threads? Which I did not do. I don't see how nylog on the flare itself would throw off the torque wrench?

Never put Nylog on the outside of the flare and never on the threads. You put it between the top of the flare and the nut, which lubricated that surface, which made it spin easier throwing off the torque. For that thing to have broken off completely, it was way over torqued. Also check your torque wrench and possibly a burr inside the nut

 

[BillK]

Just curious, What type of torque wrench did you use that would work on that type of fitting ? I am sure that one exists I just have never seen one.

 

[whatismouse]

Never put Nylog on the outside of the flare and never on the threads. You put it between the top of the flare and the nut, which lubricated that surface, which made it spin easier throwing off the torque. For that thing to have broken off completely, it was way over torqued. Also check your torque wrench and possibly a burr inside the nut

it says right in the manual to use the nylog on both sides of the flare. this should not affect the torque of the nut tightening on to the threads/compressing the flare. i believe the outside of the flare should be lubricated precisely to prevent the nut from grabbing the flare itself and twisting it, as well as to aid in sealing.

 

[Terry D]

Re: Catastrope! ...And Complete Evac & charge procedure Q's for Mini-Split

it says right in the manual to use the nylog on both sides of the flare. this should not affect the torque of the nut tightening on to the threads/compressing the flare. i believe the outside of the flare should be lubricated precisely to prevent the nut from grabbing the flare itself and twisting it, as well as to aid in sealing.

If the manual says do it, then i guess do it. To my knowledge, a small amout of Nylog is ok to the mating surfaces to the flare. That is the only place a flare seals. The threads do not seal a flare nor does trying to seal the outside of the flare. If it is leaking past where the 2 inside surfaces meet, it is a faulty flare, and you will never stop it at that pressure. You either had a faulty flare from the factory, which is possible, or your torque wrench is not calibrated. Is it a actual HVAC wrench, or a wrench with a crows foot on it. You can use a crows foot, but it has to be at 90 degrees from the wrench. Not straight out, this will lengthen the wrench, and give false readings .

To me, putting Nylog on the ouside of the flare is about the same as putting it on the threads. It is lubricating it and reducing resistance, thus making it easier to tighten, giving you a false torque reading

A properly made flare will be pretty close to the same ID as the nut. But the nut should still spin freely. And the flare should be seated fully in the nut. Hopefully you checked that before putting it together

What is the brand and model number of this mini split. Also post a picture of your torque wrench

Sent from my SM-G960U using The Garage Journal mobile app

 

[whatismouse]

that's true that the main seal is between the inside of the flare and the face which it seats on. Small and slow leaks that make it through here still have to travel either through the threads and out or around the back side of the flare and through the nut. Potentially nylog in both of these places could reduce the speed of infinitesimal leaks. I imagine at this level of persnickipicality, even every professional would have a different opinion and method about it. If torquing to spec wasn't made more difficult, I would presume that it would be preferable to use nylog in all three places.

I used a regular torque wrench with a crows foot at 90*. I checked the flare and the nut for proper mating & spinning prior to installation, but perhaps not super thoroughly. I think the most likely explanation here is user error, and over-torquing somehow.

It is a Pioneer 19 Seer 9000btu. My lineset is about 18 ft long, I cut them down from the original 25 ft. The two flares that I made myself are connected to the ODU, the one that failed was a factory flare on the 1/4" line connecting outside the IDU.

 

[BillK]

I think the most likely explanation here is user error, and over-torquing somehow.

What is the range of your torque wrench ? One of the biggest mistakes I see is guys trying to use a 250 lbft wrench to torque a 15 lbft fastener. It just will not be accurate. In your case I would be using a torque wrench with a range to about 50 lb ft. I always like to be close to the middle of the torque range. 15 lbft really is not much at all.

 

[whatismouse]

I used one from work when I set it at the initial 11.8 ft/lbs. Later when I noticed the leak I grabbed a cheap tekton with a range of 80 ft/lbs. I set it to about 13 and it did tighten the nut about an 1/8 of a turn, exacerbating the issue immediately before the flare blew off.

 

[brewchief]

The flares that come on the lineset should always be cut off, they are often not that great and more often then not the flare nuts that come with the unit are a bit better.

I had a Daikin class a couple of years ago and there were 2 factory reps there, they said over 90% of the problems they saw were either with the flare connection or caused by sealant of some type added to try to make the flare connection seal(nylog,leak lock, etc).

Sent from my SM-G965U using Tapatalk

 

[LS6 Tommy]

Wouldn't this be the case if I were to use nylog on the actual threads? Which I did not do. I don't see how nylog on the flare itself would throw off the torque wrench?

Because it reduced the friction between the flare nut and the copper on the backside of the flare. The point of using a torque wrench on a flare fitting is to establish the proper "crush" on the flared copper.

Tommy

 

[whatismouse]

I wonder if the manufacturer (pioneer) accounted for the nylog in the torque settings (since they ask that you DO apply nylog on both sides of the flare). 11 ft lbs is not that much. I don't know what other manufacturers call for for the 1/4" line.

 

[Ua189ammoniatech]

Valves on the outside you mean the suction and discharge of the compressor correct
Yes you will need to open those valves for a vaccumm test and a nitrogen test if you don't have nitrogen use and inherit gas like argon, I would pressure test both high and low side to 100#psig
Upon blowing the pressure test gas if possible stop a lil above zero 1-2 psig and hook up a vaccumm pump pull the system down to 30hg if you have a vacuum gauge use it try to get to at least 500micron or lower
If you still have r410 in the system you will need to recover that "cough,cough" because r410 is a blended refrigerant and if you just top the charge of your pressure and tempature correlation will be out of wack
While charging the system with r410 it must be charged as a liquid, attach your manifold set to the jug upside down to allow liquid to flow charge the system through the discharge valve on the compressor to not allow and chances of slugging the compressor
And as for your leaking valve I've seen to many times piss poor quality control straight from the factory don't blame yourself
If you need anything else just message me

 

[motterpaul]

There is a lot of confusing information in the answers here. This video shows a pro using nylog on the threads, but he also under-torques at the minimum of the recommended range.

Pioneer does include its own version of Nylog with the unit (its in your packing slip), so I assume they include it in their specs. I just installed two new Pioneers, and we blew through a couple of flares when we were testing the torque wrench for feel. But when we installed the units we felt pretty confident of what we were doing - so far so good.

 

[Ohmthis]

Someone stated earlier that the factory flares are garbage. When I took the installer class from Mitsubishi, they were adamant about cutting off factory flares and redoing them with proper flaring tool. They also were adamant about using the flare nuts from the unit itself.

 

[machnikt]

Prior to my mini-split install I watched numerous videos and read many threads to familiarize myself with the tools and the process. Install was successful (seemingly) and had been running for about 24 hours. I thought to check my four lineset connections one last time before buttoning up things. While slightly moving the 1/4" line at the connection point near the indoor unit I heard a leak. Mind you the whole system had held a vacuum prior to releasing any refrigerant, and I had leak tested all connections with soap bubble solution while on heating and cooling mode. (I did not do a nitrogen pressure test) But now there was a definite leak, that I could hear more or less depending on how I wiggled the line. Everything I had read had addressed how to deal with leaks prior to refrigerant being in the system, or perhaps how to correct slow leaks (where you could pump-down system, fix leak, add back proper amt of refrigerant). I did not know what to do with a significant leak on a pressurized system. I decided to torque it down, thinking that perhaps it was just under-torqued, increasing torque from 11 ft-lbs to 13 didnt seem to work, and in fact the leak was getting worse from the manipulation. (I had turned system off upon first noticing the leak). And then *POP*, the flare broke off completely releasing all refrigerant. At this point, all i could think to do was the close the high and low side valves at the outdoor unit, cap off and seal both ends of the open copper, leave the toxic vicinity, and rack my brain as to what went wrong.

My conclusion is that I must have either forgotten to, or applied an inadequate amount of nylog on the outside of the flare. Then when torquing it down to spec, the nut was grabbing the flare and twisting it. My attempt to tighten the connection to stop the obvious leak was the final straw that broke the flare clean off. That or the factory copper lineset and/or flare was faulty/poor quality.

It's quite likely and worth noting that even though the issue passed my vacuum test and initial leak detecter tests, It most likely would have failed and been noticed by a proper nitrogen pressure test, avoiding this disaster.

My initial response was to throw my hands in the air, and hire an HVAC professional to re-flare, re-connect, pressure test, vacuum, and re-charge.

My secondary response was to do more research, and figure out how to fix this problem myself. I am generally pretty precise about things and don't make mistakes like this often. I have no reason to believe that I'm just totally inept and shouldn't touch my system any more that I already have. Or maybe I'm just stubborn and want to fix what I mucked up.

In any case I've begun to familiarize myself with the process of adding refrigerant, and weighing it in. (after re-doing the flare, and evac-ing the system again)

My little questions regard the valves of the outdoor unit, and whether they should both by opened now when I draw a vacuum again. (previously the valves were closed when vacuuming because the ODU was full of refrigerant and I was only to vacuum the IDU & lineset. But now there is no refrigerant in the system, so I imagine it is proper to draw a vacuum on the ODU as well, by having both high and low valves open. I just want to make sure that there is no issue drawing a vacuum on the ODU, and that this is proper procedure for an empty system.

The second question is, while adding refrigerant, would I leave both high and low valves open, or perhaps close the low side so no liquid r410a enters the outdoor unit? (when I say high and low valves I'm referring to the outdoor unit valves, and not the high and low pressure guages on the AC guage set) Most youtubes concerning this step are in reference to a system where the outdoor unit is factory charged, and thus valves remain closed, extra refrigerant is added to the lineset to account for longer lines, and then valves are opened to release factory refrigerant. In my case the factory refrigerant has all been released to the atmosphere, and thus I'm charging the entire system with external r410a.

I much appreciate any expertise and advice!

*************

Hey there, W. Were you able to fix your system? I just ran into a similar problem (Clarion, PA) and would be very grateful if you could share your experience. Is there any way I could chat with you?

*************

 

[bluedog225]

Subscribed

 

[Syberia]

To answer your original question, open both valves and pull a vacuum on the entire system. I suggest getting a 410a charging adapter which is basically just a small orifice that goes on your tank which forces the liquid refrigerant to evaporate before it enters the system.

You cannot charge a mini split by any other method besides weight.

 

[TTTTTT]

With any copper fitting, washer, flared I always torque it down, then back it off and then retorque. That how I wash showed way back and always do it that way. " crush" the fitting first . Did my Senville mini split that was precharged 6 years ago . Still working perfectly.

 

[4x4Pete]

As a tech working in the field, we use refrigeration oil on the flares. It works everytime but we also make our flares too. Wholesalers sell all kinds of gimmicky products that are for problems that don't exist or band aid a specific problem. Why would the line set come pre flared? That in itself would cast doubt in my mind. What are these lines made of that they need to spend the time and expense to flare them?