Mini split - freezing capillary tube

[Th.Mihov]

Hi everyone.
I'll try to be short:
Before some time ,i bought a used mini-split system,with "electrical problem",at least that said the man from which i bought it.But when i went to get it ,he said also that not long ago,the conditioner was out of refrigerant,and some guy "repaired" it quite cheap.After i bought it and went home,while i was preparing for recovering the refrigerant,i saw that there was a leak from the two valves of the outdoor unit,so i have recovered the refrigerant (there was about 300gr (R407C) of 870gr as pointed by the nameplate) and changed the valves.I was purging nitrogen while i was brazing the new valves,so that no oxidation scales can form and possibly plug the cap tubes.After that i performed few times a nitrogen blow from both valves and evacuated to 150 microns (it took about 7-8 hours).Then i charged it to the name plate's charge with scale and ran it - everything was fine-it was heating the room fine,the current consumption was OK-perfect.Then i left it for few days,and when i had some free time,i decided to test it again,and the problem arised - when i turned it on in heating mode,there is no heat.In heating mode,the thicker pipe should be hot (it's the discharge line - hot gas),but it wasn't.I've checked the pressure-only 15 barg (normally it should go to 20-25barg,according to the indoor/outdoor conditions).Also when i stripped the covers ,i found that the common cap tube's end(there are two cap tubes and one check valve-in heating mode,the refrigerant passes trough both of them,and in cooling,it passes only trough the one of them (common) and the other is bypassed by the check valve) was frosted together with 10-15cm of the pipe after it.Since the charge was right,the only two options for this scenario are plugged cap tube or moisture which freezes inside the cap tube.Since checking for freezing moisture was easier ,i grabbed the hot-air gun and heated the cap tube and the frosted pipe,and voila-everything went fine again...until the next long period of not using the conditioner,and the outside temperature drops about 0deg C.Few days after,i unbrazed the cap tubes,the outdoor heat exchanger and the tubings and blown them with nitrogen,than evacuated to 150 microns again (3-4 hours),charged and nothing changed.Then i tried to evacuate again,but this time i was heating the compressor's bottom with hot-air gun to make the evaporation of moisture of the oil easier (at 150 microns ,the water needs to be about 50deg C to boil),but still no luck.
Finally i added a filter drier and sight glass with moisture indicator at the liquid line ,but again that doesn't solve the problem - the sight glass is green i.e no moisture,but the problem is still present.
My guess is that once the system was out of refrigerant,it was left open for too long for some reason,and moisture was left in and since it's a R407C system and uses POE oil ,which is hygroscopic,now the moisture is in the oil,and it cant get out.
My last hope was the filter drier,but since it doesn't solve the problem,i'm out of suggestions,and i need advice.
Any help would be appreciated.
Thanks in advance.

[frank]

R407C and moisture turn into a black sludge. With the unit being open to atmosphere for a long period, I would suspect that your compressor oil is compromised resulting in the cap tube blockages you are experiencing

[Rob White]

Hi everyone.
I'll try to be short:
Before some time ,i bought a used mini-split system,with "electrical problem",at least that said the man from which i bought it.But when i went to get it ,he said also that not long ago,the conditioner was out of refrigerant,

Snip


My last hope was the filter drier,but since it doesn't solve the problem,i'm out of suggestions,and i need advice.
Any help would be appreciated.
Thanks in advance.

Was the dryer Bi-directional?

Heat-pump dryers are both directions unlike normal dryers which are only one direction.

I think the issue is the oil and it could be contaminated.

If the dryer is not Bi-directional, when the system reverses any debris inside it is blown back out.

If it is Bi-directional then look to the oil.

There is no easy fix here because if the oil is contaminated it will have to be removed, the circuit flushed, cleaned and dried before new oil is added.

Rob.

Ps
One last thought did you use virgin refrigerant to put in it or was it reclaimed refrigerant?

.

[Th.Mihov]

Firstly,thanks for the replay!
Sounds completely plausible,but seems strange to me that,the problem only occurs while the system was off for long time.When i start it after such period,it shows the symptoms i described earlier,then if i left the system operating for say about 5-10 minutes,and switch on and off the 4-way valve for few seconds,so that the somehow warmer refrigerant can pass trough the cap tube,the problem is gone until the next long pause.Even in defrost cycle,when the refrigerant change it's direction again,the problem doesn't show.
Could it be deu to the density of the oil or the sludge when there is cold outside ? - Than it's very dense and acts like a "plug"?
Changing the oil seems good idea,but it's hardly achievable in that case ,because it's a rotary compressor,and the discharge port is on the top of the compressor,but the end of the discharge pipe which is inside the compressor is lower than the shell,and so when the compressor is turned upside-down ,not all oil can flow out.A while ago i was trying this with burnt compressor,just to see how much oil i can drain that way,and when it stayed upside down for few hours,than i cut it open and drained the remaining oil.I don't remember exact numbers,but with turning it i was able to drain about 2/3 of the oil.Is that enough for practical purpose,or it should be drained completely?.If it must be drained completely,than that's bad because the only way to do it is to drill a hole at the bottom of the compressor and then weld it back.Technically it is not problem for me-i can drill,weld,etc ,but i don't think it's a good idea,since 100% there will be at least burnt oil from the heat of the weld and possibly metal chips.
So is it acceptable to change only 2/3's of the oil + flushing the coils or not?

[Th.Mihov]

Rob White,no the filter dried is not bi directional.It was saturday or sunday,when i decided to mount it,and it was the only one i got.I know that for heatpumps it must be bidirectional,but as i suspect moisture i rather use it as a drier,than a filter.I think if the problem was a debris,in defrost cycle it should plug back,but it is not.Even more-as i said,i had unbrazed cap tubes,coil,and inside tubings (everything except the 4-way valve) and blown them with nitrogen at 20 barg.If everything was there,after that it should not be anymore,or at least i hope so.
The refrigerant is recovered by me from one of my other air conditioners which is running fine,because now in my country everything got too expensive (electricity,bills,even the food...) for very short period of time,and i'm not okay to spend XX euros,just for testing,given that i already have all necessary to test it without expenses.I always recover with filter drier,and recharge trough a filter drier,if i need to put it back in.

[Rob White]

.

I would swap the dryer out and put a new one on. Any moisture in it will be removed with the other dryer.

I have seen a glorified syringe with a long thin plastic tube in that you put down the suction pipe.
Obviously you would have to un-braze the suction. You put the thin pipe into the comps sump and
suck the oil out.

Add new oil and run the compressor for a few hours to warm the compressor through and then repeat
the process.

I have removed and changed the oil in a hermetic scroll compressor a few times and it is a real pain.
If you do decide to remove the oil then you will have to do it more than once.

I would be tempted to blow nitrogen (OFN) through the circuit. Un-braze the capillary's and blow OFN
through so it blows back through the circuit and can blow oil, crap, dust and once even a mouse out
That is another story

keep safe.
Rob

.

[Th.Mihov]

Yep,seems like the problem really is not with moisture in the system,but with the sludge.Yesterday i have changed the FD,because the old was crappy no-name ****.I have changed it with molecular one,but after 10 hour running,still no change - solid green indicator,but still couldn't "start" this morning,although the temperature last night,was always above 0C here.
My guess is that when i heat the cap tube with the hot-air gun,or by the 4-way valve,i just make the "wax" in it more liquid,so the pressure can pull it out,but after stopping the machine and everything cools down,the wax comes hard again and plugs the cap tube.
So obviously i need to say goodbye to the old oil...
Rob White,thanks for the advices about multiple oil changes-that way ,really everything will be changed,but in my case there are few hurdles in changing oil with the "syringe method" you have mentioned.
As i said,it's a rotary compressor,so it's kinda upside-downed scroll - the pumping part is at the bottom of the shell,and the motor is pointing to the top of the shell , in contrast with the scroll where the pumping part is on top and the motor is pointing down.For that reason,and because the discharge port is at the center on the top of the shell,if you put a thin hose,tube or something else,it will reach to the rotor of the motor,it's not going to the sump.In the scrolls ,or at least in Copeland's scrolls i have dealed with,the suction is closer to the sump,so it really sounds more realistic to reach it with something like thin fairly hard tube.Reaching the sump of the rotary trough the suction also sounds impossible since,firstly there's an accumulator on the suction,but even if there wasn't ,the suction is going directly in the compression chamber ,so it's nothing common between the suction and the sump.I'll pour out as much as i can by turning it upside down,and than i'll lay it by it's side with some incline to the discharge port,and then with some almost L-shaped tube,i'll try to suck it out with the vacuum pump and and empty jar or something similar.If i succeed ,i'll post pictures-who knows,may be it'll be useful for someone,someday
The info that i can't find is about the viscosity of the oil.Somewhere on the web i found document from the same manufacturer (Midea),about oils used in their other R407C models and it says RB68AF,which i suspect means that is ISO68 oil .Any suggestions?
As for the nitrogen - i already have unbrazed and blown everything with it,but i'll do it again when i'm changing the oil ,just to be sure
P.S - Excuse my bad english,but i rarely need to write.
P.S2 - What OFN stands for? (shamed).It's clear that it's a slang,i googled it but can't find something that sounds plausible?

[Jeff Winsor]

OFN Oxygen Free Nitrogen

[Rob White]

.


As above

It is basically pure Nitrogen

Rob

.

[Th.Mihov]

Got it
Any advice about the viscosity of the oil,or i should take the trial and error path?

[Rob White]

.

For Copland scrolls but in supermarket packs not AC I always used
the Mobil Artic range.

Mobil Artic 32 and others.

See

https://www.orionairsales.co.uk/mobi...um-11597-p.asp

There are two main types one for low temp applications and one for higher temp applications.

Hope this helps.

Rob

.

[Th.Mihov]

I'm back.
Today i had some free time,so i decided to try to drain the oil.Actually i started yesterday by hooking the compressor to a hanger on my ceiling in the garage with the discharge point down.

After hanging for (about) a day,i got 300ml oil in the container.There were about 20-30ml in the accumulator,and went to the floor when i was first turning the compressor.
Than i started to dig in my bench for something suitable to fit in the discharge port (3/8").Then i saw a clear plastic 6mm hose.It's perfect since it's very flexible and in the same time a little hard.As a bonus,i can see what's happening.It's used in the food industry and it's easy to find on the markets.
Another plastic tube is needed too,but with 10mm diameter.
The next "special tool" is a jar with a metal threaded cap.
The setup looks like that-one end of the 6mm hose goes in the compressor's discharge port,the other end goes in 5.5mm hole, drilled in the jar's cap.One end of the 10mm hose goes in a 9.5mm hole, drilled in the jar's cap,in the other end i have put a standard 1/4" schrader valve with longer stem.After that it's connected to the vacuum pump with standard 1/4" hose.

The tricky part is putting the hose in the compressor right.After pulling it in,it bumps into the rotor (or maybe the winding-who knows) and starts to bend back so it forms a turned down U-shape,so it can reach to the point where the oil was collected.After it's bent it's all about pulling a little in or out and twisting the hose with fingers while the vacuum pump is running so it can be seen if we are on the right place and the oil was sucked out in the jar.

With that method i've managed to get another 80ml of oil,so the whole quantity sucked out is 400ml.The total amount in the compressor is 480ml.It turns out ,that i've managed to get about 85% of the oil out,so i'm happy with that number .I guess another 20 to 50ml are in the coils,so IMHO it can be said that practically the whole amount is out.
By the way,the oil looks very brownish.Here's a sample.

So to summarize,for me it's a working method,albeit kind'a rednekish,but atleast i't doesn't require any special tools,it's easy,and it's cheap.
After few oil changes (as Rob White suggested),i think the old oil would be completely gone
While digging in my bench's depths ,i've found a bottle with a viscousier (ISO 22) POE oil,which i think to use for the first change,and than i'll search for the right oil (again thank for the link Rob White .I'll take a look).
I'll report for any further progress.

[frank]

Be careful when you are charging the new POE oil. It is VERY hygroscopic and will absorb lots of moisture - exactly as you have found due to the compressor being open for a long time.

If you don't use fresh oil into a clean tight closed system you will be back to square one with contaminated oil.

Good effort so far.

[Rob White]

.

Your doing a great job

With the amount of time and effort you have invested in this you deserve a win

If you have gone to the trouble of removing the compressor I would be tempted
to grind the paint off low down on the sump and drill a small hole.
I would then silver solder in a brass 1/4 straight flare connector in the hole.

I have done this before and it is a way to get a connection into the sump. Just
paint it with black paint and you then have a 1/4 connector in the sump to drain the comp oil.


Rob

Ps I seem to be telling you all my bodge secrets

.

[Th.Mihov]

Thanks for the kind words .
I'm that type of guy-if i decide that something should happen,it WILL happen,no matter how hard it will be.I hate the word impossible
frank,i cant remember when i bought it,neither why,but it's clear,almost full and it's tightly closed,but after i read your post,i realized that really it's not worth risking to ruin the efforts so far,so i'll use brand new even for the "flushing".
Rob White,sounds neat,but (for now ) i prefer not to do it,since i'm afraid,that a chip could enter in the sump,and block the oil inlet,although i have done it before (with a vessel,not with compressor,but was also important not to let chips inside) by slightly pressurizing the vessel,pointing down the spot to be drilled,and just for prevention - adding few strong magnets near the drill bit,to hold the chips and everything went good,but for now i'll skip this
I know it depends on the humidity,temperature,etc. but just of curiosity-(indicatively)how long is too-long for POE oil to stay in contact with air?
After more reading and googling,i bought Suniso ISO68 oil,and if i have few free hours tomorrow,i'll put in the new oil,blow everything with nitrogen again ,put back everything together and do a test run,and of course i'll report

[Th.Mihov]

I think the things are going in the right direction
Today,i was unable to complete everything,that i mentioned in the last post,but atleast i succeeded to blow everything with N2 again,pour in the new oil,pressure test everything and evacuate to 150 microns.The good news is that i achieved that vacuum level for less than 5 minutes in contrast with the previous 7-8 hours

[Rob White]

.

When it comes to POE oil I was told and always work to instant.

use the correct size oil drum, 500ml, 1ltr, 5ltr and such and as soon as
it is open transfer it. Any left over oil will need to be treated as waste.

Don't think you can seal the container up, you can't soon as air is in connect
it is ruined. In 24 hours the oil has gone.

Rob

.

[Th.Mihov]

Still no luck
Yesterday i bought and charged with fresh refrigerant,tested-everything was ok.
Left it off overnight and this morning - same as before....
Any advices?
Before some time,i was wondering if i could replace the cap tube with TXV for better performance?.Is it worth to do it ,i mean will there be a significant gain in performance or not? In my opinion it should be,since as we know cap tubes have it's peak performance only for one set of pressures,and from there on is a compromise with performance because of the cheapness of the tube vs the TXV.After all,caps are used in these cheap mini splits,because of there simplicity and low price,not because their efficiency.
I'm also familiar that r134a cap tube systems with POE oil being clogged,and the problem is sometimes solved by replacing the cap with TXV.I know that the problem is not caused by the cap tube itself but for example a dirty condenser and anything else that cause high head temperature,which destroys the oil...

[Th.Mihov]

Good morning
It's me and my haste again...
After my last post,i was annoyed and i decided to leave everything for a while.
Today i got some free time and i decided to check what's happening with the AC.
Turned it on,and i get to put in order my bench,and surprisingly after 4-5 minutes ,hot air started to blow
Last time i let it ran for about 2-3 minutes,before i judge that is not working since my other AC heats up the indoor coil for that time.And as a result from my frustration,i even don't get my gauges hooked up to see what's happening....Not a news,but i'm very impatient person
Obviously it just takes little longer to start heating,but it's working fine
Now here it's -2C ,and it's misty - perfect time for performance tests - cold and humid.
I'll let it run for few hours ,and see what will happen.
Again - thanks for the advices - i really appreciate that .
P.S - Any suggestions about the changeover from cap to TXV - to try it or not?

[Brian_UK]

Just a thought on changing to a TXV.

You will need to fit two valves and two non-return valves because it is a heat pump with bi-directional flow, increased labour and costs.

[Th.Mihov]

Brian_UK,thanks.I'm aware of that .But i'm still not sure about the end result.