strange compressor issue?

[alan_man]

Hi, I was asked to look at a friends small split system that has been installed on his caravan (much the same way as you would usually install a split, back to back). We're both unaware of its history as he bought the caravan with the system all ready installed. It is a strangely small Mistral running R22 (needs a small 590g). The issue is it has never blown cold air, only ambient temperature (since he has owned it). After a quick look I found that it seems the compressor would turn on for a few seconds then switch back off repeatedly. I felt the discharge line cool for a fraction so assumed there was some amount of charge in it. I hooked up my R22 gauges to see if I could get any decent static reading to work with and noticed no gas discharge and 0 PSI? I know the gauges work fine as I use them regularly. I have plenty of installation experience however know little when it comes to fault finding split systems which I would like to start learning. Could any one give me an idea what is happening here? Shraeder valve/compressor issue? Do most splits have a low pressure switch for the compressor? Thanks in advance

[NoNickName]

If the discharge line is cool, it means there is no refrigerant in the unit, which is in agreement with zero psi.
The discharge line should be warm.

[paul_h]

Sound like no refrigerant.
No, most splits don't have pressure controls, only bigger, ducted or better brand ones do, small chinese cheapies don't.
So if it's cutting out all the time, I'd guess current sensor error (most splits do check current draw and shut off the compressor if current too high*).
What goes wrong in a small cheap chinese non-inverter split system:
fan motor (indoor) sensor - shuts off if fan doesn't run or thinks it isn't running even if it is.
indoor thermistors - coil and air sensors.
compressor current sensor.

That's about it, though some do have an outdoor thermistor too. Larger unit's/inverters obviously have more PCBs and thermistors in the outdoor, but I'm talking about fixed speed conventionals like your small mistral.

Besides those faults that bring up fault codes or cycle the a/c on and off mentioned before, there's basically only 3 more other types of breakdowns, compressor seizing/burn out, capacitor faulty (causing seized like symptoms) and outdoor fan or capacitor faulty (unit keeps running until compressor cuts out on internal overload or over current sensor mentioned before).

If none of that applies, the systems short of refrigerant will keep running until the compressor seizes or burns out.


*Edit: Of course there's such a thing as the sensor being faulty, not the compressor actually drawing high amps too.

[paul_h]

If the discharge line is cool, it means there is no refrigerant in the unit, which is in agreement with zero psi.
The discharge line should be warm.
I think he means the 'liquid' line out of the condensing unit, which is usually cool as it's after the flow control eg capillary is in the outdoor.

[alan_man]

Thanks for the great info, so basically there are a few things to test for. I am aware of the indoor temp sensors etc and also liquid line freezing up if low on refrigerant, superheat, subcooling PT charts etc pretty much the basics. What i would like to know is there any test/something to look for that would tell me the compressor is gone, such as a current test with amp meter? I have electrical testing equipment (megga, clamp meter, multimeter etc). Will the compressor keep running constantly without refrigerant or cycle on and off? I am trying to figure out if its worth pressure testing the system, evacuating/charging however would like to know what state the compressor is in other wise the previous efforts would be one big waste of time and refrigerant. I looked for leaks around the flares but couldn't seem to find oily residue that was obvious. I guess since my RHL licence only allows me to install, commission,decommision systems up to 18KW (and also only pipe repairs) I need to find that point weather or not I can successfully diagnose a fault and fix it or successfully diagnose a fault and turn around and say it's beyond my scope. Sorry for all the Q's but Im eager to learn, thanks again

[alan_man]

I forgot to mention, I checked the light status of the indoor and it seemed nothing strange was happening, just the solid green power light, no flashing green or red anywhere

[paul_h]

Compressor will keep running even with no refrigerant.
That's why I said you have other issues if it keeps cutting out; over current, themistor fault, indoor fan fault are the only things that will make the system stop on fault.

A capacitor meter (jaycar sell a good one for $70) will tell you if the cap is OK, of course most are bulged out if metal type, or blown apart if plastic type, but sometimes visible check isn't good enough.
Then just check the winding of the compressor. Common to start is higher than run to common, but the sum of both of them should be start to run resistance.
You could megger the compressor windings.

But I'm going to guess that if you put you clamp meter on the comp the amps may be high, and it's cutting out on overcurrent.
If amps aren't high, you could always see if you can find the sensor on the indoor pcb (compressor wire goes through a loop like your clamp meter), pull it out and reroute not going through it. See if the comp cuts out or not still and what amps you pull through your clamp meter.


edit: I'm not sure even if your system has an overcurrent sensor on the indoor PCB, most do even if they are cheap chinese units. The question is answered if the whole system shuts down (does have a current sensor on a PCB), or if the fans outside and inside are running. If all fans including the outdoor is running, it might be cycling on the external overload on the compressor, under the terminal cover instead. Same applies though, bypass and check current draw with your clamp meter. Sky high = bad compressor (if the capacitor is good), normal current means bad overload sensor tripping for no reason.

[alan_man]

Ok thanks again for the great info, I'm going past for another look to see what I can find. From memory both indoor and outdoor unit fans were fully operational so from what your saying it's either current sensor stopping comp from running due to bad comp drawing too much current or if normal current is being drawn it could just be the faulty overload sensor playing games.

[alan_man]

Sorry, or blown capacitor for the comp. I assume IF I was to replace it it could be any type as long as the rated microfarad is the same on the new one

[paul_h]

Voltage too. Start caps are 240v (like smaller refrigeration gear uses ), permanent split caps (PSC) motors (larger refrig and all A/C gear) run caps are ~330v as they are always in circuit, no relay to disengage them.
The bakelite style caps are usually start caps and 240V. Most wholesalers sell the plastic 330V PSC caps though, up to 60uF. Pretty much only manufacturers sell the metal cased capacitors.
But I've used heaps of the plastic cased caps from actrol in a/cs, go through about 7 a week in summer, most common cause of breakdown really besides systems short of refrigerant due to dodgey flares.

Did one today (fixed speed daikin that use they crappy chinese capacitors), whacked in a plastic 50uF cap and of I go to another job... edit: that capacitor in that daikin did not look blown (bulged out or leaking), but was definitely gone, needed my jaycar capacitance meter to prove it - recommend getting one.
Every time I see a cheap fixed speed a/c about 3-7 years old that's not cooling (omni, mistral, sharp), it's nearly always a blown capacitor if there's no leak. Other issues are soft starters for the larger 6kw+ a/cs of course, and the occasional seized compressor - or burnt out terminals because they came loose, faulty PCB or fan, but faulty capacitor is the most common other than refrigerant leaks.

[croc1774]

Are the valves open on the outdoor unit ?

[alan_man]

Excellent help again, many thanks paul. croc, the valves were definitely open. I did manage to find the problem, before anything I decided to vac test the system which told me the issue almost instantly. System would barely hold the vac,since there was basically zero refrigerant I decided to perform the test with service valves still open and go from there. I closed up each valve and it would vac down and hold no worries. Problem must be in the outdoor unit. Cracked open the suction valve and it would still hold vac, as soon as I cracked open the discharge side valve I instantly lost complete vac. Passed some nitrogen through and you could clearly hear the hiss coming from inside the outdoor unit, it basically just passed through holding no pressure. I removed the cover and noticed a dusty, oily deposit underneath the compressor so basically I'm thinking a braze/pipe failure has occured around the condensor coil area. At this point my mate said just leave it, he'd prefer something newer in place as well as a better brand. If I was to try repair it ,once located is it possible to repair the leak using a touch of brazing?

[paul_h]

Yeah, common place of leaks in elbow at the bottom of condenser, crack in discharge line, pipes rubbed together somewhere or strainer. Can weld them up in most cases (except for cracked pipe, cut out and replace section with crack)

In your case though I wouldn't have - remember there's a problem with compressor as well as it's cutting out.
I also don;t bother with systems that have a leak on the suction (sucks in air and oil ruined).
But liquid line or discharge side, if the compressor isn't cutting out, I weld them up, pressure test, evac and recharge.

[SeanB]

I have had a few capacitors fail this summer, mostly going low in value ( lowest was a 50uF that measured under 5 on testing) but without any outward indications, just a very hot compressor and no cooling. One that puzzled me for a month would trip the breaker intermittently on overcurrent, but would reset ok and carry on running. I placed on a hard start, which reduced it to every so often tripping, but changed the capacitor, even though it measured within 5% of nominal. This solved it, after replacing the breaker as well, it was not happy after all the trips, and would still trip on occasion. I then checked the capacitor on the megger, and you could hear it breaking down at around 800V across it, while the spare ( I bought 2) was very happy at withstanding 1000V across it. Tested the next dead cap, and it too broke down at low voltage. I now will test all suspect capacitors for capacitance and voltage breakdown in future. Note that the capacitors are supposed to withstand 1000VDC across them for 30 seconds as part of the insulation test requirements.