Aircon: Testing That It'S Working Properly

[Greenside]

Had some trouble with a 13,000BTU two year old Mitsubishi split unit system in an apartment. Finally, the repair crew say it's fixed and simply by holding my hand in front of the outlet I could feel it was blowing but not cooling much. So after some discussion they fooled about with more refrigerant and what looked like a small compressor or vacuum pump and the output was significantly better. I've not paid them yet (insisting on several days of testing) but I wonder if there are any standard (and less subjective) ways to check if an a/c is doing what it should. BTW, these guys initially relied on waving their hands around in front of the vent and then jammed a couple of simple wires into a multimeter and positioned them in the airflow about 20cm apart. This gave a reading (of what, I'm not sure) of 19 which then reduced to 16 when the air began to get cooler. Moisture content maybe?

[stgrhe]

To properly test any refrigerating system one must apply a pressure gauge to get accurate readings of its function. From the readings one can easily see whether the system needs more refrigerant or if it is too much refrigerant. If the indication says the system needs refrigerant one should first find out if there is a leak and in such a case fix that first before adding more refrigerant.

If the system uses R12 or R22 one common cause for poor efficient is moisture in the system. The moisture should in such a case be cleared using a silica gel dehumidifier.

[longball53098]

I have a multimeter with a temperature probe. It may look to the casual observer like two wires plugged in to the meter but the meter has a temperature selection and gives readings in F or C. Could be what they were using!

[Greenside]

I have a multimeter with a temperature probe. It may look to the casual observer like two wires plugged in to the meter but the meter has a temperature selection and gives readings in F or C. Could be what they were using!

Could be - pretty well nigh invisible if it was.

The point here is that they say it's got the right refrigerant readings etc but it simply doesn't feel very cool to me (not nearly as cold as the a/c in my truck, for example) but I can see that getting any improvement with just my subjective assessment is going to get tedious.

What I need is something along the lines of a temperature probe or a simple thermometer that I can hold in the airflow right up against the fan and a table that says that a 13,000 unit on full cooling should give a reading of approximately X degrees after 10 minutes running. I'm surprised no one markets something like that, or maybe they do?

Edited July 16, 2010 by Greenside

[elkangorito]

The gas used in Thailand car airconditioning is most likely (could be) R12 for older vehicles, which has a higher boiling point compared to R22. As such, R12 was used in applications that required very low temperatures. I say was because both R12 & R22 are being phased out in developed countries.

Here's a graph that may be of some help. Unfortunately, it was made by non Metric people (backwards country obviously) so temperature conversion from F to C is necessary.

[Pib]

I have a multimeter with a temperature probe. It may look to the casual observer like two wires plugged in to the meter but the meter has a temperature selection and gives readings in F or C. Could be what they were using!

Sure sounds like that was what they were using and they got a reading of 16 degrees Centigrade which is about 61 degrees Franheit. Output air temperature wise, that sounds about right assuming they didn't test very long, the room temperature was already high during the day, and the room hadn't had a chance to fully cool down yet.

Would think with the A/C's room temperature setting of 25 degrees C, the A/C should be blowing around 8 degrees C (46 degrees F) from the output duct at night after the room has cooled down....I base this on doing an actual test right now with a temperature probe in the output duct of my York split air 18,000 BTU...it's now 8:45pm at night...during the day it may not blow quite as cold due to the higher heat load...the A/C compressor/unit would just run longer to pull the room temperature down. When I lived in the the States I had a home central air system and it would blow around 10 degrees C air at the output of the cooling fins for a room temperature setting of 23 degrees C.

Edited July 20, 2010 by Pib

[bankruatsteve]

I don't know the ways to test an A/C, but it seems that's not the point. Why not try setting the A/C to some setting - say 23 degrees C - and wait 30 minutes - or whatever seems like a good time. Check the temp in the room and then (assuming it's not a concert hall), you can conclude that the A/C is working or not. IE: why care what the temp is coming out of the blower if it reaches the room temp you want within reasonable time? If not, then it's up to the apt owner or the A/C provider to make it right. Or, what is it that you're paying for? Again, the spec should be a room temp of x degree after y minutes and whatever else is causing that to not happen is what you expect to pay for to get fixed.

[stgrhe]

... why care what the temp is coming out of the blower if it reaches the room temp you want within reasonable time?

Because if the unit is running short of refrigerant the compressor will be hunting with too long "ON" per duty-cycle. This will drive up the electricity bill and is an inefficient way to operate an AC. In the long run it will also cause additional wear and tear on the compressor itself.

Edited July 21, 2010 by stgrhe

[Pib]

... why care what the temp is coming out of the blower if it reaches the room temp you want within reasonable time?

Because if the unit is running short of refrigerant the compressor will be hunting with too long "ON" per duty-cycle. This will drive up the electricity bill and is an inefficient way to operate an AC. In the long run it will also cause additional wear and tear on the compressor itself.

Big 10-4 on your explanation.

[elkangorito]

I don't know the ways to test an A/C, but it seems that's not the point. Why not try setting the A/C to some setting - say 23 degrees C - and wait 30 minutes - or whatever seems like a good time. Check the temp in the room and then (assuming it's not a concert hall), you can conclude that the A/C is working or not. IE: why care what the temp is coming out of the blower if it reaches the room temp you want within reasonable time? If not, then it's up to the apt owner or the A/C provider to make it right. Or, what is it that you're paying for? Again, the spec should be a room temp of x degree after y minutes and whatever else is causing that to not happen is what you expect to pay for to get fixed.

This is why I posted the graph, which is specific to DX units (Direct Expansion). The graph is for R22 charged units.

The usual way a refrigeration technician tests the temperature of the air coming off the evaporator coil (of a DX unit), is to first use a psychrometer to establish the ambient temp & the Relative Humidity of the room. They then use an electronic temperature device, which is placed directly in the air output of the evaporator but not touching the fins (not touching anything).

Most experienced refrig technicians don't need a graph as they have this info in their heads.

Edited July 21, 2010 by elkangorito

[Greenside]

Many thanks for all the suggestions.

It's not really practical to have the aircon repair guys hang around for an hour to see if the unit can maintain the room temp at 22 and since their way of proving they'd completed the job was simply to wave a hand in front of the unit and say "OK now?" I was looking for something a bit more instant as an indicator. The person living there tells us that it seems to be working now so that it about as good as I can get. We'll check his power bill at the end of the month to see that it isn't outrageous.

[Pib]

What really comes in handy to easily check/test temperatures coming from A/Cs, temperature inside a refrigerator, etc., is one of those cheap digital thermometers with an "external/outside temperature sensor." Normally the sensor is on then end of an approx 6 ft long wire. They only cost a few hundred baht and you can buy them in many places. Of course you could use a plain old thermometer, but one with external/outside sensor also allows you to easily check the temperature reading without removing the sensor, opening the frig door, etc. And most of these digitial temperatures have a max and min temperature memory which tracks remembers the highest and lowest temperature during a timeframe. Well worth a few hundred baht investment.

Edited July 22, 2010 by Pib