Only my servers have AC. I just offer my own personal thermal pain up to the saints.

Lots more impact that you might think. You should consider finding, and sealing all the leaks.

Yes, the insulated one is the cool one. When you say "unless it was able to thaw out over a few days", were you running the AC during those few days? It will thaw out in a couple of hours, typically, but you usually need to turn it off first. You want to run the AC for 8-10 hours constantly and then check for freezing. The evaporator, in the area you describe as "pretty cool" should be cold. Unambiguously.

The evaporator might make a hiss sound because you could be hearing the refrigerant escaping through the orifice tube/expansion valve. You probably know this already, but the AC works exactly like the cold created by discharging a CO2 fire extinguisher. The rapid expansion and transition from liquid to vapor, because of a pressure drop, absorbs heat. In an air conditioner, they just collect all that gas and then re-compress it and condense it in a cycle. So you can hear the hissing of the refrigerant some times, which is the gas expanding out the nozzle.

I can't tell from your photos, but is the evaporator coated with a spider-webby looking grunge? If so, that is going to have a big affect on the cooling. I am pretty certain your system is not operating at capacity. There's only about 4 things it can be: dirty/frozen evaporator, dirty condenser, low refrigerant, malfunctioning/clogged expansion valve. Well, add two more: there are two fans in the system. Is your condenser fan working? If the coils are clean, the fans are running, and the compressor is working and it's not getting cold on the evaporator, you should have the refrigerant level checked.

There is usually an oil in these systems mixed in with the refrigerant, so I don't really buy the "bubble" hypothesis.

A tiny amount of water vapor in the system has a HUGE effect on the performance of the refrigeration cycle. This is because water has an enormous latent heat of vaporization (which is the same reason you should turn on your AC and let it run all summer without opening the windows if you live in a humid climate). Anyhow, a contaminated system would do this, and then perform well once it was evacuated (all air/water vapor removed) and recharged. Of course, the system *became* contaminated with humidity because there was a slow leak in it, and eventually the performance would degrade.

It was kind of hard to look at the heat exchange fins due to the way the mechanism was mounted and the location of the access panel, but they seemed perfectly clean. Shiny metal and nothing left on my fingers when I touched it.

Just to be certain, the condenser is the big fan unit outside, right? That may well need to be cleaned. It's worth looking at, but I'm not going back outside in the heat today if I can help it.

Like I said, the cold pipe was about as cold as the outside of a glass of iced tea. I would have hoped that it would be cold enough to be unable to hold onto it for very long, but I'm pretty sure that I could have held it indefinitely.

As I understand it, the AC cycle goes like this:

1. Compress refrigerant, which generates heat
2. Dissipate as much heat as possible in compressor unit
3. Pipe refrigerant to evaporator
4. Allow refrigerant to expand, absorbing heat
5. Pipe expanded refrigerant back to compressor
6. Start over

Does that sound right?

My thought at this point is that it is not able to dissipate enough of the compressed refrigerant's heat at the compressor. Does that sound reasonable? (It could still be undercharged, but I'm going to save that as my last resort.) If so, the problem seems likely to be that the compressor isn't able to move enough air over the hot tube, which could be several things, but dirty coils seems likely. I'll check that when the sun goes down.

Maybe better to not buy my fuzzy paraphrasing.


Aside from achieving a capacity bump, one reason we replaced the systems was agae (18 years) and that it did seem to have a slow leak in a coil. So maybe the 4th visit simply finally purged some water out of the system. Anyhow, if Bitt's system is old, I guess my take home is that simply recharging might not fix a malfunctioning loop.

Regardless, my room is nice and cool now and I haven't touched the thermostat once (I probably helped ice up the old system once by asking it for more than it could give).

A coil leak can be fixed by brazing. I had a leak in my coil where coil vibrated enough to wear a few small holes into it (the coil would hold refrigerant for about five days, so it was prety slow). After fixing the vibration problem the holes were brazed. So far that fix has held for 5 or 6 years.

Sounds pretty clean. We can tentatively rule out low airflow across the evaporator, I think. The furnace fan *is* running, correct?



Yep, that's it. Also, if the fan isn't running (burned out motor, bad switch, whatever) that would do it. The condenser fan should run with the compressor is running.



Was the evaporator quite wet? I want to know if the evaporator temperature is below the dew point.



That is *exactly* right. The evaporator and condenser coils are "heat exchangers", which attempt to change the temperature of what's inside them (the refrigerant) to the temperature of the ambient air blowing through them. That's why airflow through these coils is so critical.



Yep, that's reasonable. Do you have an air compressor? If so, figure out which direction the fan is blowing the air, and then use your air compressor hose to blow the opposite direction through the coils and blow all the dirt and leaves and shit out of it. Really the only other thing it could be on the condenser side is the fan not coming on.

I expect you are low on refrigerant. Like the other Jim said, an AC repair guy is just going to put some more in and say "there you go" (because people seem inherently lazy). What you really want is a full evacuation. Ideally, you'd pump it down to a good vacuum and let it sit evacuated for a while, watching the pressure gauge (pump turned off). If it loses vacuum, which you can see on the gauge given some time, then you have a system leak. If you have a leak, then you have to find it. This is typically done by recharging with a refrigerant containing an ultra-violet reflecting dye. You charge with the dyed refrigerant, then run it for a while, then come back with your blacklight and look for where the refrigerant is leaking. Looking at your evaporator, there seems to be quite a bit of corrosion on it, so I would not be surprised if there was a pinhole leak where one of those tubes joins the heat exchanger area.

Jim

Oh, definitely. Quite a lot of air rushing out of the access panel when I was inspecting the evaporator.


Well, it runs (at least more-or-less) when the furnace fan is running. Actually, I assume that the noise I'm hearing is the fan and not the compressor itself. I should probably doublecheck that.


Oh, yeah. Dripping. Well, at least the copper tubing leading to the evaporator. I don't recall the fins being wet; in fact, I'm almost certain they were dry.

I'm going to clean the compressor unit tonight, and observe how dirty it is. I'll also verify that the fan is actually working.

I was surprised at the amount of rust on the evaporator, too. You'd think that a device that's designed to be wet wouldn't be made from carbon steel.

If my system had a leak, would it be able to maintain enough refrigerant to work as well as it does during 90° weather? Maybe if the leak were on the low-pressure side?

Also, I found a more precise reference for the 20° thing from earlier. It's referencing either side of the evaporator. Since the air temperature in the house is at 75°, and it's not replenishing with hot outside air, it should be able to pump out air at least as cold as 60°, using their low estimate of a 15° difference.

I have a suspicion that the AC repair folks talked themselves out of a service call.

Yeah a 20 degree drop across the evaporator seems reasonable. That's more a function of the surface area of the heat exchanger itself. Keep in mind, however, that as the ambient air inside the house cools, that 20 degree drop will continue, *progressively* cooling the air more and more.

If the system has a leak, even on the low-pressure side, it won't maintain refrigerant. These are volatile compounds that evaporate readily to the atmosphere (making them good refrigerants!). Also, even a small leak will get water vapor in the system, totally screwing you.

I think you need a new evaporator and consequently a evacuation and recharge. Too early to tell 100%, but that's how it looks to me.

J
Edit: the fact that it cools below the dewpoint means that it's got some refrigerant, so possibly a full evacuation and recharge could be enough.

Right, I meant to point that out.

Which direction is the refrigerant flowing? I think I was thinking about it wrong. There's the insulated line and the bare line. I was thinking that the insulated line was transporting refrigerant from the compressor to the evaporator, but that's backwards, isn't it?

Right, the insulated line is the suction line (after the evaporator and leading to the compressor). It's insulated because it's cold, and the insulation keeps condensation from forming all along that line and dripping all over your house.

It's not what people think intuitively. People think that a cold fludi of some kind is being pumped to the evaporator, but that's not the case. Hot liquid goes to the evaporator under pressure (the high-side, or uninsulated side), and is expanded into the evaporator through a nozzle, causing it to evaporate (hence the name). The vaporization of the refrigerant takes a lot of heat (the heat of vaporization), and it gets this heat from the environment, thus cooling the coil. So the refrigerant leaving the evaporator is cold, but going to the evaporator it's a warm liquid. The cooling state change actually happens right inside the evaporator.