cheap window reverse cycle ac build

[Xringer]

Is this how you did your 'Hot gas defrost' mods?



In block diagram form, it don't look too hard to do..


http://www.ebay.com/itm/120836674490

[jeff5may]

Yes. The only thing missing in the diagram is a metering device in the defrost line. You don't want the defrost fuction to blast full discharge pressure into your evap. The pros use modulating pressure regulator valves. For my dehumidifier heater, i used a 1 foot piece of cap tube to connect the solenoid valve to the evap inlet. I replaced the cap tube with a TXV (danfoss 1/2 ton) and used a piece of the extracted cap tube for my defrost regulator.

[jeff5may]

Update with some lessons learned:

Since the vibration problem was fixed, this unit has been running like a top! No problems whatsoever out of the unit. I have the programmable thermostat set to automatic mode with an upper setpoint at 74 degF and a lower setpoint at 68 degF. It automatically changes from heat to cool when it has to, which has been awesome this spring. I had it set to 70/74 originally, but my wife said it made her sweat in bed at night when winter tried to move back in. So I dropped the lower setpoint a bit.

I had been taking power usage measurements with my newly purchased kill-a-watt meter until we got a brief power outage. We lost power for like fifteen seconds, then when it came back on, the unit tried to restart and got a locked compressor. Before the overload blew, the fuse in the kill-a-watt meter blew. I traced this fault to my non-heat pump thermostat. Since it was in heating mode, it tried to run the "furnace" when the power was restored. Lesson learned: use a heat pump thermostat with short-cycle protection built in.

Nevertheless, I gathered good data on the unit's heating mode of operation. It starts up quickly, then draws around 1000 watts during pump-up, floating up to between 1300 and 1400 watts at full pressure. It produces around 10kbtu of indoor heat at 25 degF outdoor temp climbing to about 16kbtu at 60 degF. On the coldest day of the season, it got down to 7 degF, and the unit was still blowing out 80 degree air, but could not keep the upstairs warm on its own. I would say it performs well down to about 20 degF. The heating circuit is metered by a TXV recycled from a Lennox heat pump, set for 10 degF superheat right past the TXV sensor bulb. I imagine superheat at the compressor inlet is a few degrees more.

After the initial hack, I had a vibration problem caused by my plumbing job between the reversing valve and the compressor outlet. This line is basically a straight run with u-bends at each end. I had created a pendulum for my reversing valve to swing from, using the compressor's natural vibration as locomotion. Looking back, I should have run a few more inches of tubing and bent it to keep the vibration from stress cracking the tubing (more like a trumpet or a trombone), as well as anchoring the tubing to something solid.

I have yet to measure the performance of the unit in cooling mode, since my kill-a -watt meter is now dead. I do know it works as it should, because it has been doing a good job of cooling the upstairs lately. The cooling circuit is metered through the original cap tubes, charged for 10 degF superheat at around 100 degF outdoor temperature. When it gets hot outside this summer, I'll adjust the charge again.

Which brings on a question: how would you all repair the kill-a-watt meter? I want to add a fuse that I can change (if it ever blows again) without opening the meter. Can I use an automotive blade fuse, or should I use something more exotic? If i jump out the stock fuse, will the added wiring skew the reading? I want to use something I can find locally. Waiting for a special-order 50 cent package to arrive sucks. Especially when it costs $5.00 or more.

[AC_Hacker]

Originally Posted by jeff5may

...We lost power for like fifteen seconds, then when it came back on, the unit tried to restart and got a locked compressor. Before the overload blew, the fuse in the kill-a-watt meter blew. I traced this fault to my non-heat pump thermostat. Since it was in heating mode, it tried to run the "furnace" when the power was restored. Lesson learned: use a heat pump thermostat with short-cycle protection built in.

I'm glad you spelled this out in this post. This issue needs to be put before people early who may want to hack HVAC machines. You might consider editing this valuable insight into some of your previous posts, as a constructive caution to those who may try this later.

Originally Posted by jeff5may

Nevertheless, I gathered good data on the unit's heating mode of operation. It starts up quickly, then draws around 1000 watts during pump-up, floating up to between 1300 and 1400 watts at full pressure. It produces around 10kbtu of indoor heat at 25 degF outdoor temp climbing to about 16kbtu at 60 degF. On the coldest day of the season, it got down to 7 degF, and the unit was still blowing out 80 degree air, but could not keep the upstairs warm on its own. I would say it performs well down to about 20 degF.

If you have more points of data, it would be very useful to have a graph of performance. I don't know how handy you are with graphing... I can help if you need help.

Originally Posted by jeff5may

After the initial hack, I had a vibration problem caused by my plumbing job between the reversing valve and the compressor outlet. This line is basically a straight run with u-bends at each end. I had created a pendulum for my reversing valve to swing from, using the compressor's natural vibration as locomotion.

Just in case you took any photos along the way, this would be a dandy place to edit them in.

Originally Posted by jeff5may

Which brings on a question: how would you all repair the kill-a-watt meter?

I have one that I hooked up to a high-watt resistance heater. It still registers, but the readings are off by around 100%. Fixing these things could be a thread in itself.

* * *

Congratulations on your success with your hacked unit!

I'm really looking forward to seeing what you do next.

Best,

-AC

[jeff5may]

Summer update:

The unit has been running on autopilot for the whole summer. I've had no problems at all, it just does what it should. I now trust it.

One minor modification I did to the unit: I taped up the drain hole I cut in the drain pan for frost removal during heating season. In cooling mode, the drain pan fills up with condensate from the indoor evaporator. When the pan fills to a certain point, the outdoor fan slings the water on the condensor coil to dispose of the water. This sounds like a gimmick, but it really improves efficiency of the unit, especially when it's hot outside. The cool water from indoors helps subcool the liquid line on its way indoors. The water slung on the condensor coil drastically lowers high side pressure, allowing the compressor to move more refrigerant. And indoor exhaust temperature is much lower than with the drain unclogged, so much that the evap coils try to freeze for the first few minutes of operation. The unit actuall blows frosty fog for a minute while the condensor heats up outside !0

I still haven't had a chance to fix my kill-a-watt meter, but when I do, I'll post energy usage data vs. heating/cooling capacity. In my opinion, the unit has already paid itself off by leveling out my utility bills. And my wife doesn't have to mess with the central air system when she's feeling hot or cold.

[jeff5may]

Fall update:

The unit finally couldn't heat the whole house last night. For the first time this season, the gas furnace came on around 2AM for a minute. Then every so often until sunrise. Outdoor temp bottomed out at 26 degF overnight. Brrrrr...

When temps rose close to freezing this morning, the unit started banging and clanging like an unbalanced washing machine on spin cycle. It woke me up! I threw on a robe and went outside to investigate. I then had a "Doh" moment.

I had forgotten to unplug the sump drain! The frost that had been forming and defrosting all night had frozen in a ring around the shroud of the outdoor coil! When outdoor temps approached freezing, and the unit went into defrost, lots of this ice melted and filled the drain pan while the fan was turning slowly. When the unit went back into its heating cycle, the outdoor fan slung this warmish defrost water onto the ring of ice. Some of it separated from the ring and got stuck in the fan blade, causing all the racket.

When I saw what was happening, I quickly yanked the tape off the drain hole. I grabbed a rubber hammer and gave the unit a couple of whacks here and there. It quit making noise, so I went back inside and back to sleep. After I woke, I went out to check on the unit. It was still working fine, and the drain pan was bone dry.

Later that evening, my neighbor behind me asked me why I was still running a window a/c in the heating season. He was concerned that i may have killed it last night! How embarrassing.

[Xringer]

LOL!! So, was the reason the gas came on last night:

A. The HP couldn't keep up, at 26F?
Or,
B. High water & ice resulted in degraded HP operation?

[jeff5may]

Probably both. I probably need to do the math that AC_Hacker had suggested as to the COP and raw btu output of the unit as the temps drop off. Can someone lead me in the right direction here?

I can rig a data logger to the unit to get good data. I have some arduino parts laying around.
Last year, I did some rough calculations using basic equipment such as:
Kill-a watt meter (before it died - I can probably resurrect it)
Digital handheld anemometer/thermometer (indoor discharge)
Digital indoor/outdoor thermometer (ambient vs. evap)
Analog alarm clock (cheap hour meter)

If I look hard enough, I might be able to find this old data, but it was basically done to verify superheat / temp split / dT, etc. and is not complete in any shape or form. I just ran a few numbers to make sure the monster wasn't just gobbling energy.

[AC_Hacker]

Originally Posted by jeff5may

... I probably need to do the math that AC_Hacker had suggested as to the COP and raw btu output of the unit as the temps drop off. Can someone lead me in the right direction here?

Doing a COP analysis for a water-in-water-out heat pump is pretty straight forward, since you can work with a known volume of water and measure the before & after difference. You can also work with measured flow rates of water and do the same thing.

But with an ASHP, calculating COP is not such an easy matter, since air, a gaseous mixture, diffuses out into the environment.

However, I did see a university doing it though... by capturing air output in a very large plastic bag (yes, really), they were able to measure the volumetric rate of the output... I think they did something similar for the input, too. Then measure the entire heat content (sensible and latent), of the input air stream and the output air stream, to find out the work done on the air to change its temperature & humidity, etc. Then divide by the electrical power required (Kill-a-Watt used here) to do that work.

To build the 'COP vs. Ambient Temperature' curve, several of these tests would need to be done with various incoming air temperatures, to establish enough data points for the curve to be reliably created.

That's the best I can come up with.

With a water-in-water-out heat pump, it is really easy, just run a data logger with a temperature probe in each water container, and periodically measure the power consumed. Since there is no change of state, and the effect of heat on water is linear, the process is very straight forward.

Best,

-AC

[jeff5may]

Yep. I know what you mean here. Water heat flow is easy to calculate. Air heat flow is not. The main difference is that air heat is measured in enthalpy.

What is enthalpy, you ask? Enthalpy is the sum of sensible and latent heat in the air. You see, dry air has hardly any latent heat in it, where moist air has lots of latent heat in it. The difference in humidity affects both the heat transfer capacity and the density of the air. What this means is that 1 pound of dry air will not move as much heat as 1 pound of moist air.

I'm not trying to start a research project on my homemade heat pump. From general observations, I can tell when the unit is operating how much heat it is actually moving. I am highly satisfied with its operation and money saving ability during most of the year.

Its output is massive above about 40 degrees F. Depending on outdoor conditions (mainly humidity), the coil begins to need to defrost below 40. In the dry weather, the dewpoint is low enough that not much frost builds up on the unit, so it only defrosts for less than a minute. But when it's damp outside, the evaporator starts to clog up with frost pretty quickly. By the time the next defrost cycle begins, the unit isn't putting out much heat at all. It takes much more time to defrost as well. Once it defrosts, the heat pours nicely indoors until the outdoor heat exchange coil clogs again.

Below about 30 degrees, the unit hardly frosts up unless it is actually snowing outside (or trying to). It doesn't put out quite as much raw heat as above 40, but then again, it doesn't defrost for much time, either. It also draws less power from the mains.

Due to the above observations, I have set the defrost control to its minimum setting of 30 minutes. When the unit is operating in the "frost collecting" range of 30-40 degrees F, the outdoor coil gets defrosted before it has a chance to completely clog up. If it had a larger evaporator or more airflow, I imagine it would still need defrosting in this range to keep up with mother nature. I am happy with the defrost control; ICM obviously did their homework when they designed it. The thing really can sense when the coil is free of moisture and quickly terminates like it should.

The only thing I really have left to do with this unit is to figure out whether it puts out enough heat below 40 degrees to justify running it. I do know that it annihilates my gas heating bill down to around 30-35 degrees, depending on humidity. I haven't run it long enough with the increased defrost frequency to tell how much better it will perform.

Given that my normal average daily temperature is 35 degrees or higher for the whole year, my average winter low temp is 29 degrees, and my 99% heating design temp is 15 degrees, I know the unit will provide more heat than a 1500 Watt space heater throughout this range. Thank you NOAA. At roughly freezing outdoor temps, the unit begins to defrost and the home's heat loss catches up with the unit's output. I'm just not sure how to calculate whether running this thing year round would cost much more in electricity than the gas usage it offsets.