new air conditioner for brunnhilde - part 2

As with any major renovation for Brunnhilde, no chances is too small for risk taking. Before committing to proceed with the air conditioning system replacement, I set out to conduct a number of bench testings.

i remove the enclosure shroud for the evaporator housing to review what are all inside; of the most important, i want to see how the condensate drainage design; the shroud is very well made with sturdy glass reinforced plastic; note also the aluminum foil insulation

the evaporator sits on a long rectangular trough of the air conditioner's base pan lined with black foam; as the condensate level rises and reach the bottom of the adjacent it spills over into the square bowl that is the optional condensate pump mount; in the square bowl are two pre-drilled holes that serves double duty - 1) for used by two screws to secure the condensate pump, or 2) serve as condensate drainage holes if there is no pump

 in all the base pan is very well design and made

the bigger holes scattered around the base pan are for provision of the last resort condensate drainage should the two small holes are clog

 here you can see the two spillways formed by the two vertical slots of the condensate pump housing

 these are the two condensate drains

 i set up the AC for bench testing using saw horses

first figuring out the six wires and their functions at the connector harness; i did not purchase the ceiling air distribution box (ADB) as I want to save money (yes, i can be very frugal) by reusing the thermostat and the rocker switch from the Kerstner system

 the squirrel cage blower top cover has no fasteners; it is held in place by the evaporator top shroud cushioned by a layer of foam

 plastic impellers

 compressor label

 condenser fan motor label

 evaporator blower motor label

the fan can easily be removed to gain access to the electrical junction box; note the double washers likely a design afterthought to tune (minimize) the fan noise

 there is a PTCR (positive temperature coefficient resistor) - i assume is a part of the start capacitor network for the compressor

 looking up from the bottom, the smaller of the two port is the cold air exit into the cabin

my solution to my office that is in the west side of the house in a 90-degree day

 i saved about $60 ordering this mounting frame, instead of the ADB

don't do this at home - risk property damage, injury, or worst death; i am jury-rigging a crude wiring for to bench test the AC without any switch, or a thermostat; there are two objectives - 1) a complete function checkout, and 2) to see if it can be powered by the Honda generator in Eco mode

don't do this at home - risk property damage, injury, or worst death; my wire gauge is too small but they were what i have on hand to sacrifice for the causes; the under-sized wiring must contribute to more difficult of the starting as the motors takes longer to gain speed due to excessive voltage drop

  i first use household power to make sure everything works

then the cliff hanger moment - will the Honda able to power it in Eco mode? yes, even with all the motors and the compressor were to be started together from off

The technical manuals always tell you never to short cycle a refrigeration system; if you restart the compressor shortly after it just been shut off. The restarting will result in very high starting current because the compressor has to work very hard against high head pressure. To confirm it,  I tried it (without waiting for a minutes or two) and sure enough the Honda generator over-load protection tripped.

Taking advantage of the hot day, I conducted countless cycles of on/off tests, each time after a reasonable running. Because the air condition is cooling the ambient air it gets a very good work out. The Honda starts it reliably each time in Eco mode.

Next I set out to see if letting the blower constantly run, if the starting is easier for the generator. I do this by jury-rigging a SPST switch to simulate the thermostat. There is no perceptible difference which make sense as the blower motor consumes the least current.

the SPST rocker switch simulates the thermostat

It should be note that my tests are conducted at near sea level, albeit in a rather difficult condition for the air conditioner. It is possible at higher altitude the Honda may be challenged. However, in my observation the main challenge is not so much of the Honda's maximum load capacity. It is it's ability to spool up the RPM to create enough voltage and current for the electronic inverter to sustain the starting transient current. When the air conditioner's compressor and motors first turn on and create a huge aggregate surge current the Honda responds by spooling up the RPM. As this takes time the inverter's output voltage sags. It is a fine line when the engine can spool fast enough before the overload protection trips.

So what can I do to improve the situation if the interoperability deteriorates? There are two. Coleman has a soft starting kit one can add to the compressor starting circuit. The other is my own idea. Since the condenser blower fan is next higher power consuming motor, I can delay it's starting by a few seconds. This can be done with off-the-shelf time delay relay - which are used in commercial HVAC for staggering motors starting.

What these bench testings tell me is, I was wise to decided against buying the 11,000 BTU Dometic Penguin II which was one of the two finalists in my selection process. I have no doubt the Honda will have trouble coping with it (in Eco mode) due to it's overall higher power consumption.

Update 07.07.2015 20:00
In the late afternoon, I conducted one last test case with the Honda. The Mach 8 Cub AC's blower has 2 speeds. I had been testing only with the blower hardwired to low speed. I decided to also test the high speed mode. I have alleged above that the challenge for the Honda set to Eco mode is more so how well it can spool up in respond to the huge startup current surge, rather than its being taxed by getting too close to running out of maximum power headroom. If my logic is correct, setting the blower to high should make it easier for the Honda to respond.

The blower fan in the Mach 8 Cub moves plenty of air even at low speed. I cannot see that for this 9200 BTU unit there will ever be a need for setting it to high speed.

I proceeded to conduct the test with the blower set to run constantly at high speed. It turns out that the Honda can start the compressor and the condenser fan much easier! I ran a number of cycles with reduced off time after a reasonably on time. I used 1 minute of off time which no thermostat will have a shorter time. The Honda spooled up perceptibly faster than with the blower running at low speed!

Why? It is very simply explained. With the blower fan set to low speed, the generator's engine is barely idling. When the compressor and the condenser fan turns on, it has to spool from idle to beyond full load striving to respond to the transient load. Gasoline engine has very low torque at idle RPM and it struggles to ramp up.

When the blower is set to high, the generator RPM is somewhat higher than idle which has higher torque as well as increased flywheel effect. It is better able to spool up from a higher initial RPM.