Inverter A/C running at low(est) power

liamg

Hello,

I have a 18000btu split type inverter air conditioning system and I want to run it ONLY at reduced output power (low input current), as it have to share the same power inverter with a refrigerator and a well pump. If the refrigerator and the pump start simultaneously, the inverter almost enter the "red" zone. That's it, it would not accept any other semnificative load during that time.

To avoid the inverter shutdown, I thought of forcing the inverter air conditioning system to work at a reduced power despite of its thermostat settings or room temperature. Let's say, I want it to run at 1/5 of ful power all the time.

Does anyone knows if that's possible without hacking the A/C internal circuits? 

bill von novak

liamg

To avoid the inverter shutdown, I thought of forcing the inverter air conditioning system to work at a reduced power despite of its thermostat settings or room temperature. Let's say, I want it to run at 1/5 of ful power all the time.

Does anyone knows if that's possible without hacking the A/C internal circuits? 

Short answer - no.  You would have to hack at least the thermostat wiring to ensure that it only "sees" a minimal temperature rise; this would likely keep the system in lower power mode.  If it's the Fujitsu mini-split you can get a remote wired thermostat and hack that instead.  It uses a three wire connection, though, and I have no idea what signals the interface uses.

liamg

Many thanks for the tip. I have a Daikin regular(?) split system. It has a infrared remote control (not sure about a wired remote control interface).

Do you know how low an inverter A/C system can go? I didn't find any reference in the specifications. 

And how to hack the thermostat to only see a minimal temp difference (as long as the room temperature won't be constant)? My device is out of waranty thus I could eventually hack the internal circuits but I prefer a "cleaner" and reversible solution. 

Does anyone have a service manual (full schematic) of such a device? I wonder if I could by-pass the internal MCU and drive the compressor with an external/proportional signal to ensure a low power output. 

Later edit:

The right question is: how to manually modulate the output power? 

BB.

Refrigerator running, inhibit well pump?

Well pump running, turn off refrigerator?

Use current sense relay on primary load.

- Bill

liamg

Thanks, @BB.. that should be the next best solution but still it would be nice to take a bath while the fridge is running!

Btw, that's a temporary setup untill I'll upgrade the inverter. If I could magically twist some wires to gather a manual modulation control..

bill von novak

liamg said:

Do you know how low an inverter A/C system can go? I didn't find any reference in the specifications. 

And you probably won't.  Even average power specifications are often not based on actual running current.

Does anyone have a service manual (full schematic) of such a device? I wonder if I could by-pass the internal MCU and drive the compressor with an external/proportional signal to ensure a low power output.

Don't think you'll be able to do that.  The "internal MCU" is probably also generating the gate drive signals for the inverter, and unless you are an EE with years of experience with motor drives, you don't want to try to replicate that.

The right question is: how to manually modulate the output power? 

Well, the way to do it manually is to turn up the temperature setpoint until the A/C takes the least amount of power.

liamg

bill von novak said:

Don't think you'll be able to do that.  The "internal MCU" is probably also generating the gate drive signals for the inverter, and unless you are an EE with years of experience with motor drives, you don't want to try to replicate that.

Actually, I am an "EE with years of experience" (though not necessary with motor drives but power electronics) but I was looking for a cleaner/reversible solution. 

bill von novak said:

Well, the way to do it manually is to turn up the temperature setpoint until the A/C takes the least amount of power.

That's not going to work, as the room temperature is not constant hence the unit will automatically increase/decrease its power to reach the setpoint.

I found some service/engineering manuals for a similar Daikin unit. From what I've read so far, one quick solution might be to replace the internal thermistor with a potentiometer, to simulate a constant room temperature (near the setpoint). 

But, if the internal MCU is smart enough, it could detect this "status quo" (e.g. taking hours trying to reach the setpoint) and might gradually increase the power to finish its job (setpoint temperature = room temperature).  

A full detailed schematic would be a life saver, as I don't have enough time to dig inside the unit to decrypt the PCB traces & components.

littleharbor2

Have you considered getting a smaller unit?

mvas

Here are two alternate solutions:

Would it be OK to shutdown the Fridge whenever the Well Pump is ON ?

High Priority = Well pump
Low Priority = Fridge

Solution #1:
http://www.ebay.com/itm/CR-Magnetics-CR4395-EH-240-11-X-CD-ELR-I-Rev-A-Current-Sensing-Relay-/262915695146?hash=item3d36ff222a:g:VREAAOSwhQhY3G9h

Use this device to sense when the amps are flowing in the hot wire of the Well Pump
Wire the Normal Closed NC contacts to disconnect the Hot wire of the Fridge

NC Contact rated at 10 Amp 
Needs 24 Volts to operate.

When the A/C and the fridge are BOTH operating and then the Well Pump tries to start 
will this switch turn the fridge off FAST ENOUGH before overloading the Inverter?

Solution #2:
If the above is not acceptable the add a Relay after the Water Pressure Switch.
That relay will turn the fridge off, BEFORE turning the well pump motor on.

mcgivor

liamg said:

bill von novak said:

Don't think you'll be able to do that.  The "internal MCU" is probably also generating the gate drive signals for the inverter, and unless you are an EE with years of experience with motor drives, you don't want to try to replicate that.

Actually, I am an "EE with years of experience" (though not necessary with motor drives but power electronics) but I was looking for a cleaner/reversible solution. 

bill von novak said:

Well, the way to do it manually is to turn up the temperature setpoint until the A/C takes the least amount of power.

That's not going to work, as the room temperature is not constant hence the unit will automatically increase/decrease its power to reach the setpoint.

I found some service/engineering manuals for a similar Daikin unit. From what I've read so far, one quick solution might be to replace the internal thermistor with a potentiometer, to simulate a constant room temperature (near the setpoint). 

But, if the internal MCU is smart enough, it could detect this "status quo" (e.g. taking hours trying to reach the setpoint) and might gradually increase the power to finish its job (setpoint temperature = room temperature).  

A full detailed schematic would be a life saver, as I don't have enough time to dig inside the unit to decrypt the PCB traces & components.

The MCU is relying on two inputs, a thermistor, along with a set-point  in this case, I would doubt that it could detect that you were attempting to fool it by adding a resistor to simulate a particular  room temperature, but  in all likelihood it would  either  shut the compressor down  or make it run at maximum speed,  rather than reducing the compressor speed for any length of time , depending  wether  the thermistor has a positive or negative temperature  coefficient. In the shutdown situation there would be a brief period where it would reduce compressor speed before shutting down, but once satisfied it has met it's goal it would remain off, unless there was a way of inputting a moving target electronically, but that sounds complicated. 

Good brain food none the less.
 Just some thoughts 

liamg

@littleharbor2: 

 It's a temporary condition, not worth exchanging the A/C unit. 

@mvas: 

 Thanks, that's an easy solution but I still prefer to lock down the A/C power instead. The fridge has a LCD display & stuff and it starts blinking, lose the settings (time, temperature) and even starts the audible alarm in case of power lost so I better not touch it.

       

mcgivor said:

The MCU is relying on two inputs, a thermistor, along with a set-point  in this case, I would doubt that it could detect that you were attempting to fool it by adding a resistor to simulate a particular  room temperature

I'm afraid that it would detect the lack of change in the room temperature for an extended period of time hence it will try to increase the power.

I thought of setting the fake room temperature at 25C and the setpoint at 24.5C to ensure a minimum output power from the A/C unit but, depending on its "brain", it might or might not keep that minimum power for ever.

mcgivor said:

it would  either  shut the compressor down  or make it run at maximum speed,  rather than reducing the compressor speed for any length of time , depending  wether  the thermistor has a positive or negative temperature  coefficient. 

I thought about replacing the thermistor entirely with a potentiometer, not putting a pot in parallel with the thermistor. I guess I'll give it a try and see what's the MCU answer..

Estragon

What if you had the ambient reading taken at or near the cool air outlet of the unit, so it always thinks it doesn't need to cool much?

mvas

liamg said:

@mvas: 

 Thanks, that's an easy solution but I still prefer to lock down the A/C power instead. The fridge has a LCD display & stuff and it starts blinking, lose the settings (time, temperature) and even starts the audible alarm in case of power lost so I better not touch it.

  

OK, what if you ... LOCK-OUT the A/C when both BOTH the Fridge and Well Pump are on ?

liamg

> @Estragon said:
> What if you had the ambient reading taken at or near the cool air outlet of the unit, so it always thinks it doesn't need to cool much?

Sounds fair, I'm gonna play with the remote control before I'll try to hack the indoor unit.

@mvas:

I have a back-up inverter (1kW) so I could isolate the A/C from the fridge/pump for now but still I have to limit its output power to 800W or something.

mike95490

Sadly, I could not figure out a way to lock out appliances on a 90 acre ranch, so I went overboard on the inverter so it could manage it all.  Since I have gravity fed water, I restrict the 1/2hp water pump to prime solar hours only.

Dave Angelini

liamg said:

Many thanks for the tip. I have a Daikin regular(?) split system. It has a infrared remote control (not sure about a wired remote control interface).

Do you know how low an inverter A/C system can go? I didn't find any reference in the specifications. 

And how to hack the thermostat to only see a minimal temp difference (as long as the room temperature won't be constant)? My device is out of waranty thus I could eventually hack the internal circuits but I prefer a "cleaner" and reversible solution. 

Does anyone have a service manual (full schematic) of such a device? I wonder if I could by-pass the internal MCU and drive the compressor with an external/proportional signal to ensure a low power output. 

Later edit:

The right question is: how to manually modulate the output power? 

There is a company that makes a retro for motels. It basically allows the unit to be set to temperature like a thermostat.I will dig it up later if you ask. The other way is to use a aftermarket adapter to control with your smart device. The best way is to do that and control the heat pump based on an inverter or CC AUX output based on battery voltage. I do this for my clients.

Your problem would be lessened by not using that big heat pump. Offgrid, it is best to use a .9000 btu or a 1 ton as they are much more efficient than the large units! Splits that are inverter based have 3 or 4 power levels. A 9000 btu unit will draw about 300 watts if the room temp and the programmed temp are about the same. There are many strategies but it is best to cool the room down early to avoid the room getting hot in the first place. Since you are solar powered it is the way to go. I tried this strategy in 2008 when we had a wildfire arround us. There is an update here.
http://forum.solar-electric.com/discussion/23007/mini-split-update-for-offgrid/p1

bill von novak

liamg said:

That's not going to work, as the room temperature is not constant hence the unit will automatically increase/decrease its power to reach the setpoint.

I found some service/engineering manuals for a similar Daikin unit. From what I've read so far, one quick solution might be to replace the internal thermistor with a potentiometer, to simulate a constant room temperature (near the setpoint). 

But, if the internal MCU is smart enough, it could detect this "status quo" (e.g. taking hours trying to reach the setpoint) and might gradually increase the power to finish its job (setpoint temperature = room temperature).  

Right, so you have to spoof it.  Measure the current.  If the current is too high, then decrease the sensed temperature.  You could do this with a DC bias, a digital pot, even a Peltier device next to the temperature sensor (if your goal is to not mess with the board at all.)  No sane system will increase the power if it thinks the room is 20 degrees below setpoint.

In general it's easy to find the temp sensor.  It's usually a 10K thermistor and they look pretty unique.

liamg

Measuring the current it's a good idea (as I already have some Hall current sensors available). Using this information as a feedback (to adjust the fake room temperature) it's brilliant, too.

But, before trying that, I'll keep searching for an even simpler solution, like setting the MCU "maximum output power" parameter using the remote control or some PCB jumpers.

mcgivor

Is the unit a 2 stage compressor or is it an inverter type?

liamg

Dave Angelini said:

A 9000 btu unit will draw about 300 watts if the room temp and the programmed temp are about the same. There are many strategies but it is best to cool the room down early to avoid the room getting hot in the first place.

It would be nice if I could keep the unit "brain" (including sensors) unaffected, to rely on it's actual setpoint and the (true) sensed room temperature. If only I could lower its maximum output power..

Btw, I don't mind if it runs continuously, as I have enough solar/battery for that amount of power (5-600W).

@mcgivor:

As I did mention in the first post, it's an inverter type unit (split/wall mounted).

mcgivor

@liamg
What model is the mini split? Was able to find a service manual with wiring diagrams but no idea if it would be relevant. 

liamg

@mcgivor:

Thanks a lot for your support!

It's not exactly a mini split but a regular one: Daikin FTX50GV/RX50GV combo (18000BTUs).

I did not find its service manual neither. Maybe I should do more using the wired remote control interface? Or maybe there is a service code available to limit its maximum output power?