Ämne: Panasonic IR remote controlling

[apophis]

Hi all! In Finlands local heat pump forum we have been trying to find out how to control air-to-air heat pumps using PC with infrared transmitter. So far attempts have been made to record remote control codes and trying to solve what the codes contain. Overall it seems that the codes are very long and therefore cannot be learned to 'one-for-all' remote controls.

I've been looking into Panasonic controlling and now it seems that the code format is solved in general and I even got my E12 to reveice codes sent from PC!

First I used an own IR recorder to record raw codes from the Panasonic remote control. The code is stunning 216 bits (27 * long and contains a lot of information. Here is an example of setting heat mode and 20 degrees:

0220E004000000060220E00400492880A40D000EE000000100069D

The temperature is directly set in 7 bits starting from bit 112.

Code has a checksum in the last byte (9D) calculated as follows: f4 + 02 + 20 + ... + 00 + 06 (= 9D in the case above). Summing all but the last byte to f4.

The code signalled using 40kHz carrier like this, all times in microseconds:

Header:

3535 carrier on
1705 off

First 8 bytes, LSB first:

For each 0 bit:

450 on
400 off

For each 1 bit:

450 on
1269 off

After 8 bytes, another header is sent:

474 on
9963 off
3538 on
1677 off

And then rest of the 19 bytes are sent as the 8 above.

Last a trailer is sent:

495 on

I have tested the sending with a very simple IR hardware where 2 IR leds are directly connected to PC audio output L and R channel outputs. The diodes are in opposite polarities, no connection to ground, but directly in L and R.

I thougt somebody might find this subject interesting and might want to give it a try. Please also find some custom software to record and playback IR codes and some logs from Panasonic remote. Feel free to use and modify the source, should compile with Visual C++. I couldn't attach the .zip files here, but you can found them from the original thread (ir.zip, pana.zip).

Link to the original thread: http://lampopumput.info/foorumi/index.php/topic,1904.0.html

[Paxmax]

That was an original idea for sure!
Good job and congratulations, I'll make sure to bookmark this.

Yes, there is a heck of lot of information in Panasonics transmissions.
Maybe it can also be decoded for different angles of the horizontal and vertical vanes too?
ION on/off, powerful on/off, "pump on" statement and "pump off" statement.

Also there is the time and date information and possibly even TIMER settings.

Once more, GOOD JOB! and thanks for the information!

[apophis]

Thanks! I think you can do anything with this code. Some of the commands like Quiet and Powerful are shorter than the normal commands. I haven't yet recorded all combinations and solved all bits. It seems that you can set the temperature to any value, for example below 16 degrees, assuming that the indoor unit won't limit it. Also custom vane settings are surely possible.

Also there might be interesting test commands etc. that could be utilized. My goal has been somehow to extend the defrosting period at low outdoor temperatures when the air is dry. As you know the outdoor unit does unecessary defrosts when temperature is below -10 which affects the COP heavily.

There is a simple way to prevent this in E9/E12 by changing the mode to Cool just before the defrosting starts. Then immediately switch back to Heat and it continues to heat another 60 mins without defrost. This can be repeated indefinetly.

But anyway, good that you find this useful! I'm looking forward others to play with the IR remote comands and see if we can dig out something interesting.

ps. I used a simple and common IR receiver with Vcc, Gnd and data to record the signals. I powered the receiver just 4.5V battery. Data is connected to the serial port DCD pin and the software just polls for the pin and writes out timestamps.

[Paxmax]

Well... As for E9/E12 DKE (-3) it seems like extending the time between defrosting is a death sentance for the unit!
I'm sure you have read about the H16 problem?
Since the Pana units oil is really too thick in viscosity for our climate it gets stuck in outside evaporator.
That coupled with oil level being slightly too low effectivly kills the machine if you extend defrosting time.

Otherwize I agree with you, the defrosting times are retarded.

I'm also trying to attack the defrosting timer with manipulating the units temperature readings.
The unit has two sensors, outside temp and temperature of outside evaporator. By reporting +C temps to the pump defrostings are suspended. At the same time you have discret electronics measuring actual values and judging wheter defrosting is needed or not. I mean, at around 0C defrosting is heavily needed at every 40 minutes (Pana thought -"well 80 minutes will do", the evap is an icepack!

Then the other way around.... -9C outside... no friggin' ice and the units descides to defrost.... lunacy!
(Except for the oil returning to the compressor at defrosting)

So, what is needed before modifying defroster intervals is an improved oil return... somehow.
Before the oil return is fixed, it's highly likely any modifications (extending) of defroster time will have a negative impact of units life expectancy.

[apophis]

Uh oh! Thanks for this information! I have heard about the H16 problem, but I did not know these details about the compressor oil. Extending the defrost time has been tried in Finland with simple parallel resistor to pipe temp sensor. It works just fine and very long operation cycles can be achieved. I better share this warning at our forum...

I have also been investigating how to really measure the amount of frost in the evaporator. One idea I tried was to use optical measurement using IR led and phototransistor and measure the light reflectance from the frost. The results were really encouraging, as the frost accumulates, reflectance steadily increases. Using this with a mod in the pipe temp would give a possibility to defrost at optimal intervals

Another idea that was considered was to measure outdoor unit fan power consumption. As it decreases, more frost is accumulated. This would require a hall sensor in the fan power wire or something similar.

A simple solution for the usual bad defrosting algorithms would be to invite the far east engineer to Scandinavian countries in winter when there is -15 degrees outside and the air is dry. Temperature+time based defrosting would be ok, if it would be tuned also to lower temperatures than -7

[Paxmax]

Yeah, Pansonic does only admits to E9/E12 DKE units having a defective spring.
In secrecy they have been modifying some componenets to increase the oil return. They have added small drain holes to the fluid/gas separator to allow fluid to seep through slowly. Some oil collects in the separator and the drain holes would increase oil return.
Thing is, you never know which separator your particular unit is equipped with...

Adding oil helps. It makes sure the compressor can operate longer in "oil starvation mode" I.E. times when the outside evap is covered in frost and it is still some long time until defrost.
During manual water spray-on tests the evap temps dropped to -30 something in 25 minutes!
At those low temps the "68"-classed oil becomes more like wax in consistency.
So normally, oil would return slowly once the separator is filled up to the drain holes level. However, with the outside evap being 10-30 below freezing the oil begins to thicken considerably and sticks inside evap.

There is more about this in this forum "H16" part of Panasonic. The original H16 thread is about 1100 posts long!
Easiest is to study the "Oljeinjektering" thread.