Controlling a DC inverter machine

[ham789]

If it's standard RS-232 protocol, I have a TEK 308 logic analyzer.
Has a serial mode for RS-232.

[Acuario]

You'll probably need to build a level converter interface (something like in my schematic using MAX232 and MC1489) to interface to the 5V logic - RS232 is +/-12V signalling so you can't connect directly from the RS232 to the machine.

Actually, with the analyzer it may not matter as it may be able to work at 5V.

You will need to trace the circuit though as I wouldn't advise connecting anything to the wires that go from the internal to external units as they are most likely at line voltage.

Nigel

[AC_Hacker]

Originally Posted by Acuario

You will need to trace the circuit though as I wouldn't advise connecting anything to the wires that go from the internal to external units as they are most likely at line voltage.

That was what I was trying to determine when I read the potential from ground to the data wire with my multimeter.

The DC potential was 14.77 volts and steady and the AC was not steady and about the 0.2 volt level.

I have a boat anchor scope that I will double check the situation with today.

-AC

[Xringer]

I think it's likely an RS232 level.. Or maybe a modified version that runs 0.0v to +15 volts.?.
I've seen a few of those. I've also seen some devices that had 0.0v to +9 (even +5)
and they worked just fine, talking to 1488 & 1489 chips..

The RS-232 standard defines the voltage levels that correspond to logical one and logical zero levels for the data transmission and the control signal lines. Valid signals are plus or minus 3 to 15 volts; the ±3 V range near zero volts is not a valid RS-232 level. The standard specifies a maximum open-circuit voltage of 25 volts: signal levels of ±5 V, ±10 V, ±12 V, and ±15 V are all commonly seen depending on the power supplies available within a device

[Acuario]

I've had a couple more hours free to play a bit. I've now tidied up my program and removed loads of duplicated stuff, I've also cracked what is going on in some more of the frames.

I can now reliably communicate with the machine (it was a bit flakey before) and turn it on/off in any mode I want. I've also found the bytes that report the status of the compressor and fan. I've attached a photo of the main screen of my program.

I've ordered a PIC prototyping board that has a 16F877 processor, display, keypad, RS232, Temperature sensor and more off of ebay. My plan is to use this as my controller (no point building it from scratch as the built board is only $42 including shipping.

Nigel

[cadillackid]

This is some fun stuff!! one of the members here found me on a Prius forum.. 3 years ago i did a lot of work on this with CHIGO branded minisplits to use in my Home automation system...

CHIGO is sold under Grunaire, amcoraire,turboair and other names.. These units actually have a connector for a wired control that can be soldered to on their main board...

unlike Daikin, or mitsubishi these things are not designed to run in a networked environment .. whereas daikin and mitsi have their own networks and computer control interfaces available for purchase..

my original project was so that I could control these units in a non standard way for z ducted zoned HVAC system in my house.. I not only wanted 1 direction control but I wanted to get feedback from the indoor unit;s room temp sensor... as my idea was to "feed the unit" its indoor temp from a micro controller..

I ended up using zbasic.net micros and creating a control board that has run them for 3 years now.. I never got into the indoor to outdoor comms much.. although I did get frames and start decoding them by picking up the signals to and from the indoor board right as they came off of the micro and before they went through the optos and superimposed on a higher voltage (30 VAC) line.

I can surely share any information you guys are interested in.!

couple things i found with the chinese units.. one is that the comms protocol really attempts to match IR remote streams.. they use the same algorithms and just apply the characteristics to wired lines ..TTL levels at the micros themselves. (in my case ATMEGA8 on Chigo units)...

the data streams were sent and then sent again inverted for integrity checking...

ie 00000111 and then right after 11111000.

they used a Pulse distance encoding style...

to start I'd get a 20 ms low... then a 5 ms high... then the data..

a "1" was a 2.1 ms high... then a .768 ms low (spacer), a 0 was then a .768 ms high then a .768 ms low (spacer) and so on.. 8 bits per word in this manner.. my units sent 5 bytes.. and the same 5 bytes inverted for a total of 10 bytes...(the byte and then its inversion right after)..byte-1...byte-1 inverted... byte-2.. byte-2 inverted etc

each byte being bit mapped such as 3 bits of 1 byte for the mode, 5 bits of one byte for say the setpoint, 2 bits of another for swing, etc.. I mapped out most of the registers.. but many i didnt care about...

im pretty sure each manufacturer uses a different algorithm.. but I do know that the IR remote commands for many of the chinese units are similar in how they do it just different bit maps for the commands..

I also was succesful in learning the LONG IR remote commands and controlling them with a global cache GC-100 box.. I did that for a friend who didnt want to go through building boards and the like...

the physical installation of my units is such is completely custom in that I disassembled the indoor units and built my own coil boxes and installed them in the ducting of my home behind zone control dampers.. one unit for each of 3 zones.. one central blower.. BAPI-HVAC controls static pressure and air velocity sensors allow me to control the VFD running the blower to maintain proper and desired CFM across each coil.. I allow the units to do their own thing to maintain their own coil temps and such.. I send the units their setpoints and mode.. and i use a digital pot to skew the indoor room temp sensor.. as Mini-splits were designed to be installed high up on the walls they are designed to turn on and off at wildly varied temps.. heat mode shuts off when its room sensor is 3 deg C above setpoint.. cooling at 2 deg C below setpoint.. that wouldnt fly in a central system.. so I fake the room temp based on actual to make the unit think differently... the digital POT faskes the thermistor and I read back what the unit thinks its indoor temp is via the comms link to adjust properly..

for 3 years now the system has run very well except for defective Capacitors on the PFC boards of the units.. easy fix to replace the MOSFETs and caps that died in the first 6 months of service.. no more issues after that..

I havent really tried to measure the COP on the heatpumps but I do know since I put the system in my comfort level went up and my bills went way down from the 2004 10 SEER A/C that was installed when the house was built so I let that speak for the success..

the HVAC is just part of a complete home automation system running on a mixture of windows (Homeseer) and some custom linux stuff..

hope that helps you guys out a bit.. I love this kind of stuff its a BLAST!
-Christopher

[AC_Hacker]

Acuario,

I'm about to build the interface as in your schematic or similar.

I found this part (DS275)...

...and I was wondering if it would be suitable for this project?

Best,

-AC

[Acuario]

In theory it should be fine. I did have problems trying to use 1 chip for both TX and RX though, that's why I reverted to 2. Try it and see - if it doesn't work try a MC1489 on the RS232 TXD leg.

Nigel

[ham789]

Originally Posted by AC_Hacker

Acuario,

I'm about to build the interface as in your schematic or similar.





...and I was wondering if it would be suitable for this project?

Best,

-AC

That probably works if you really have RS-232 on the other end.

You have a scope, what voltages did you measure on the port when
transmitting? What's the quiescent voltage when not transmitting?

What part of the problem are you trying to solve?
If it's serial data and you want to look at it, why not hook it to
a real serial port? I understand the trepidation of hooking your
$2,000 laptop to an air conditioner, but you can buy laptop
computers at garage sales for less than the shipping cost of the chip.
Ditto for PalmIII PDA's with serial cradles. The newest PDA's don't
have serial ports, so stick with the old stuff with a serial cradle.
FREEWARE comm sw
is available for both.

Don't be put off by the price sticker. It's not unusual to get old
laptops marked $50 for a buck by the end of the day.

I've interfaced a bunch of computer related stuff.
A diode will connect a PIC processor to RS-232, you just have
to set the compiler directive to invert the data.
A pair of transistors makes a great inverting interface.
0-5V works fine on most serial inputs with short cables.
Never had that not work, but I've never interfaced an air conditioner.

You may be able do do something interesting with the IR port.
Like receive the remote and send it out the serial port.

Make it work with a computer interface, then when you're done,
convert it to a microcontroller.

[cadillackid]

In my case I used a Logic analyzer to pick off the Signals on my units..

if you look at the boards of your unit and are trying to interface to the indoor-outdoor Comms link that all ends up at 5 volt TTL levels to hit the micro when all is said and done.. past that it is put through transistors and optos and boosted up onto a higher voltage carrier to deal with the crazy EMI around a VFD motor drive, and also to beter handle surges that occur on a line...

since most of these units arent capable of calculating LRC or CRC checks on their data streams they either rely on it to be correct all the time or invert it after the initial send and if they dont have a match. discard the record...

the Logic analyzer I use is a Salae LOGIC and it will capture data for you so you can go back and analyze what you are getting across the comms line...

the other thing to remember is that these units do electronically what traditional units do mechanically or with temperature change, etc.. they are engineered pretty tight... if you have the inclination to directly control your compressor drive, expansion device opening, outdoor and indoor fan speeds then you are a brave one..

having worked on both computers and HVAC i can honestly say the algorithms inside these units are pretty much dead on as far as the mechanics and thermodynamics of it all...

they know more than you think they do.. my units for instance have 5 temp sensors outside... 3 on the coil, an outdoor ambient, and a compressor disscharge.. and 4 on the indoor.. 3 on the coil and room temp...

the VFD board provides feedback to the current being pulled by the motor and the voltages at the motor.. so now with all that info the unit can pretty much calculate the line pressures low and high...

in playing with them they will do some wild stuff to keep themselves running.. ie kill the indoor fan and it floods the evaporator and slows the outdoor fan, raising the head pressure and evaporator pressure so it wont freeze up. eventually it shuts down as it knows something is wrong after 5 or 6 minutes of freeze protection...

kill the outdoor fan and it ramps down the compressor and indoor fan in hopes it may still gain some cooling.. after a couple minutes it shuts off if its hot enough out..

as I mentioned in an earlier post, getting feedback as to what the unit is doing wont be possible via the IR port.. the IR is ONE way on most all A/C units.. it sends a LONG code each time a button is pressed... ALL parameters are sent in each remote command.. mode, temp.. fan speed, swing, pitch, etc are sent at each remote press to keep the remote in sync with the unit..

if you take the remote into a closet and jack up all the settings. then walk out to the unit and press just ONE button, the unit will then take on ALL of the settings the remote has on it.....

cool stuff!
-Christopher