Thanks. I added a row for your unit to the compatibility matrix in GitHub.
Do your units communicate with indoor via S1/S2 bus or LNS communication? I also have one Midea DLCSRB unit that uses LNS and that does not have a 5v TestPort connection in the same place as the light commercial units that use S1/S2 (for those units the TestPort is on the small circuit board right under the handle assembly where power etc comes in).
Where is the TestPort in your units? Just wondering if my LNS communicating unit might have a TestPort on the main control board also, before I go through the hassle of opening it.
Awesome projectāinterested in contributing!
Questions
- Does it seem likely from the data below that your project might work for my unit? I know the info I have is a little limited, just wondering if you think itās worth it to depower the unit and pry further.
- Do you have a photo of the dangerous capacitors on your board? My plan was to open the breaker, put on some PPE, disconnect the mains, wait 10 mins, and then measure the voltage across the largest caps I could find on the board
. The board is upside down though so Iāll have to pull it off to even get at it.
Details
Me: Mech Eng by training. I have done a fair bit of microcontroller programming in my day but little proper EE knowledge aside from a few small hobby projects.
My heat pump: Carrier 38MARBQ24AA3 (outdoor unit)
LLM assessment of the suitability of your project for my unit: āYour outdoor unit is Midea-manufactured hardware under a Carrier badge. Mideaās own OEM parts catalog lists this SKU under its own part-number system.ā
Attached some photos that may or may not be useful. I didnāt pull apart the outdoor unit further for more useful photos yet, but if you think thereās a decent chance this would work Iāll do so.
Hi - unfortunately, I donāt see a diagnostic port on the pictures you uploaded (neither on the PCB nor the wiring diagram). The wiring diagram will show the port if itās available:
Regarding the cap: This video shows the location of the capacitor on the ODU control board: https://www.youtube.com/watch?v=poEmSZnrnjs&t=70s
If your unit doesnāt have a diagnostic port: The same information + additional IDU parameters can be extracted from the signal line that connects the IDU + ODU: Reverse Engineering Mideaās LNS Communication Protocol
However, that approach means interfacing directly with line voltage. On 220V units the signal line is referenced against one of the hot legs. This also requires a slightly more complex circuit. In the interest of not electrocuting myself, I abandoned the idea of tapping into those lines.
I see some of the pins labeled S1 / S2:
There is a separate effort to tap into the S1S2 communication: Reverse Engineering Senville/Midea S1S2 Bus
@fmckee @kydonian
I donāt think those are the same as the S1S2 bus Ive been decoding. I think Ive seen threads talk about the PQE bus before but unsure where to point you to.
Maybe if you check the screw terminal voltages to see if anything is active on them you can see about using them. If they are dead it might be because you are using a different Indoor to outdoor comm bus and they are locked out. It also looks like you are possibly using 24V communication for the indoor thermostat so that might end up being a dead end for a connection.
fmckee
As a side note the HAHB bus on my unit is insane. Fully carries XYE and S1S2.
Based on clauding only I suspect there will be a separate board with the 5V test port.
āIPM (Intelligent Power Module) ā¦ bundles the IGBTs and their gate-drive circuitry together ā¦ā
āā¦official Midea manual for their handheld diagnostic tool ā¦ directly names the connector. ā¦includes an item called āTestPort connectionā; ā¦lines up with the other Midea manual ā¦ that labels the connector
TESTPORTon the PCB itselfā¦ā
(Manuals imply the TESTPORT is on the IPM.)
Iāll open it up in the next week or so and see what I can see. Iāll look for a 4-pin JST XH connector labeled ātest portā or āCN3ā or something, and if I do put a lead on it, turn it back on, and see if I can measure 3.3VDC across two of the pins.
(Full disclosureājust had a baby and working on this project intermittently. Diaper changes, snugs, opening up my heat pump, just normal pat leave stuff.)
@fmckee
Ive noticed your āOperating Modeā changes when idle. Do you think this is caused by the Ecobee setup. I would think it would be the same while the unit is idle. If you still have the Ecobee setup and it setting the unit to āOFFā mode, do you see your Indoor Ambient Temp drop?
Saw that in the other thread. Related: Itās getting increasingly difficult for me to keep track of this Midea protocol multiverse:
- UART
- XYE
- S1S2
- HAHB
- Diagnostic Port
- LNS
We should put together some high-level documentation covering each protocol:
- Where is it used?
- What information does it carry?
- Can it be used to change settings, or is it read-only?
- How do you tap into it?
I have a pretty good grasp on LNS and the diagnostic port and some understanding of UART, but my understanding of XYE / S1S2 / HAHB is lacking.
Itās 5V. Just FYI in case you plan to connect it to an ESP w/o level shifter.
Ah, yes, that will not help with the timely completion of hobby project 
Yes, thatās caused by my contraption: It switches between COOL and FAN mode all the time.
The opposite: I see an immediate bump:
There is definitely something weird going on with the set point:
It fluctuates up and down (top), but those fluctuations donāt line up with the switching between COOL and FAN mode (bottom). Almost looks like the last bit carries a different piece of information.
BTW: I ordered a small batch of PCBs that I expect sometime this week. This will let me put together a dongle for the diagnostic port pretty quickly. Happy to assemble a spare (with all electronics, adapter cable, and enclosure) for folks who want one. $20 including shipping (continental US) would be enough to cover my costs. Drop me a PM if you are interested.
High-level documentation would be a great thing to have. I have a good understanding of the S1S2 bus Iām sniffing but Iām noticing other headers contain the S1S2 references so it gets unclear what S1S2 actually means. Iām starting to think S1S2 is just the physical terminal labels on my unit, not the protocol itself.
On your setpoint - I havenāt noticed a difference in setpoint vs mode on my unit. Maybe something to do with Heat vs Cool mode offsets.
On mode - Iād think thereās 2 different modes even on the diagnostic port, IDU mode and ODU mode. On my unit theyāre slightly different depending on whatās happened. The ODU has many overrides as fail-safes, thereās mode switching timeouts, and when the heatpump is in defrost the IDU shows Heat while the ODU shows Defrost. Maybe the mode you see is the ODU mode.
Also if you ever turn the unit specifically to OFF mode, let me know if your indoor temp shifts by 8.5C-ish. Something interesting in my system.