@william1066 Thanks for sharing all you have learned so far. Documenting what you have done so others can follow sometimes nearly takes as long as doing it in the first place. But I think that this route will unlock a lot of potential for increased visibility and control over Samsung heatpumps, especially now with many people watching their electricity bills escalate. And obviously, all risk is assumed by the person who runs the code.

Do you need to run mbpoll initially or is this a step you carried out to verify the data transmission?

I think I have the same USB adapter, I bought it from the OpenEnergyMonitor store.

Good advice in relation to incrementally approaching the read, write, control actions. I will follow your advice on this.

I had started configuring InfluxDB but never completed it. I am going to follow the Everything Smart Home tutorial on YouTube to get this set up in advance of acquiring the modbus data.

Posted by: @redzer_irl

↑

Do you need to run mbpoll initially or is this a step you carried out to verify the data transmission?

My understanding is that you only get access "out the box" after installing the modbus interface module to the registers in table 2 in the manual (starting page 13).

If you want access to the "new modbus register table" which is table 3 on page 18 you need to "program" the mim module with the correct commands to send to the unit.

I think one of the messages sets enables registers that sends commands to the outdoor unit, and the other message set enables registers that sends commands to the wired remote.

I needed flow rate and outside temperature so enabled the additional registers right after installing.

@william1066 Ok thanks for that, I have a raspberry pi which I should be able to use for the mbpoll config

My MIM B19N board didn't arrive before the weekend so I'll have to wait until next week to start working on it.

I received my modbus board today and installed it.

The lcd on the board is flashing "00" and this is the message I get back when I run the first command in mbpoll.

Any ideas on what might be wrong?

I switched the USB over to my Home Assistant machine and am getting some data in. Great start!

I have now managed to program the new registers. I think the issue may have been down to the lack of a 120 ohm resistor on the modbus card initially. I said I would try without it but was getting frequent dropouts on the data in home assistant. 

Once I added the resistor, the data became rock solid. I tried mbpoll again and this time it worked first time.

@steven Thanks very much for posting your ESP config on github. I just copied it into mine and it worked first time 🙂

I'm using an ESP8266-12F pre-installed on this board with a builtin relay (the relay will be part of my controls). It's sat inside the indoor controllers case, powered from the indoor controller's 12V. The uart is this one from aliexpress.

Cabling is a 15 metre run of basic screwfix 0.75mm2,  5 core heating control cable, from indoor controller to outdoor unit. 2 cores for the indoor unit to outdoor unit comms, 2 cores for the modbus. no screening, no grounding. There's even a join 3/4 of the way along because the sparky who I asked to run it at the same time as the power cabling got the lengths wrong. So it seems very tolerant of cabling. I think the aliexpress uart has the 120 ohm resistor.

my ashp isn't plumbed to my CH yet (need a warm weekend to do cutover) - it currently can only pump water as far as the volumiser and back - but at least I can talk to it!

one question now based on these numbers - how real time and accurate are they? temperature and flow rate being the most obvious ones to want to get right and timely.

oh and where has everyone mounted their MIM-019's? there wasn't an obvious spot in the 16kw gen6 case as far as I could tell. so I used the "enclosed" plastic box that came with the MIM-019,  as that had a hole on that I could loop a cable tie through, from there to the same points as the cable ties that the controller cable clamps to. thus its "dangling" from that tie point. No issue with weight or stability but feels somewhat inelegant.

@iancalderbank

I had the same thoughts when wiring up my 8kw unit. I used a few cable ties also and tried to find somewhere that made sense. 

@redzer_irl yes, I tried the exact same location as you have done for mine as it seemed designed for the board, right size mounting socket. but the blue/brown data cable barely reached the main board socket, was under tension, so I didn't like that. must be a slightly smaller distance on the 8kw.

how are you guys calculating COP with the samsung? I don't see a modbus register for energy (input or output)?

are you doing

input= electrical energy measured by some other means (electric meter or CT clamp)

output = samsung modbus flow rate x samsung modbus DT

or something else?

@iancalderbank I am doing the calculations after I upload the data inputs to Emoncms.org

Input = Heat Pump Shelly EM CT clamp + Indoor Tank Shelly EM CT clamp

Output = Heat Pump Flow Rate (L/minute) x 0.16666666666666 (to get L/second) x 3.9 (using this as a guesstimate at the moment as I have some % glycol in the system) x Heat Pump Flow Delta Temp x 1000 (to get value in W)

I'm not getting a great overall COP but I'm going to monitor the data for a month and then see what I can do to improve things.

It is a renovated 1990s 2 story detatched house (A2 energy rating), 175 m² with ~60 m² UFH and new radiators. We got an air leakage result of <3.53 m3/hr/m2 and have MVHR.

I (possibly incorrectly) insisted on 3 zones - UFH, Downstairs and Upstairs. All that is to say, I was hoping for an SCOP of better than 2.5