THZ / LWZ Tecalor Stiebel Eltron Heizung

[belu]

@IMMI

Code Auswählen Erweitern

2015-01-07 14:07:24
sGlobal

outsideTemp: 0.3 flowTemp: 31.3 returnTemp: 25.4 hotGasTemp: 69.8 dhwTemp: 49.7 flowTempHC2: -60 evaporatorTemp: -3.4 condenserTemp: 27.7 mixerOpen: 0 mixerClosed: 0 heatPipeValve: 0 diverterValve: 0 dhwPump: 0 heatingCircuitPump: 1 solarPump: 0 compressor: 1 boosterStage3: 0 boosterStage2: 0 boosterStage1: 0 highPressureSensor: 0 lowPressureSensor: 1 evaporatorIceMonitor: 0 signalAnode: 0 evuRelease: 1 ovenFireplace: 0 STB: 0 outputVentilatorPower: 31 inputVentilatorPower: 30 mainVentilatorPower: 60 outputVentilatorSpeed: 24 inputVentilatorSpeed: 24 mainVentilatorSpeed: 17 outside_tempFiltered: 0.6 relHumidity: 0 dewPoint: 0 P_Nd: 3.68 P_Hd: 12.77 actualPower_Qc: 4949.592 actualPower_Pel: 1.743 collectorTemp: -60 insideTemp: 23.7

Wroks with  actualPower_Qc: 4949.592 actualPower_Pel: 1.743

But only a COP 2,83 (((

[godmorgon]

this is a philosophical/religius decision.

Many things in our lives are philosophical or debatable. Some of them are related to the efficiency (in comparision to the effectiveness).

The serial link (hardware) is not the root cause of the problem because there is no real data loss. Data loss usually occurs when the read process is slower than the data arrival rate. What I observe with fhem is the opposite - if I slow down the read process I get all the data.

Of course some users might experience data loss due to wiring or EMC issues.

The messages have a specific formatting allowing us to detect the start and the end of a message. Futhertmore, there is a checksum providing an indicator for the data integrity (its quality is debatable). You don't need hadware flow control to cleanly segment the bytestream into individual messages and to check its integrity. So you should be able to rely on the message formatting. If something does not fit then all collected bytes are discarded.

Let's do the math first.
1 byte is encoded in 10 bits (1 start bit, 8 payload bits and 1 stop bit)
9600 bps would ideally transfer 960 bytes per second. That is, 1 byte needs less than 1ms. The longest THZ/LWZ message is shorter than 100 bytes. So it should arrive at the receiver within less than 100ms. But it obviously does not. I guess nobody has ever analyzed the transmission timing at the link level. Possibly the THZ/LWZ does not transmit bytes back-to-back which would increase the transmission duration. Therefore, it might be better no to rely on micro-timeouts but rather use a bigger timeout for the DLE+ETX reception.

A serial link running at 115200 bps is 12 times faster. Nevertheless, the fhem code seems to have a timing issue. I cannot tell you exacly what goes wrong in the long chain of data exchage. But something definetly goes wrong.

If I find time I will try to propose a solution. My approach would be to have a separate thread collecting all bytes belonging to one message.

[godmorgon]

have you used my v0123 or something else?

I currently have no access to my heatpump either. So I did a quick command line test:

Code Auswählen Erweitern

$ perl -e "printf(\"%.3f\n\",unpack('f', pack( 'L',reverse(hex('45d8df9f')))));"
6939.953
$ perl -e "printf(\"%.3f\n\",unpack('f', pack( 'L',reverse(hex('3fc49ba6')))));"
1.536
$ perl -e "printf(\"%.3f\n\",unpack('f', pack( 'L',reverse(hex('45d76d7d')))));"
6893.686
$ perl -e "printf(\"%.3f\n\",unpack('f', pack( 'L',reverse(hex('3fcd9169')))));"
1.606
$ perl -e "printf(\"%.3f\n\",unpack('f', pack( 'L',reverse(hex('4600d58f')))));"
8245.390
$ perl -e "printf(\"%.3f\n\",unpack('f', pack( 'L',reverse(hex('3fd47ae2')))));"
1.660

So even if I assumed the first value to be in W and the second in kW the COP would be not very credible: 4.29 - 4.97. I wish I had such a COP during the winter season.

[immi]

If I find time I will try to propose a solution. My approach would be to have a separate thread collecting all bytes belonging to one message.

A separate process/thread for communication: It would be great. I would be happy to merge.

one more thing look at the length of your fb-response compared to mine

0100F9FBFDA8FFFC00B8009A021400A7FDA8FDA8FF8F00A82008110122012C02BC000F00160014000200000000010203CD40908DD33FA158111003
0100eafbfda8001601440108024901e6fda8fda8001c01176008110000000002bc000000000014001900000000014c05304600d58f3fd47ae207c500b1012a000001060144fff4ffde012300000000103e

immi

[godmorgon]

Now you cannot say the size does not matter! 
This is why I have more accesible registers in my heatpump.
However, I still have no clue where the values for humidity and hydralic pressure are hiding.

[godmorgon]

There's something fishy.
Your 0xfb message in the last comparision had the following hex values:

Code Auswählen Erweitern

$ perl -e "printf(\"%.3f\n\",unpack('f', pack( 'L',reverse(hex('3FA15811')))));"
1.261
$ perl -e "printf(\"%.3f\n\",unpack('f', pack( 'L',reverse(hex('40908DD3')))));"
4.517
The decoded values are represented rather differently:

Code Auswählen Erweitern

actualPower_Qc: 4949.592 actualPower_Pel: 1.743

[houseowner123]

Beheizte Fläche lt EnEV: 145 qm; 51 W/qm anno Primärenergiebedarf
Wandaufbau oder direkt den U-Wert: 0,147 W/qmK
dito für Dach: 0,178 w/qmK
Aufbau und U-Wert der Fenster und Außentüren: Alle Dreifachverglasung 0,7 w/qmK; eine Aussentür auch gedämmt Wert unbekannt.
Ergebnis des Blower Door Tests: 0.98 1/h
Stockwerke: 1, 5 Geschossig
Keller? wenn ja: Perimeterdämmung: Woraus und wie dick? Keller WU-Beton mit EPS 8cm; Treppenaugeninnenwände (innen und aussen) sind zusätzlich gedämmt mit 10 bzw 12 cm Glaswolle
Dämmung unter der Bodenplatte? wenn ja: Perimeterdämmung: Woraus und wie dick? 12 cm EPS nur Treppenauge
thermische Solarzellen zusätzlich verbaut? wenn ja: Details (Größe, Typ) nein
Holzofen? wenn ja: Details (Größe, Luft oder mit Anschluß an die FB-Heizung) Palena Olsberg 5 kW; kein Anschluss, nicht wasserführend
Ofen ist meist nur am WE an, oder wenn  Minusgrade draussen sind.

Hab kein so tolles Zusammenfassung-Bild wie ihr, alles einzeln: Meine Spreizung ist nur bei 2K. 
Steigung: 0,31 Fusspunkt: 1,5; Raumeinfluss: 80%; Flow Proportion: 30% Hyst 1: 2,8; Hyst 2: 2,4; Asymetrie: 3 K; Integral 100Kmin;  BoosterTimeoutHC:60 min; SummerMode Temp: 15°C; SummerModeHyst: 2K; BoosterTimeoutDWH:180 min; Bivalenz: -7°C

2014 Gesamtverbrauch inkl. Lüftung: 1975 kWh bei 4 Bewohnern.

(http://img5.fotos-hochladen.net/thumbnail/screenshot201cx671lvbw5_thumb.jpg)

[immi]

There's something fishy.
Your 0xfb message in the last comparision had the following hex values:

Code Auswählen Erweitern

$ perl -e "printf(\"%.3f\n\",unpack('f', pack( 'L',reverse(hex('3FA15811')))));"
1.261
$ perl -e "printf(\"%.3f\n\",unpack('f', pack( 'L',reverse(hex('40908DD3')))));"
4.517
The decoded values are represented rather differently:

Code Auswählen Erweitern

actualPower_Qc: 4949.592 actualPower_Pel: 1.743

Code Auswählen Erweitern


$message= "F9FBFDA8FFFC00B8009A021400A7FDA8FDA8FF8F00A82008110122012C02BC000F00160014000200000000010203CD40908DD33FA15811";

just add to line 1156 is decoded

Code Auswählen Erweitern


outsideTemp: -0.4 flowTemp: 18.4 returnTemp: 15.4 hotGasTemp: 53.2 dhwTemp: 16.7 flowTempHC2: -60 evaporatorTemp: -11.3 condenserTemp: 16.8 mixerOpen: 0 mixerClosed: 0 heatPipeValve: 0 diverterValve: 0 dhwPump: 0 heatingCircuitPump: 1 solarPump: 0 compressor: 1 boosterStage3: 0 boosterStage2: 0 boosterStage1: 0 highPressureSensor: 0 lowPressureSensor: 1 evaporatorIceMonitor: 0 signalAnode: 0 evuRelease: 1 ovenFireplace: 0 STB: 0 outputVentilatorPower: 29 inputVentilatorPower: 30 mainVentilatorPower: 70 outputVentilatorSpeed: 15 inputVentilatorSpeed: 22 mainVentilatorSpeed: 20 outside_tempFiltered: 0.2 relHumidity: 0 dewPoint: 0 P_Nd: 2.58 P_Hd: 9.73 actualPower_Qc: 4.517 actualPower_Pel: 1.261 collectorTemp: -60 insideTemp: -60


[houseowner123]

I cannot make much sense out of the new energy parameters when I plot them. Why the neg values? The lines are almost on top of each other. I wold expect at least double AUC from the produced energy. Does anybody else has see this too, in their readings?

(http://img5.fotos-hochladen.net/thumbnail/screenshot201e806ygpzsx_thumb.jpg)

[immi]

tom
please post which version you have
you have to update to v0.123

[houseowner123]

yep, it is version 0.123

[immi]

@IMMI

Code Auswählen Erweitern

2015-01-07 14:07:24
sGlobal

outsideTemp: 0.3 flowTemp: 31.3 returnTemp: 25.4 hotGasTemp: 69.8 dhwTemp: 49.7 flowTempHC2: -60 evaporatorTemp: -3.4 condenserTemp: 27.7 mixerOpen: 0 mixerClosed: 0 heatPipeValve: 0 diverterValve: 0 dhwPump: 0 heatingCircuitPump: 1 solarPump: 0 compressor: 1 boosterStage3: 0 boosterStage2: 0 boosterStage1: 0 highPressureSensor: 0 lowPressureSensor: 1 evaporatorIceMonitor: 0 signalAnode: 0 evuRelease: 1 ovenFireplace: 0 STB: 0 outputVentilatorPower: 31 inputVentilatorPower: 30 mainVentilatorPower: 60 outputVentilatorSpeed: 24 inputVentilatorSpeed: 24 mainVentilatorSpeed: 17 outside_tempFiltered: 0.6 relHumidity: 0 dewPoint: 0 P_Nd: 3.68 P_Hd: 12.77 actualPower_Qc: 4949.592 actualPower_Pel: 1.743 collectorTemp: -60 insideTemp: 23.7

Wroks with  actualPower_Qc: 4949.592 actualPower_Pel: 1.743

But only a COP 2,83 (((

HI Belu
actualPower_Pel: 1.743 is maybe ok but 4949.592 is not ok. 
you have probably a very new heatpump like godmorgon with different encoding of this parameters
Let us wait for feedback from more people...
I start to think that we have to fork like we did for Peter:
very old heapump like Peter -->firmware version 2.06
medium heatpump like me -->firmware version 4.39
very new heatpump like godmorgon --> firmware version 5.39 and ID 7278

let us see
immi

[immi]

Hi tom
can you post
get Mythz sGlobal


if you updated in the afternoon, you see only big values in the plot, which come from before the update

The  right values are much smaller like actualPower_Qc: 5.332 actualPower_Pel: 1.359

immi

[belu]

@ MMI
have a thz 304 5.19 software version 316
The COP I have derived from the upper values are indeed identical values other. My COP is bad ..

2,8 COP...

Wroks with  actualPower_Qc: 4949.592 actualPower_Pel: 1.743

But only a COP 2,83 (((

or
2,8 COP...

sHeatHCDay 8870 Wh / sElectrHCDay 2968 Wh = HeatDay_COP = 2,989

[immi]

@ MMI
have a thz 304 5.19 software version 316
The COP I have derived from the upper values are indeed identical values other. My COP is bad ..

Belu
you have a  problem in your actualPower_Qc: 4949.592 actualPower_Pel: 1.743
actualPower_Qc is 3 order of magnitude different from actualPower_Pel.

Can you see these value in your display?