search

arendst / Tasmota

Alternative firmware for ESP8266 and ESP32 based devices with easy configuration using webUI, OTA updates, automation using timers or rules, expandability and entirely local control over MQTT, HTTP, Serial or KNX. Full documentation at
https://tasmota.github.io/docs
GNU General Public License v3.0
22.08k stars 4.78k forks source link

Showing periodically wrong power data when using SML (energy meter) #10007

Closed Jambola77 closed 3 years ago

Jambola77 commented 3 years ago

PROBLEM DESCRIPTION

Showing wrong power amount when using SML. I got periodically wrong power data (SML 10.70.00), please see screenshots. As the data changes every second, it is hard to figure out what is wrong. Maybe some SML messages are not correct and these should be ignored. For 99% the display is correct.

REQUESTED INFORMATION

Make sure your have performed every step and checked the applicable boxes before submitting your issue. Thank you!

- [ ] If using rules, provide the output of this command: `Backlog Rule1; Rule2; Rule3`:

Rules output here:

- [x] Provide the output of this command: `Status 0`:

STATUS 0 output here: 11:25:58 MQT: stat/ESP07_Module/STATUS = {"Status":{"Module":18,"DeviceName":"ESP-07","FriendlyName":["ESP-07"],"Topic":"ESP07_Module","ButtonTopic":"0","Power":0,"PowerOnState":3,"LedState":1,"LedMask":"FFFF","SaveData":1,"SaveState":1,"SwitchTopic":"0","SwitchMode":[0,0,0,0,0,0,0,0],"ButtonRetain":0,"SwitchRetain":0,"SensorRetain":0,"PowerRetain":0}} 11:25:58 MQT: stat/ESP07_Module/STATUS1 = {"StatusPRM":{"Baudrate":115200,"SerialConfig":"8N1","GroupTopic":"tasmotas","OtaUrl":"http://ota.tasmota.com/tasmota/release/tasmota-DE.bin","RestartReason":"Software/System restart","Uptime":"1T21:59:11","StartupUTC":"2020-11-27T12:26:47","Sleep":50,"CfgHolder":4617,"BootCount":103,"BCResetTime":"2020-10-15T16:13:55","SaveCount":539,"SaveAddress":"FA000"}} 11:25:58 MQT: stat/ESP07_Module/STATUS2 = {"StatusFWR":{"Version":"9.1.0(tasmota)","BuildDateTime":"2020.11.27 09:52:16","Boot":31,"Core":"2_7_4_5","SDK":"2.2.2-dev(38a443e)","CpuFrequency":80,"Hardware":"ESP8266EX","CR":"400/699"}} 11:25:58 MQT: stat/ESP07_Module/STATUS3 = {"StatusLOG":{"SerialLog":2,"WebLog":3,"MqttLog":0,"SysLog":1,"LogHost":"","LogPort":514,"SSId":["xxx",""],"TelePeriod":60,"Resolution":"558180C0","SetOption":["00008009","2805C8000100060000005A00000000000000","01000000","00006000","00000000"]}} 11:25:58 MQT: stat/ESP07_Module/STATUS4 = {"StatusMEM":{"ProgramSize":633,"Free":368,"Heap":15,"ProgramFlashSize":1024,"FlashSize":8192,"FlashChipId":"1740EF","FlashFrequency":40,"FlashMode":3,"Features":["00000407","87DAE797","043E8001","000007CF","000013C4","C000F989","0000400C","00000000"],"Drivers":"1,2,3,4,5,6,7,8,9,10,12,16,18,19,20,21,22,24,26,27,29,30,35,37","Sensors":"1,2,3,4,5,6,7,8,9,31,53,65"}} 11:25:58 MQT: stat/ESP07_Module/STATUS5 = {"StatusNET":{"Hostname":"ESP07","IPAddress":"xxx","Gateway":"xxx","Subnetmask":"255.255.255.0","DNSServer":"xxx","Mac":"18:FE:xxx","Webserver":2,"WifiConfig":4,"WifiPower":17.0}} 11:25:58 MQT: stat/ESP07Module/STATUS6 = {"StatusMQT":{"MqttHost":"192.168.1.100","MqttPort":1883,"MqttClientMask":"DVES%06X","MqttClient":"DVES_9FB3F8","MqttUser":"DVES_USER","MqttCount":2,"MAX_PACKET_SIZE":1200,"KEEPALIVE":30}} 11:25:58 MQT: stat/ESP07_Module/STATUS7 = {"StatusTIM":{"UTC":"2020-11-29T10:25:58","Local":"2020-11-29T11:25:58","StartDST":"2020-03-29T02:00:00","EndDST":"2020-10-25T03:00:00","Timezone":99,"Sunrise":"08:20","Sunset":"16:56"}} 11:25:58 MQT: stat/ESP07_Module/STATUS10 = {"StatusSNS":{"Time":"2020-11-29T11:25:58","SI7020":{"Temperature":27.9,"Humidity":31.0,"DewPoint":9.2},"BMP280":{"Temperature":28.6,"Pressure":1028.6},"CCS811":{"eCO2":626,"TVOC":34},"WP":{"WP_Total":150.3106,"WP_HT":67.0551,"WP_NT":83.2555,"WP_Total_Liefer":0.0014,"WP_HT_Liefer":0.0013,"WP_NT_Liefer":0.0001,"WP_WL_P1":62.3,"WP_WL_P2":4.4,"WP_WL_P3":653.5,"WP_AktLeist":64.8},"Strom":{"StromBezug":2063.5,"StromLiefer":10401.9,"AktLeist":1304.0},"PressureUnit":"hPa","TempUnit":"C"}} 11:25:58 MQT: stat/ESP07_Module/STATUS11 = {"StatusSTS":{"Time":"2020-11-29T11:25:58","Uptime":"1T21:59:11","UptimeSec":165551,"Heap":17,"SleepMode":"Dynamic","Sleep":50,"LoadAvg":20,"MqttCount":2,"Wifi":{"AP":1,"SSId":"xxx","BSSId":"xxx","Channel":11,"RSSI":100,"Signal":-31,"LinkCount":2,"Downtime":"0T00:00:07"}}}



### TO REPRODUCE
Problem arises periodically during normal operation

### EXPECTED BEHAVIOUR
Wrong SML messages should be ignored.

### SCREENSHOTS
![Bildschirmfoto vom 2020-11-29 11-19-43](https://user-images.githubusercontent.com/55683273/100539363-37292c00-3236-11eb-8afb-af2ff2c6caa6.png)
![Bildschirmfoto vom 2020-11-29 11-17-58](https://user-images.githubusercontent.com/55683273/100539366-37c1c280-3236-11eb-9980-66089956b97c.png)

### ADDITIONAL CONTEXT
_Add any other context about the problem here._

**(Please, remember to close the issue when the problem has been addressed)**
Jambola77 commented 3 years ago

I am using 9.1.0. Normally the "Strom Akt.Leistung" (Actual power consumption) is not greater than 4000W.

kugelkopf123 commented 3 years ago

How about posting your script code, too?

Jambola77 commented 3 years ago

Here is the script:

>D  
>B  
->sensor53 r
>M 2  
+1,14,s,0,9600,WP
1,77070100010800ff@1000,Total,KWh,WP_Total,4 
1,77070100010801ff@1000,HT,KWh,WP_HT,4  
1,77070100010802ff@1000,NT,KWh,WP_NT,4
1,77070100020800ff@1000,Total_Liefer,KWh,WP_Total_Liefer,4 
1,77070100020801ff@1000,HT_Liefer,KWh,WP_HT_Liefer,4  
1,77070100020802ff@1000,NT_Liefer,KWh,WP_NT_Liefer,4
1,77070100240700ff@1,P1_AktLeist,W,WP_WL_P1,1
1,77070100380700ff@1,P2_AktLeist,W,WP_WL_P2,1
1,770701004c0700ff@1,P3_AktLeist,W,WP_WL_P3,1
1,77070100100700ff@1,AktLeist,W,WP_AktLeist,1
+2,2,s,0,9600,Strom
2,77070100010800ff@1000,Bezug,KWh,StromBezug,1
2,77070100020800ff@1000,Lieferung,KWh,StromLiefer,1 
2,77070100100700ff@1,Akt.Leistung,W,AktLeist,1
#
kugelkopf123 commented 3 years ago 
>D  
>B  
->sensor53 r
>M 2  
+1,14,s,16,9600,WP
1,77070100010800ff@1000,Total,KWh,WP_Total,4 
1,77070100010801ff@1000,HT,KWh,WP_HT,4  
1,77070100010802ff@1000,NT,KWh,WP_NT,4
1,77070100020800ff@1000,Total_Liefer,KWh,WP_Total_Liefer,4 
1,77070100020801ff@1000,HT_Liefer,KWh,WP_HT_Liefer,4  
1,77070100020802ff@1000,NT_Liefer,KWh,WP_NT_Liefer,4
1,77070100240700ff@1,P1_AktLeist,W,WP_WL_P1,1
1,77070100380700ff@1,P2_AktLeist,W,WP_WL_P2,1
1,770701004c0700ff@1,P3_AktLeist,W,WP_WL_P3,1
1,77070100100700ff@1,AktLeist,W,WP_AktLeist,1
+2,2,s,16,9600,Strom
2,77070100010800ff@1000,Bezug,KWh,StromBezug,1
2,77070100020800ff@1000,Lieferung,KWh,StromLiefer,1 
2,77070100100700ff@1,Akt.Leistung,W,AktLeist,1
#

Try this!

Jambola77 commented 3 years ago

Ok, I will try it with the flag=16.

Jambola77 commented 3 years ago

I have now inserted the 16, but now I do not get negativ values. As I have a photovoltaic system some values can become negativ. Here is a sample of two SML telegrams:

12:48:20 : 77 07 01 00 10 07 00 ff 01 72 62 01 62 00 62 1b 52 fe 53 aa 0c 01 01 01 63 e6 d3 00 76 05 b9 03 4e 09 62 00 62 00 72 63 02 01 71 01 63 a9 c1 00 00

12:48:21 : 77 07 01 00 10 07 00 ff 01 72 62 01 62 00 62 1b 52 fe 53 a9 59 01 01 01 63 8b 9e 00 76 05 bc 03 4e 09 62 00 62 00 72 63 02 01 71 01 63 fa 4c 00 00

Maybe the flag=16 does ignore negativ values or make them positive?

kugelkopf123 commented 3 years ago

@Jambola77 Good Question. Flag 16 is a Median on the values you read. But it should work with negative values as wel. Perhaps @gemu2015 knows something more about it.

gemu2015 commented 3 years ago

it should work with negative values. however it may have a delay of up to 5 sequences because it is a 5 value median filter. e.g. if your meter emits values every 3 seconds it may take up to 15 seconds to update.

since my photovoltaik feeds a large battery i have no more negativ values (at least in winter time) so i can't test now

it is however strange that you have wrong values only on one entry. is it really only that specific sml code that fails ? (2,77070100100700ff@1,Akt.Leistung,W,AktLeist,1)

99999 means the type of entry wasn't found

if i would silently discard this result we could not recognize that something goes wrong

try to use a hardware serial for this faulty meter interface.

using 2 software serial ports may introduce errors sometimes

Jambola77 commented 3 years ago

Hello gemu2015, I have now switched from flag=16 back to flag=0 and did a reboot. But I cannot see any negative values, only wrong values. 600W equals approx. -28W in reality.

Bildschirmfoto vom 2020-12-01 11-22-08

kugelkopf123 commented 3 years ago

@Jambola77 How about gemu2015 other questions? And what is the model of your meter? Have you tried switching to HW Serial?

I have now switched from flag=16 back to flag=0 and did a reboot. But I cannot see any negative values, only wrong values. 600W equals approx. -28W in reality. Then it seams that it has nothing to do with the median.

Jambola77 commented 3 years ago

The model of my meter is DWS7412. No, I have not switched to serial. How to switch to HW_Serial? Using the RX Pin (GPIO3) and select "Serial RX" in the template? Or leave the GPIO3 pin as "none"?

gemu2015 commented 3 years ago

the dws74 has a firmware bug! use the define in the source code to fix it. search these lines and uncomment the define

// addresses a bug in meter DWS74 //#define DWS74_BUG

Jambola77 commented 3 years ago

Yes, I have already done this.

kugelkopf123 commented 3 years ago

The model of my meter is DWS7412. No, I have not switched to serial. ...

@Jambola77 Do Not set anything in the device settings. Everything you need are you doing in the scripter. Just use the rx pin and set your meter definition to gpio 3 in place of 2. When the rx pin is connected you aren’t able to flash your device. Hold this in mind when Trying to flash it.

gemu2015 commented 3 years ago

i am not sure if every DWS has this firmware bug. did you try both variants ? i ask again. is this bug only with the current value ? (Akt.Leistung) if yes it is most likely not a serial port issue. serial issues would also affect the other values.

Jambola77 commented 3 years ago

I will try both versions. Hopefully tomorrow we have more sun, otherwise I cannot test this. :-)

It only shows up with this value, as this is the only one which can become negative.

gemu2015 commented 3 years ago

so it is most probably not a serial issue would be helpful to have an SML telegram for a positive and negative value

of type 2,77070100100700ff@1,Akt.Leistung,W,AktLeist,1

Jambola77 commented 3 years ago

12:48:20 : 77 07 01 00 10 07 00 ff 01 72 62 01 62 00 62 1b 52 fe 53 aa 0c 01 01 01 63 e6 d3 00 76 05 b9 03 4e 09 62 00 62 00 72 63 02 01 71 01 63 a9 c1 00 00

12:48:21 : 77 07 01 00 10 07 00 ff 01 72 62 01 62 00 62 1b 52 fe 53 a9 59 01 01 01 63 8b 9e 00 76 05 bc 03 4e 09 62 00 62 00 72 63 02 01 71 01 63 fa 4c 00 00

I think they are both negative.

gemu2015 commented 3 years ago

these values are both in the same polarity . (aa0c, a959) with DWS bug repair they both report positive values without DWS bug repair negativ values

i need a positive AND a negative value telegram example

Jambola77 commented 3 years ago

Here is a sample of a positive value: 77 07 01 00 10 07 00 ff 01 72 62 01 62 00 62 1b 52 fe 53 0b e6 01 01 01 63 5f ee 00 76 05 fd 81 55 09 62 00 62 00 72 63 02 01 71 01 63 5d 24 00 00 this should be a negative value: 77 07 01 00 10 07 00 ff 01 72 62 01 62 00 62 1b 52 fe 53 f9 67 01 01 01 63 fa a3 00 76 05 56 86 55 09 62 00 62 00 72 63 02 01 71 01 63 22 73 00 00

Both without DWS bug repair.

gemu2015 commented 3 years ago

the dump function shows raw values so the bug repair is not involved

your meter does NOT have this bug!

results are 30,46 and -16,89

so you should get negative values as desired

if you still have occasionally wrong values, try with median filter. it definitely works with negative values

or try the hardware serial port

Jambola77 commented 3 years ago

I have now changed the script to use flag=16 and I have no more spikes anymore. Negative values are also present. Thanks a lot for the support!!

I will close the issue.

Jambola77 commented 3 years ago

Problem still exists with 9.3.0. I will try with 9.3.1.

Jambola77 commented 3 years ago

sml.txt Problem still exists with 9.3.1 and median filter. grafik

My photovoltaic system currently outputs about 5,100 W.

gemu2015 commented 3 years ago

ok, checked that ALL these meters have the bug in consumption. https://wiki.volkszaehler.org/hardware/channels/meters/power/edl-ehz/dzg_dvs74 i only catched that bug with 16 bit values but they also have it on 24 bit and 32 bit values

can you please provide some sml consumption dumps and the corresponding power values.

i only need code lines starting with 77 07 01 00 10 07 00 ff

Jambola77 commented 3 years ago

Please find attached some more sml dumps. Values should be about -5000W. sml2.txt

gemu2015 commented 3 years ago

please try this version, should fix the bug.

xsns_53_sml.ino.zip

Jambola77 commented 3 years ago

nitems is not declared (line 2197). Do I need more files? (I tried to compile it in 9.3.1)

Edit: Will try it again with development branch.

Jambola77 commented 3 years ago

Looks great! grafik Thanks a lot for the fast bugfix!