Emoncms.org - Changing Input device, is it possible to link a new input to previous feed?

Currently all the inputs from my EmonTxV3 are posted to Emoncms.org via 433MHz to my Heapump board, which is connected via WiFi to Emoncms.org.

Today, i will be adding a wifi module to my EmonTxV3 so that it can send its values directly to Emoncms.org independantly of the heatpump board. The reason for this is so that i can run the Continuous Monitoring firmware on the EmonTxV3, which would otherwise not work with the Heatpump board in the middle.

So my question is…

When the EmonTxV3 is changed and the new inputs detected, can i somehow setup the new input to write its data to the existing feeds so they carry on?

Or even take the data from the old feeds and add them to new feeds for the new inputs?

May have answered my own question.

When setting up a new feed for an input, you have the option to select existing feeds (image below). I assume this will allow the feed to continue its data collection from a new input, as i need to do ?

(Curious, what happens if you log two inputs to the same feed :thinking: )

If the new input feeds a new feed that feeds an existing feed, that would appear to be what you want.
If new data arrives from the same NodeID and with the same name (if it’s JSON) or in the same place in the data stream (if it’s unnamed), then it will go into the existing Input and thence into the existing feed.

For a variable input “TIMESERIES” feed, input values are assigned a time based on the arrival time and recorded individually. So data from two inputs will interlace. For fixed input “FINA” feed, the data drops into a fixed time slot based again on the arrival time. The latest value to arrive within the time slot is retained, anything already in that slot is overwritten.

2 Likes

Ah, that is useful to understand.

This is why the FINA feed is set to (say) 10 s interval, but the emonTx (DS sketch) is set to slightly less, so that it runs faster than the database. The values arriving creep backwards in their slots until a value arrives before the slot expires, so the occasional value gets overwritten.

If it were to be the other way round, the arriving values would creep forwards and there would be an occasional empty slot, which plays havoc with the way the graphs pick values when looking at extended time spans.

1 Like