Issue with changing burden resistor specification in 02_03_18

Release 02_03_18 contains some changes to support IoTaWatt manufactured with different burden resistor values. The latest production run now uses 20 ohm instead of 24 as in older units. Without debating the relative merits of the two, suffice to say that on balance there should be no practical impact for the vast majority of users.

This problem is with the user capability to specify new burden values. To put it simply, that capability is not working. Sounds awful, but the reality is that hardly anybody changes the burden values. Where this has cropped up is in a workaround trick that I suggested to allow using some selected voltage type CTs without modifying any hardware. The short story is that it’s done by changing the burden specification which isn’t working.

Mea culpa. I had made the proper changes and it was working great, but I got sloppy with my merging of configuration app changes and the changes were lost. I’ve reinstated the changes in the latest github version and they will go out with the next release. In the meantime, if there is anyone that neads to change their burden specification, you can either:

  • Download and install the updated configuration app index.htm (94.3 KB)
    from github.
  • Change the burden specification in the spiffs file using the file manager /esp_spiffs/config/device/burden.txt
  • Post a support request and I’ll talk you through one of the above.

How can I know which version do I have ?
Is the version shown on the log ?
10/29/18 22:56:01 device name: pinguIW3, version: 3

Just click Burden Resistors in Setup. The values will be shown. But you don’t need to know. This issue is only a problem if you are changing the values.

Thanks a lot for this post! It seems to be just what I needed, because I just started working with voltage type CT’s and prefer not to solder if it isn’t necessary. Some small question though.

When looking at Setup => Burden Resistors, the system said 24 Ohm, but when looking at “/esp_spiffs/config/device/burden.txt” I was told they are 20 Ohm. Which one is right?

This method worked well for me. To start I assumed the burdens in my IotaWatt are 20 Ohm and I set the equivalent values (for a sensor resistance and a burden resistance in parallel) accordingly following this method. My readings have become a lot more accurate, seemingly within a percent or so of the full scale of the sensors.

I am really starting to become a fan of the IotaWatt! :slight_smile:

Without opening it up I can’t really say, but I have not yet shipped any 20 ohm versions, and all of the units shipped by OEM since they restocked are 20 ohm, if that helps. The spiffs values should be correct as they are set during manufacture.
EDIT: I just checked the code and that is consistent with the bug that was fixed. If you download the fixed config app above and install it, the app will match (and update) the spiffs.

This method is really a bad hack, and I’m sorry I didn’t approach it a different way, and perhaps I will in the future. The spiffs burden file is intended to be a true representation of the actual burden resistors in the IoTaWatt. The ultimate intent is that if you move an SDcard with configuration from a 24 ohm unit to a 20 ohm unit (or visa-versa), everything will still work. When the spiffs doesn’t have the “real” values, all bets are off. I always say one bad hack leads to another.

In retrospect I think a better way to handle this is to allow configuring a voltage type CT even when there is a burden, and compute the melded burden at startup. That way, there is no math to configure the burden, and you could move the configuration between different IoTaWatt with different internal burden values.

I have to say that those voltage type CTs are less than optimal for this application. The 020’s on the mains are probably fine, but any of them, and especially the 030 and higher, are a poor choice. If you decide to upgrade, the iotawatt stuff site sells echun 50A CTs that would be more appropriate. YHDC SCT013-000 would be better than any of the voltage types.

You might want to check that the orientation of the CTs to be sure they are all the same. It looked to me like the rightmost of your mains CTs might be reversed.