You are quite right I got that completely wrong (just tried editing post to cross out my suggestion so as not to mislead, but there is no strikethough so I have had to delete it) there is a way to do it along those lines as I have done it before, but again you are right the “ship next” will do it and do it in a simpler and neater way. I tend to forget about the extra processing functions as they not available on emoncms.org and I get used to doing things the hard way.
@pb66:
The left-hand edit window has the ability to accept html, so <strike>works</strike>
to give you works.
It’s very useful for Ω and μ too, especially if you want μΩ 
1 Like
Hi Robert,
I picked up your suggestion about my 9v wall-wart not being properly included in the setup.
Doing nothing else other than rebooting the Pi has solved my problem. I now see 0W or 10W solar generation at night-time where previously I saw 300W. My emoncms readings are now very similar to my Utility meter readings for both Grid imports and Solar Generation.
So I think my original problem was caused by the Pi booting before the 9v was present.
But in my opinion this is an unfortunate limitation of the emonPi design which is not sufficiently prominently flagged in the installation instructions.
Even if I had been made adequately aware of the critical importance of connecting the 9v before booting the Pi, I can still see a scenario where things might go bad.
If I power my Pi from an online UPS or a battery it will ride out any power failure in the Utility supply. But it will not be appropriate to power the 9v wall wart from the UPS because the circuits that the CTs are measuring are assumed to be phase-related to the AC which is powering the 9v source.
So in the event of a Utility power cut, upon resumption, won’t the 9v voltage reference be lost?
Is it impossible for emonPi to look for 9v reference after booting?
Many thanks for your help getting to the bottom of this.
Cheers, David.
Hi Robert,
Can you give me a link where I can learn how to do these manipulations in emoncms.
Thanks, David.
No. When the mains goes away, the 9 V a.c. will be lost and that will give you zero power - which should be the truth! When the mains comes back, the a.c. reference will come back and you have a voltage to multiply by and will be able to calculate power again. It’s only if the a.c voltage isn’t present when the ATMega328P inside the Pi starts up that it assumes you don’t have an a.c. adapter and substitutes a fixed 230 V instead.
There’s a 100 ms delay after the sketch starts before it begins to read the a.c. voltage, and that’s after the 5 V has become established and the 328P has booted, so normally, it’s enough to power both at the same time. The caveat is the a.c mustn’t be later than the 5 V d.c.
I wouldn’t have said impossible, but it isn’t done. You’d better ask G&T (Glyn Hudson & Trystan Lee - the designers). There are a few interesting questions there about what you want it to do, that need careful thought. If you do pick up and use the a.c. reference once measurements are established, I’d suggest it’s a one-way process, you never revert to the fixed 230 V assumed value. But I suspect it’s a question that’s open for debate.
I’m afraid that, as far as I know, there’s no complete and comprehensive documentation for the available operations. If you’ve got the Help in your Pi, that should be a start, otherwise the help is present at emoncms.org. But I usually tell people to try it out and see what happens. The general idea is that the result of an operation is passed down the chain to the next step, but there are exceptions to that rule. You can have multiple ‘log to feed’ in the list, and the result at that point is saved in a feed.