Archived Forum

If your 8V doorbell transformer plugs into a standard ring-main socket, then that should be fine.  But if it has to be wired directly into the mains, that would not meet the safety conscious standards that the OEM site needs to have. 

The post-transformer amplitude of the AC supply doesn't really matter as most of the signal is removed in the divider stage.  Whatever amplitude is left just needs to occupy most the the ADC's available range.  Accuracy can then be fine-tuned by using  calibration factor(s).

What is the regulation of your doorbell transformer, i.e. what happens to the output voltage when the bell-push is pressed? Just asking....

When it is pushed? I have no idea how if/how it collapses I should go and measure that actually, good point. While you don't have to share the transformer, I did mention it as something that is very easily obtainable. So an 8V design has its merits.

There are several kind of doorbell transformers. Some are neatly mounted inside the panel (my personal preference) some have a regular plug. Mine for example has a regular 2 bladed euro plug connecting to a socket especially created in the meter cabinet for the doorbell transformer (and an internet modem).

Mind you, I have seen plenty of homes/meter cabinets, where the doorbell transformer is part of the breaker panel and no extra outlet was available at all, so a 9V AC adapter wouldn't have worked anyway.

So personally, I say standardizing on 9V seems random, when 8V is readily and easily available. I'm sure there's 8V power bricks available as well. You have to admit, it does make sense ...

While you're in measuring mode, check the phase shift as well! It's quite hard to find a transformer that doesn't introduce harmonic distortion and phase shift, and both are important when measuring power.

In the UK, bell transformers are often suppied for surface mounting with screw terminals for a direct connection to a fused spur - or often into a lighting circuit, like this, with several voltage options.

Before I do anything dangerously stupid, it should be safe to connect my oscope to the 8V output side of the transformer, right? I'm pretty sure it's safe, but just double checking before I create an gnd / earth short :)

Having said that, here in NL older houses metercabinets look like this:

Here you clearly see the transformer to the right and an outlet where it plugs into next to it. Ignore the stuff on the right, no clue what that is for, looks like telephone equipment.

Newer houses may use nice built in ones:

Those are nicely tucked away in the beter cabinet like such (click for big version):

To me personally, the last solution looks the best. It's nicely tucked away, with only a wire (or two more on that below) coming out connecting to a nicely packaged EmonTX for the metercabinet.

Now while it could need to be investigated if the doorbeltransformer can be combined with the doorbel, the drop/distortion when pushing the button can be really bad, but then again, how often does that bell get pushed?

That said, using an 8V AC-AC source sounds far more 'standard' because of this then 9V AC-AC. So why not standardize on 8V, that really was my question ;) Prices would suggest that those are decent transformers, they feature a power switch and have overheating protection, so seem far better then 'cheap chinese brand wallwarts'. But obviously, for a 'fully safe consumer friendly' setup, supplying a 8V AC-AC wallwart can still be done reliably?

Again, 8V AC-AC is just so easy to get, while anything  AC-AC in the walwart factor is much much harder.

Any level of AC voltage should work equally well: 3V, 6V, 8V, 9V, 12V ... whatever.  

In all cases, the peak-to-peak amplitude will need to be reduced to whatever is right for the ADC.  With an Arduino Uno, or similar, that's 5V by default.  With an emonTx, it will be 3.3V. 

I would go with whatever safe source of AC you can devise, then run any of my simple test sketches to find out how much of the ADC's range is being used.  If the results are not to your liking, just change the ratio of the divider resistors.

I wasn't after my personal use case, but rather 'wouldn't this be a more generic voltage for all users'.