24Vac to an arduino usable voltage.

I've seen older/similar posts so I thought I'd refresh it.
and add the things I know from working on HVAC systems.

I am working on a project that powers off an HVAC system.
the transformer in the HVAC is rated to turn 230Vc into 24Vac
unfortunately the power companies pull a little trick to save themselves money.
the input voltage is 248Vac which puts the transformer output at 28Vac.
24V rectified will easily fall below the 36V max on normal regulators.
28V rectified ends up rectifing to about 40V. even at 1/2 wave I end up over 36V.
result: my LM7812 overheats even at idle with a heat sink.
(my project has 12V components I have to use.)

my plan is to give this a try:


I'll post after the parts come in.

Pololu has plenty of DC-DC converters, with 42V input and 5V output.

No waiting weeks for delivery either.

Before Purchase check the current requirement in secondary side . i,e load .

Can yo share your previous circuit diagram to understand why it is getting heated.?? Why the input voltage is fluctuating that much???

the input voltage is 248Vac which puts the transformer output at 28Vac.

Where in the world are you?
This is illegal in the UK and the power compines can be fined if they do this.

Under no load you will always see the peak voltage after rectification, this is 1.414 times the RMS value.

Read Product safety for businesses: A to Z of industry guidance - GOV.UK
With UK old voltage 240 volts and + 6 % you have an upper limit of 254 volts

With the new european voltage 230 and + 6 % the upper limit are 243

So where in the world are you?


The peak voltage of the mains depends on the rms voltage but also on the amount of
harmonics on the supply, which these days can be large due to lots of SMPS's, or in
an industrial setting from motor-driving circuitry.

You cannot rely on 24Vac transformer secondary output peaking at 34V, you should
expect harmonics and noise on the mains, so 40V wouldn't surprise me at all. 100V spikes
quite likely too (fluorescent lighting and fridge motors generate spikes for instance),
but those wouldn't be energetic enough to affect a smoothing capacitor much, its
just the rectifier diodes that need to survive those.

Also an unloaded transformer secondary typically generates 10% higher voltage than
at full load, so unless the transformer is matched to the load the voltage can be
10% out anyway.