Sunday, December 15, 2013

12VDC -> 220V AC step-up transformers, a.k.a inverters and laptop chargers don't always work together well.

The other day, I had to go on a long road-trip and wanted to make sure I would keep access to my laptop throughout the journey.

Without giving it much thought, I went to the car parts store and bought me one of these:





The idea was to convert the 12V DC output of my battery to a good old 230V AC that I could plug my 220V AC to 20V DC laptop adapter in.

First, it's kind of sad to jump through all these hoops just to do a 12VDC to 20VDC step-up, but I was in a rush.

Second, a bit overkill on the power rating side, but you never know what you'll end up needing to hook to your car battery when on a camping trip.

Well ... turns out buying this things was a bad idea: contrary to expectations, the laptop refused to charge.

Back home a couple of days later, I decided to investigate ...

First step, let's tear the box open to see what goes on in there:



At first glance,  the inside looks very much like most inverter circuit schematics you cand find on the net:
  • A DC input stage on the right and top with a bunch of what's likely foolproofing diodes and noise filtering caps.
  • A good old LM324N quad op-amp, maybe used to produce the AC oscillation
  • Two ICs that look like voltage regulators or voltage references, codename KA7500B
  • A bunch of power transistors hooked up to the case which doubles as a heat-sink. Transistors look like they are organized in push-pull configuration to current-amplify the oscillation.
  • A bunch of transformers to do the 12V -> 230V step-up.
  • A big fat 400V-rated electrolytic cap, not sure what for, maybe some more filtering.
  • The 230V AC output stage.
In other words, nothing fancy.

Yet things don't work.

Time too hook the whole thing up to the scope to see what's going on:


Now let's feed it 12VDC on the DC input side (current limited to 1A, you never know):



First thing I notice: the darn thing sucks up 8 Watts of power without any load on the AC side :-(
Don't let it hooked to your car battery with the engine off.

Anyhoo, continuing, Fluke says 12V are going in, everything looks good.


And let's take a gander a what comes out the other side:



Ahah.

And ugh.

So, sure enough, at first glance, we get "mains-like" juice:
  • 50Hz says the scope
  • 296V plus peak says the scope
  • -300V minus peak says the scope
  • 244V RMS says the scope (close enough to 230V)
But but but ... the likeness stops there: the waveform is not exactly what I'd call a nice sine wave ... it's in fact an awful bad PWM-like approximation of a sine wave. Yuck.

This is very likely where the problem comes from: I bet my laptop power adapter sorta expects a real sinusodial waveform coming in.

This half-assed approximation of a sine wave is probably making it choke and shut down.

Solution ?

I think there are a few possible avenues:
  • Hacking the beast above by adding a large filtering cap after the output. Not confident enough in my electronics skills at this stage, so not going to try it.
  • Buying a "pure sine wave" adapter. I guess that's what I should have done, had I thought of doing my research :(
  • Buying a DC-to-DC laptop adapter which will at any rate likely be more efficient and more silent (hopefully no large fan and heatsink). Probably something like that.
At any rate, caveat emptor:

If you plan to charge your laptop in your car with a power inverter, you had better triple-check that the so-called "AC output" of the inverter you plan to buy is actually compatible with the input expected by your laptop adapter.

Best way to do this ? Confirm with salesman, buy it, try it, bring back to salesman if it doesn't charge your laptop, iterate.

But don't do what I did and expect it to work simply because the inverter says "230V AC output" and the laptop adapter says "230V AC input".