In article ,
Paul wrote:
You might ask yourself, if the car battery impedance is as low
as we think it is, how do *70V spikes* appear on the car
electrical bus ? As a Natural Philosopher, you can go find
an answer for us. I haven't seen any oscilloscope pictures
of car electrical systems which prove these spikes are present.

Nor have I. Despite playing with car electrics more than most. Except if
you were stupid enough to disconnect the battery with the engine running.

As the saying goes, you can make things fool proof. Idiot proof, even. But
not c**t proof.

As an illustration of this, the specsmanship, we can use a
TDA2003, an audio amplifier intended for cheap car radios.

UTC TDA2003 LINEAR INTEGRATED CIRCUIT

ABSOLUTE MAXIMUM RATINGS(Ta=25°C)

PARAMETER SYMBOL VALUE UNIT
Peak Supply Voltage Vs 40 V
DC Supply Voltage Vs 28 V
Operating Supply Voltage Vs 18 V

Nominal

Supply Voltage Vs 8 to 18 V

Now, that's an integrated circuit, where the nominal battery
voltage might be 12.6V, or maybe 14.4V fully charged or under
charge or whatever.

The 18V value is what results, if you jam the field winding
to battery, as an alternator test. The alternator can run
the battery up to around 18V, by forcing a lot of current
into the battery. Your mechanic at the shop might try this.
That's why this stereo expects on occasion, to see 18V, and
the car radio will continue to play.

I don't know of any modern alternator where it's remotely possible to
bypass the regulator. Certainly not by the average grease monkey. Who
knows very little about car electronics, other than fitting new parts.

Well, what do the other voltages mean ?

The 40V value is a measure of load dump. If there is a
transient on the car electrical bus, the 40V value can be
achieved. The TDA2003 is designed to disconnect itself,
at the 40V level, in an attempt to protect the IC innards.
The behavior of some part of the IC changes, when the supply
goes over 18V.

BUILT-IN PROTECTION SYSTEMS

LOAD DUMP VOLTAGE SURGE

The UTC TDA2003 has a circuit which enables it to withstand a
voltage pulse train, on pin 5, of the type shown in Fig. 23.

If the supply voltage peaks to more than 40V, then an LC filter
must be inserted between the supply and pin 5, in
order to assure that the pulses at pin 5 will be held
within the limits shown in Fig.22.

A suggested LC network is shown in Fig.23. With this network,
a train of pulses with amplitude up to 120V and width of
2ms can be applied at point A. This type of protection is ON
when the supply voltage (pulsed or DC) exceeds 18V. For
this reason the maximum "operating" supply voltage is 18V.

There can be spikes on battery, but you would hope that things
like CANBUS or whatever, the spikes would not be seen there.

What is the battery doing when those spikes are present ?
You got me there. Looks like the battery impedance at
frequency, is insufficient to deal with the spike in any
useful way. It suggests an impedance versus frequency curve,
of weird shape or behavior. Even bench DC power supplies,
have impedance versus frequency curves.

A cheap battery charger, is a half wave or full wave rectified device
running off a transformer. A pulsating DC waveform is being applied
to battery, but staying at relatively low voltage levels. A good reason
for not making sparks around the battery area, is igniting any
evolved hydrogen (more of a problem around the battery types
where the battery caps are removable, as those have vents). It's
for that reason, you could connect the charger while it's off.
The electrical spike issue, the car design has taken some
of that into account, with load dump specs and the like.

Love to know where you'd by a cheap battery charger that has a transformer
these days. SMPS have been cheaper for many a year.

--
*Gun Control: Use both hands.

Dave Plowman London SW
To e-mail, change noise into sound.