Re: Split charging and charge lights

Roger Glover wrote:

[snip]


"Gibbo" <gibbo@xxxxxxxxxxxxxxxx> wrote in message

Have you tried this in *several* installations?

Yes. Automotive, marine and static installations both 12 and 24 volt.


I can see *many* instances that would blow the transistor.


[snip]

I'd appreciate if you would reveal what you see as the '*many* instances
that would blow the transistor', as I've found, in practice, using suitable
components, this method of energising a 'split charge' relay, without
loading the initial field excitation, has proved to be both economic and
reliable.


Firstly I'll confirm that your choice of emitter follower is the
correct choice. And of course no base resistor is required.

What first brought failures in this type of circuit to my attention
were random failures about 10 years ago on various bits of kit that
seemed to work well in many instances but showed repeated failures in
others.

This isn't really the correct place to discuss this but as you asked
I'll continue. The boaters are, of course, welcome to ignore it :-)

You may be familiar with ISO7637-1/2. If not, they are the industry
standards that equipment designed to operate on 12 and 24 DC volt
systems should be designed and tested to.

The transistor across the warning lamp will fail in almost *all* of
these tests.

It needs to withstand a load dump. This can be up to 400 volts positive
on the collector. Obviously the choice of transistor can allow for this
on the collector. It is doubtful that is the case for the base which
can also be subject to the same voltage. Further, if the collector is
not subject to the load dump but the base is, then this will forward
bias the base/collector junction and blow it.

The collector needs to withstand *minus* 600 volts with respect to the
emitter! This will require a forward biased diode in series with the
collector. This diode will need to be an ultra fast recovery diode so
that a fast reverse transient does not continue to conduct. It will
also be necessary to insert a resistor in the collector lead and a
zener from collector to ground. Obviously this resistor has to pass the
full relay coil current so it needs to be of very low value. Which
thereby necessitates a very high power zener and series diode.....

The same precautions will be required for the base.....

These relate to just *two* of the tests in the standard. There are many
more.

And so the problem goes on.

Now I'll admit that I'm the first to ignore the bullshit regulations
from Brussels when they have no bearing on reality but this isn't just
theoretical bo*&^%x due to some new regulation. Our experience shows
that unless equipement *is* designed and tested to meet this standard,
then failures become very common in the field. They sometimes work for
months or even years then fail, for no apparent reason.

I have been involved in failues of this type for many years as
representative of other companies and also got stung by it on one of
our own products. After this experience we test everything to the
highest level (there are four levels) of severity in ISO7637.

Any equipment that does not repeatedly pass the most stringent tests in
this standard *will* fail in the field.

On *most* (not all) mitsubushi and hitatchi alternators simply
switching the ignition key off will blow the transistor with the shown
circuit. This is because many of these alternators use separate stator
windings to feed the D+ diode trio and if the ignition switch is turned
off at a time when the regulation transistor is not conducting (highly
probably once the batteries reach a higher charge state) these separate
coils have no load and will produce an effective load dump pulse
directly into the base.

The idea for the circuit is good. Implementation of it is somewhat more
invloved in order to get a fully reliable circuit.

I think a small relay is probably simpler and more reliable. For
vehicles it also doesn't breach e-marking requirements!

Gibbo

.

Relevant Pages

  • Re: Simple 555 PWM - disappointing performance
    ... For three amps of collector current it ... > a voltage of only one volt? ... This implies that either the transistor is not turning ... base signal goes well over 1 Volt, ...
    (sci.electronics.design)
  • Re: Transistor Characteristics
    ... Is there any way to dig up what the coefficients are for the diode ... diode connected transistor is actually the world's simplest feedback ... and its GBW is roughly proportional to collector current. ... low currents, but the collector current pretty much follows the ideal ...
    (sci.electronics.design)
  • Re: real working of transisitors
    ... >> diode even though the collector base junction is forward biased. ... > If you hook the transistor in upside-down mode, ...
    (sci.electronics)
  • Re: Transistor base as a diode
    ... a diode where the base is also short circuited to the collector. ... The most common case that comes to mind is a current mirror circuit, ... is as the OP mentioning a transistor with its CB ...
    (sci.electronics.components)
  • Re: Transistor base as a diode
    ... a diode where the base is also short circuited to the collector. ... is as the OP mentioning a transistor with its CB shorted and the BE used as a diode - as opposed to in a configuration such as a CM. ... One application that springs to mind is that it yields a slightly lower Vf than a regular diode, the collector will begin to draw current at a VBE of around 0.5V. ... Saves inventory costs, rigging costs etc. ...
    (sci.electronics.components)