I am creating a new thread, since the original thread's topic is not directly related to the problem I'm trying to solve. I've also taken the liberty of quoting said thread's participants who were kind enough to answer my questions and offered advice, for the purpose of letting them know that the topic being discussed has been moved here.

Now I get it, the response time of the alternator regulation system is slow, when the battery impedance is removed, the voltage soars!

I worked in the automotive electronics industry for 15 years. I can tell you that those products are fully characterized and protected for load dumps and many other abnormal events.

You could also likely use three ≥100A standard (not Schottky) silicon diodes in series to perform the protection function.

You should limit it to at most 0.3C if you care about battery life. You can push it harder if the battery is near being flat but the LA electrochemistry starts to be pretty lossy near 80% Soc so it starts to get hot (warping plates and degrading separators). Regulate the charging voltage to limit current and them limit that voltage to stop gassing.

Would two (or three) of these diodes do the trick?

Recap: I want to install a battery disconnect switch in my car, but I want to protect its operation by making sure that if some idiot (such as me) has the bright idea of disconnecting the battery while the motor is running and the alternator is delivering a considerable amount of current. The protection is needed because under said scenario, the instant the battery gets disconnected, then the alternator would experience a sudden increase in its output voltage due to its internal regulator being too slow to react. If that were to happen, the car's entire electronics (including the alternator's internal diodes themselves) would be fried ... and that would be a rather undesirable outcome ... This can happen because the battery in the car also acts as an overvoltage dampener of sorts. And if said dampener is suddenly removed, it could have very adverse effects under the aforementioned circumstances.

Anyway, the simplest idea so far was proposed by crutschow: Use two or three normal silicon diodes in series with a high current capability so as to take advantage of their inherent voltage drop. Said voltage drop would prevent current produced by the alternator from reaching the battery the switch has been opened, but the diodes would start conducting if the alternator's voltage exceeds that of the battery by the amount of voltage being dropped by the diodes.

This is what I gathered from the way he described his idea:

​

The diodes I am planning to use are Infineon Technologies' IDWD150E65E7XKSA1, which have the voltage drop curve shown below:

​

Said diode is rated at 200A, so it should be enough to protect the car's electrical system for the half a second or so that the alternator's voltage surge is expected to last.

Any thoughts or comments?

 

Why not use just one diode?

 

I really don’t understand how someone could have the bonnet/hood of the car open (to disconnect the battery) and not know that the engine is running (or not).

To disconnect the battery, always first disconnect the connection at the negative/chassis terminal – that way if the spanner/tool accidently makes contact with the vehicle chassis you will not short the battery.

 

Why not use just one diode?

The voltage drop needs to be significant (in the order of about 4 to 6 Volts) to make sure that current stops flowing from the alternator into the battery once the regulator stabilizes after opening the switch.

 

The voltage drop needs to be significant (in the order of about 4 to 6 Volts) to make sure that current stops flowing from the alternator into the battery once the regulator stabilizes after opening the switch.

Isn't the point to prevent damage?

Why not just let the current flow until you shut off the engine? Then it stops for sure.

 

The point is to disconnect the battery so that it doesn't spontaneously discharge while the car is being kept in storage for prolonged periods of time.

And yes, an ordinary switch is normally used for that, but I want to go the extra mile and make extra sure that under no circumstances the car's electronics are at risk of damage due to a better idiot bypassing the idiot-proof arrangement.

 

@crutschow , I have a question. Would this arrangement work the same as the one shown on my first post? ... My instinct tells me that it does, but I don't trust my instinct alone and I'm trying to visualize if there's something wrong with the circuit:

​

 

Been there and done that back in 1967. Got a ride in a 1966 Rambler American with a failed battery that had to be push started. There was enough residual magnetism (I guess) in the alternator to deliver enough power to fire the plugs enough to start running if I got it moving very well before the driver popped the clutch. That usually only took running a very few steps. And the alternator with no battery load never fried the radio or the instrument cluster.

 

And the alternator with no battery load never fried the radio or the instrument cluster.

The problem is not the alternator with no battery load. The problem is the sudden disconnection of the battery while the alternator is delivering a high amount of current. Such scenario could prove catastrophic for the car's electronics due to the regulator's slow response.

 

@crutschow , any comments on my post #7?

Will placing the diodes and switch on the negative pole work as well?

 

Addressing post#7: the voltage surge caused by suddenly disconnecting the battery when the alternator is delivering maximum output is caused by the removal of the load while the rotating field is at maximum. Which side of the load makes no difference at all. The voltage rise is the result of the residual magnetism that remains for a while after the field current ceases.
Auto manufacturers have understood this process from the start and so have chosen to assure that all the devices are designed to survive the "Load Dump" transient voltage spike. The magnitude and time duration are usually defined in the vehicle power description documents.
The obvious protection scheme would be a Zener-diode type shunt able to conduct the alternator maximum output current for the duration of the voltage peak, connected directly across the alternator output. GOOD LUCK on creating that package in a cost effective device!!