AC current in the DC

Joined
Mar 6, 2020
Messages
3
Location
Pennsylvania
One of the things that most would over look in a charging system is a failure of an alternator diode. You might have an electrical symptom where components are working strange. When a diode fails it will either short or open. Most of the time it shorts or leaks AC. When this happens you get AC intrusion into the charging system. This can damage electronic components or make them operate wrong. This AC intrusion is in a frequency of 1000- 3000 htz. Your alternator is just that. An AC generator before the diode clusters. To properly check you need a good brand of digital vom. Or if you have an analog meter like a Simpson 260 you need to add a diode to the ground probe as a DC blocker. Any AC in the DC will read by about half of it's actual voltage. If you find AC in your charging system you need to replace the alternator as soon as possible. A weak charge is also an indication of a bad diode or just a bad regulator.
 
Originally Posted by atikovi
You lost me at, leaks AC. How does electricity leak?


Current can absolutely leak if given a path, no different than a pipe can leak if there's a tiny issue in a solder joint or pipe wall.

Leak isn't the right term for the ac component of an anticipated dc waveform.

One has to start by realizing that three phases rectified to dc is intrinsically horrible dc power. It's not a flat, continuous voltage, it's horribly bumpy. Even fancy multi-pulse power electronic rectifiers can only start to emulate true dc. It needs to be "filtered".

The fewer diodes, the lumper the rectified dc looks. Lose one diode in an alternator, that sc comes straight through, positive and negative parts of the waveform, and makes it even more bumpy and lopsided.
 
I think Bon Scott was certainly the superior singer compared to Brian Johnson.

Oh wait a minute, this is about electricity? Disregard.
 
Originally Posted by Chris Meutsch
I think Bon Scott was certainly the superior singer compared to Brian Johnson.




lol.gif
And i agree! Although i like Brian Johnson too, but of the two, Bon.
 
OP is correct. Google alternator ripple current for more info. A multimeter set to millivolts AC will aid in diagnosing an alternator problem.
 
I think I would almost consider it contamination when an AC ripple occurs in DC, and it can play havoc with anything electronic, even the battery (which basically acts as a big capacitor and smooths out the "bumpiness" caused by the full wave rectifier in the alternator when the engine is running).
 
Originally Posted by atikovi
When a pipe leaks you see water leaking on to the ground. How does that relate to electricity?


Think of the wires and those little traces on printed circuit boards as a network of water or hydraulic pipes, only instead of transferring water or hydraulic fluid they're moving electricity. The various components like resistors, diodes, and whatnot are like pressure regulators, check valves, et cetera.

When a pipe bursts or a valve fails, you get fluid where you don't want it. Same thing happens in electronics, only instead of it physically falling onto the ground, it's going to some other component that it shouldn't be.
 
Last edited:
Originally Posted by atikovi
When a pipe leaks you see water leaking on to the ground. How does that relate to electricity?


Stray current paths dissipate as heat. And can damage components.

Pressure and flow are analogous to voltage and current in many ways.

The only difference is that water in a pipe WILL flow out onto the ground. Electricity will have the potential to do so, but wont go unless it has a path.

To the OP's point, AC ripple on the DC bus will cause over voltage which degrades capacitors and other elements in electronics. Ripple that goes positive and negative inside a battery will cause heating. Heat causes degradation in chemical batteries.
 
Basic Diode Operation and rectification:

https://wiki.analog.com/university/courses/electronics/text/chapter-6

A healthy diode in an AC circuit only conducts current when the polarity of the AC waveform is of the proper polarity. A "leaky" diode on the other hand will allow AC current to pass no matter the polarity of the AC waveform.

Here is a PDF file that explains automotive alternators:

https://www.microcharge.de/downloads/Understanding_Generator_Ripple_Waveforms.pdf

To properly view the alternator output waveforms you need an oscilloscope.
 
Last edited:
AC current is forever variable, the diode redirects it positive. Direct current.

None of this really matters, you don't repair diodes and rebuild alternators anymore...you hook up a little hand held charging system tester and it tells you if the rectifying bridge passes, charging system output, battery voltage, damaged cells ...the whole test takes three minutes.
 
Last edited:
When I first started working on electronic devices for automotive applications, somebody gave me a good paper that started off saying something like, "When you connect to a car's battery, you are plugging into the power supply from [heck]."
Load dump, reverse battery, double battery...all kinds of crazy stuff to worry about, and even normal operation is pretty ugly in itself. A messed up alternator makes things even worse, the battery actually makes for a darned good filtering element if it is working (sort of like a big capacitor) but it can only do so much. Load dump is just the removal of the battery from the alternator while it is charging, the current in the coils cannot change instantaneously to respond to the loss of load and a big voltage spike results (V=L*dI/dt). Having a loose battery connection can make you get load dumps over and over.
Luckily, most of the automotive stuff I worked on worked off a regulator from the battery, but I did have some parts that had to deal with this hellish stuff. Add in the fact that every customer had their own definition of load dump with differing peak voltages and attack/decay rates, or maybe multiple definitions...ugh.
 
Last edited:
Diodes can absolutely leak current, it's specified on the data sheets among many other parameters that define that a diode is not ideal. We won't even get into reverse recovery time and junction capacitance. Leakage current and reverse recovery time usually get worse with heat (positive temperature coefficient), to the extent that you can get into a thermal runaway scenario which may lead to the destruction of the diode, other components in the circuit, and possibly the equipment connected to it.
 
Originally Posted by atikovi
When a pipe leaks you see water leaking on to the ground. How does that relate to electricity?

with electricity a meter is used.
 
Originally Posted by JHZR2
One has to start by realizing that three phases rectified to dc is intrinsically horrible dc power. It's not a flat, continuous voltage, it's horribly bumpy. Even fancy multi-pulse power electronic rectifiers can only start to emulate true dc. It needs to be "filtered".
Originally Posted by Virtus_Probi
the battery actually makes for a darned good filtering element if it is working (sort of like a big capacitor)
This is exactly right. The battery IS the filter. The engineers that designed automotive alternators knew this. In modern computerized vehicles, a bad or weak battery can cause all kinds of problems, many of them related to this.
 
Originally Posted by atikovi
When a pipe leaks you see water leaking on to the ground. How does that relate to electricity?

We can make an analogy between an ideal diode and a one way valve in a plumbing or hydraulic system that is only supposed to let water flow in one direction...same with a diode, it is supposed to let current flow in response to an applied voltage in its forward direction but not in its reverse direction.
It may take some finite pressure for the flow valve to start allowing fluid to pass...that is analogous to the forward voltage of a diode, it takes a nonzero voltage for the current to start flowing in the "right" way.
That valve may also leak a little when it has back pressure on it, and that is similar the reverse current of a reverse biased diode.
You put too much reverse pressure on a valve and it may break down and start letting fluid pass in the "wrong" way...the diode also has a reverse breakdown voltage past which it will conduct current, and this may or may not be destructive to the diode depending upon its nature and the degree of reverse stress. I don't know that much about valves, but I'd think this would inherently destroy them unless this feature was actually designed in for protection.
A valve can be stuck open or damaged and allow fluid to pass in either direction without much restriction...this is like a diode that shorted out and doesn't block anymore. You can also have a valve that is stuck shut and a diode that is open circuited that blocks everything.
In general, voltage (aka electric potential) can be considered as analogous to water pressure...just because you have a voltage difference in a circuit doesn't mean you have current flow, just like you can have high water pressure in a plumbing or hydraulic system without any fluid actually flowing. Current can be thought of as water flow, and the electric charges like the water itself. The voltage is related to current by impedance, with resistance causing current in phase with the voltage and capacitance and inductance causing out of phase current. Almost any practical load impedance is going to be some combination of resistance, inductance, and capacitance. A hydraulic analogy to a capacitor one of those spring-loaded anti-hammer pipe attachments you can buy, a hydraulic inductor can be a turbine. Electric resistance is like the resistance of a narrow pipe to water flow. Power factor correction is the art of making a load that is not purely resistive look more like a resistor...but that is probably a topic for another post.

One thing I don't know is if any automotive alternators increase efficiency by replacing the diodes with switches that have their gates controlled such that they emulate diodes with near zero forward voltages...this can increase efficiency a fair amount, but requires intelligent control and is probably not generally as robust as the diode bridge solution. It is often tempting in automotive to go with robust and simple...
 
Back
Top