Hi Uffe,
Quote:
[b]You mention switching design, I guess you mean transistors or IGBTs - controlling the output voltage of the alternator?
No, I was referring to how switching transistors or other devices can change DC, or 0 Hz AC into higher frequency AC current, which is affected by inductance, to highlight that it's the changing currents that can be problematic when they're present but not wanted. Inductance is present in all wire, just in different amounts. DC isn't affected by it after steady state current flow is reached through it. It does have an effect on initial power up, as a state analysis shows, but again, the DC value is rising and is not steady state during that brief time.
If you have a friend with a ham radio with some tube finals (its visually more dramatic than with transistors to my eyes) in it, ask him to put a scope on the parallel tuned tank circuit, and a DC voltmeter on the top side of that same tuned tank circuit which supplies DC B+ to the plate of the tube. DC will flow from the B+ supply, right up through that coil of wire with no noticeable dc voltage drop, yet the scope probe connected to the same point will show high frequency RF voltage waveforms. At the bottom of that same coil will be the same DC voltage tied to the DC source, but little to no RF voltage. The RF sees too much impedance in the coil (parallel RC tank circuit) to short to ground or B+, just like it won't on your alternator wiring. RF is nothing but a term for AC of a designated frequency range. DC and RF existing on the same coil of wire, by design. In your vehicle power bus, very undesirable. At the right frequency, that wire from your alternator to your battery (and elsewhere) looks like an inductor coil with high impedance and inductance. Any fast rising current through that wire will behave just as in any other reactive circuit, just "undefined" as they like to say for states not mapped or desired.
Our CRD alternators don't using switching inverters, just the diodes arranged in a full wave rectifier configuration. That diagram you posted of the full wave pulse shows both halves of the alternating sine wave produced by the alternator, gated so that both halves of the sine wave are always pulsed in the same direction, or no reversal. That upper portion of the sine wave that can introduce higher order harmonics, which can be heard on some radios at certain rpms, is of interest. The RMS value of that half sine wave, or about 70.7% if memory serves correctly is what the dc equivalent power rating comes from. But those upper portions are still changing values and under certain situations can be heavily distorted into noisy spikes. Harmonics of those even without distortion can be heard in your radio sometimes, but switching distortion of those diodes when driven hard enough with heavy current can produce narrow width, high voltage spikes, right into the vehicle wiring. Narrow width is another way to say short time duration, which is another term for high frequency. The low impedance of the vehicle wiring for DC is also high impedance for those high frequency spikes. As I'm sure you know inductance is frequency related.
Guess what they can distort into when driving a weakened battery with enough internal resistance, when the alternator sensing 7,8 or 9 vdc right after starting, opens up at full bore, driving those diodes into heavy over saturation conduction, while another vehicle alternator begins doing the same thing? They end up looking like your picture, except with some very, very high, narrow spikes on top of them. Those short duration, high spikes are literally of RF frequency and can ride the outer layer of the wire(RF skin effect) from the alternator to the battery, into the vehicle buss and right into the electronics. Even then, most of the time, they are dissipated safely, but it's the one time they aren't when it costs dearly. If the conditions are right, they can be hundreds of volts high. That is the voltage spikes I'm referring to, not the surge. With a decent oscilloscope, you could see them if set to capture to a sub-microsecond time base.
That is one issue that can damage the ECM - another is ordinary voltage surge, brought on by the time lag of two alternators charging heavily into a weakened battery, driving over voltage very briefly before the regulators catch up. Nothing ever works instantly, even electronic voltage regulators. If you measure with enough time granularity, you can see this. If you drive 270 - 300 amps into a high resistance load (like a weakened battery with a partially open cell) you can go well above 14 vdc easily for very brief instants of time, milliseconds, but enough to occasionally damage sensitive electronics. You will never see it on a DVM or standard voltmeter or an old Simpson 260. Even a really cheap scope wont capture it. It can be difficult to grasp without the tools to view it, but I've worked with RF electronics long enough to know that many RF engineers grasp it because they have access to the tools to be able to see it visually. One look at it on a scope is easier to grasp than a weeks worth of reading.
Quote:
I would have guessed on diode rectifiers with capacitors to filter the higher order harmonics - the bigger the caps the "better" the filter.
I doubt they bother to use filter capacitors in automotive alternators, at least I've never seen any, depending instead on the huge capacitive value of the battery (at 0Hz or DC) to filter down those pulsating outputs.
Quote:
In such passive designs, I see very little chance of creating a voltage spike caused by reactive currents, since you are rectifying your voltage passively, you cannot inject reactive currents unless your load is inductive. This type of reactive current will not produce a voltage spike unless the inductive load is suddenly disconnected (cables, and as tony points out, beware of the time when you pull the cables off).
If you believe a DC powered automotive system is nearly immune to induced voltage spikes, think about the military electromagnetic pulse shock weapons. They can disable everything electronic in a modern automobile. How do you suppose that works? Why doesn't the battery filter it out? Where is the inductive load in the vehicle coming from?
Understandable, but as I stated earlier, its easier to grasp the first time you see it on a scope and it's easier when you work in a field where those tools needed to observe it are readily available. The next statement about disconnects is not entirely correct, all that is needed for inducted current is a change in the field surrounding the inductor or a changing current through it, not only a disconnect. A disconnect is one of several ways of causing a changing field. The heavy current switching distortion I mentioned is affected by the inductance of the wire, with smaller reverse kickback, sometimes called ringing. They exist as a circuit, but it doesn't exhibit undesirable behavior until certain conditions are met, such as rapid switching transitions under very heavy loads. This is also one of the reasons why running a vehicle with a bad battery can be very costly with all of the electronics on board. It wasn't a problem at all on the vehicles I grew up with, with generators, no computers of any kind to ruin. Not a single semiconductor in the entire vehicle. How things have changed. Maybe we should keep an old mechanical diesel around just to be sure
Quote:
The sparks created by connecting two batteries could be caused by voltage potential differences between the cars at connect time. Again I assume the battery has enough capacitance to filter out the voltage spikes.[quote/]
Usually, but again, if the spikes have very high frequency components in them, they can ride down the impedance of the jumper cable and both vehicles battery supply lines. 99% of the time it won't bother anything.
Regards,
Ranger1
Login
Register
FAQ
Search
