Hi Finn

Really appreciate your reflections and design ideas.

I had graded the risk of my smallish ~45amp alternator going HV as low, and had ‘assumed' that in such an event the 40Ah battery bank and ~20amp system load would ballast the system voltage for some time (minutes) before a critical engine failure was imminent.

I think if you had an ANL fuse or fusible link in the B+ lead and good batteries to absorb the rising voltage from a real runaway alternator situation, the fuse would probably blow and protect the batteries and system from damaging voltage. But I guess that would require some data on what the batteries can handle and what they will do compared to the characteristics of the ANL fuse (how much current for how long before it blows).

My engine monitor will let me know ‘audibly’  as soon as the system voltage exceeds 15 volts.

‘Assumptions’ right! I hadn’t got real confirmation of the real voltage/current over time in such an event.

The truth is I have always struggled to get a handle on how a three wire automotive alternator works electrically.

The 'Sense', I thought was simply the feedback for the regulator which I have from memory via a 5 amp fuse to the engine buss.

Yes, in the RX-7 it goes to the battery via the ignition key.

With no OVP crowbar, I think you'll  be fine.

I'm fairly sure that my OVP crowbar was what killed my alternator output. Very, very stupid to connect the OVP crowbar over the field winding. It should only have killed the power to the OVP relay. I obviously did not consider the failed open battery possibility and lazily connected it also  to the field winding terminal. If I'd know that the alternator could boot up with power supplied to the Sense terminal I would have done so. Especially as its current draw is only 0.27A compared to 3.25A on the field terminal with alternator stopped.

Note that the diode shown in the sense terminal may not exist. Some of the schematics in the Haynes Mazda RX-7 book does not show it.

But what is this ‘lamp/field’ wire??

In my alternator schematic, you see those three diodes I blanked out in one of my posts? Once the alternator is producing power, they will supply the alternator field winding and voltage regulator with the needed voltage to operate. The field winding is the lower coil in the alternator schematic and rotates in the alternator. The three coils in Y-configuration are stationary and are what produces power on the B+ terminal (via a big bridge rectifier).

However, on startup there is no magnetic field being produced by the field winding and the alternator won't produce power. You have to supply that either directly to the field winding (output from the three diodes), or to what I labeled "Startup/sense". In my alternator's voltage regulator there is a path from the startup/sense terminal via the top resistor to the top field winding. Applying power to that will gradually build up enough magnetic field for the alternator to start producing power. That supplied power could then be removed. But that terminal is also (or normally) being used by the voltage regulator to more accurately adjust the alternator's output voltage. Another possibility would have been to supply startup power to the field terminal via a diode after the OVP crowbar.

In the RX-7 there is a "Alternator Warning" lamp relay. One side of its coil is connected to ignition (12V battery) and the other side to the field winding. When alternator is not producing power the field winding is near zero volts and the relay activates and the warning lamp comes on (like when you turn on the ignition key but engine not running). Once the alternator is running and producing power, the three internal diodes produces near +14V (or whatever the voltage regulator is set to) and the relay releases. I suppose that the relay coil initially helps supply voltage to the field winding.

Other alternators may be different and some may even have enough residual magnetic field in them to be bootstapping themselves and not need external power to start.

In mine it’s not connected, and again I ‘assumed’ that as my idle is ~1900 rpm this was spinning enough to excite the system and self sustain the magnetic field at that point.

Seems correct.

Perhaps I ought diode feed this from the Engine Buss?

Don't think so, if your alternator comes on-line with no problem.

I understood that because of the above characteristic once these alternators were generating power you couldn’t turn them off by de energising a ‘field’ etc. wire and therefore your plan to disconnect its output/ B terminal is required.

That is correct. The internal diodes provide voltage to field winding and voltage regulator. I'm still uncertain on my theory that my OVP crowbar was able to short the field winding and thus collapse the field. But it's the only thing that makes sense. Still haven't hooked it up on the bench to test it.

I’m thinking that in your circuit, when your Over Voltage relay is de-energised by the OVP popping the CB, like disconnecting an alternator from its battery, (at high RPM, not simply idle) might damage the alternator diodes/regulator though this is an acceptable outcome if it keeps the engine running in a critical situation.

Good question and point. Definitely something to test with a running engine and loads. Disconnect the B+ lead and see what happens on the B+ terminal with a scope. I think I'd  have a big cap on the B+ terminal when first testing it to get an idea of how big the "load dump" surge will  be (hopefully absorbed by the cap).

What I would like to know in order to make an realistic risk assessment is:

1. How these alternators actually work electrically

Hope I've clarified that above, but no idea on how your alternator may differ.

2. What a High Voltage on the B lead will do over time.

A. In the immediate mSec of an event and

B Over the next 5 minutes

My belief -- still subject to testing -- is that a big cap on the B+ terminal should protect it (and system in case of open battery failure).

Of course in the event of a real  alternator runaway and an OVP relay disconnect be prepared for an exploded cap. I intend to put mine next to the alternator and definitely not in the cockpit.

Another unknown is how robust the rectifier diodes in the alternator are. Will they simply fail open from the over voltage current spike absorbed by the batteries? Maybe there's no need for an ANL fuse in the B+ lead.

In my RV-3B, Rusty simply put an ANL fuse in the B+ lead and there was no OVP circuit when I got it. I need to look up what alternator it is, but definitely smaller than my RX-7 alternator.

How is your RV4?

I hope you can get back in the air soon.

Hope this year, but considering how slowly I work, maybe not.

I'll be putting my attention  on the RV-3B next so I have something to fly while I work on the RV-4.

Thanks again for the reflections.

Steve Izett