Blocking Diodes

[Cantata]

Last year I fitted a Sterling regulator, charging through a blocking diode to 2 battery banks. Works as advertised.
When replacing the starter motor yesterday after overhaul, I found out to my surprise that the line from the alternator to the diode is live. (I knew this from the size of the spark when I inadvertantly earthed it!).
I understood that current can't pass back thro a diode. Or does it only stop current flowing from one battery bank to the other?

<hr width=100% size=1>

 

[halcyon]

It is possible that you have a diode gone, they can fail open circuit , you get no charge to that battery, or they can fail short circuit, this means that one battery can feed the other. Check the blocking diodes with a diode test on multimetre, also check the wireing that the blocking diode is correctly installed.

Brian

<hr width=100% size=1>

 

[tome]

I've seen it posted several times that a diode can fail short-circuit, but I have never experienced this and think it may be a myth. A diode is a semiconductor junction and the laws of physics plus 30 odd years experience lead me to believe that this simply isn't possible. Open circuit, certainly.

The output from the alternator should go direct to the central post of the BD. If, when disconnected, you get a spark then it is live. Most likely explanation is that it's a faulty installation and that the original line from alternator to battery is still intact. It shouldn't be.

<hr width=100% size=1>

 

[Cantata]

I'm certain it is wired correctly, and the original line is not there. The installation has worked fine all last year.
Done nothing since except take home the alternator (which I do every winter to keep dry and warm). Hence the wire was dangling!
Halcyon, I have a position on my multi-meter with a picture of a diode but I've lost the instruction book. Any idea how to test and what the readings will mean, please?

<hr width=100% size=1>

 

[halcyon]

I've spent the last 25 years designing marine charging systems, over that time we have found a good percentage of power diodes failing short circuit.

Brian

<hr width=100% size=1>

 

[halcyon]

Set to diode picture, the flow is the from the red probe to black probe.
So if you put the red probe on the alternator terminal, and the black probe to the battery terminal ( batteries turned off ) you will get a little beep from the meter, and you have a reading on the meter 300 / 800. Reverse probes, and you should get nothing, and the meter will read open circuit.
If you get a continuous beep, and a reading of 0 on the meter you have a short circuit.
Disconnect the alternator, with batteries connected, and see if you have a voltage on the blocking diode, if not, put the wire's back on the alternator, less the connection to blocking diodes, and see if you have a voltage on the alternator output connection.
Should then tie it down a bit closer.

Brian

<hr width=100% size=1>

 

[VicS]

Not so sure about the multimeter beeping. They don't all have an audible signal and those I've come across that did only had it on a continuity test function. The important thing is that when testing a diode you get a high resistance reading one way and a low resistance reading the other way. Incidentally you must use the diode test function not one of the ordinary resistance ranges.

<hr width=100% size=1><font color=purple>Ne te confundant illegitimi.</font color=purple>

 

[Cantata]

Thanks all. Will try out in a couple of days.

<hr width=100% size=1>

 

[tome]

Ok, I stand corrected but have never experienced this as far as I can recall. Any explanation of how this occurs? Time to dust off my old textbooks, perhaps.

<hr width=100% size=1>

 

[halcyon]

Re: Blocking Diodes tome

tome it was mainly on battery charger rectifiers, were we have had diodes and thyristers fail both short or open circuit, though I have also come across a 1N4001 feeding an LED fail short circuit.
As the faults give typical unit life of 12 / 15 years use it's always been put down to normal wear and tear, and not warrented a lot of investigation, but seemed to be allied to a high induction load.

You have got me thinging about something we stopped making 8 years ago, thank you very much.

Brian

<hr width=100% size=1>

 

[VicMallows]

Re: Blocking Diodes tome

Confirm also seen diodes fail short-circuit. Usually the heavier types.

Vic

<hr width=100% size=1>

 

[William_H]

Halcyon on my multimeter the diode test function is a measure of voltage drop across the diode so in the forward direction it gives a reading of .7 aprox while in the reverse direction it shows overload a shorted diode will give a zero reading in both directions. The resistance reading in the forward direction on ohms scale is a function of the internal arrangement of the meter and certainly doesn't indicate the diode has 300 ohms or so it's what the meter interprets as resistance with the .7 ohm volt drop. So contradicting another commenter you can use a multimeter on ohms function either digital or analogue but expect to get a reading of some hundreds of ohms in forward direction and infinity in the other. Incidently the old analogue meters will usually give the opposite effect regarding red lead to the anode compared to digital multimeters. ie red lead on anode (alternator side) on digital gives conduction .7 volt or 300 ohms . Black lead on anode on analogue meter gives conduction reading. Regards will

<hr width=100% size=1>

 

[roger]

Are you sure you dont have a possible current path of
battery via engine control panel to alternator? I would imagine this could only operate if the engine key switch were on.

<hr width=100% size=1>Roger

 

[Cantata]

Well, I remain certain that the wiring is all correct. OK, I might be wrong, but I am very careful with this stuff.
I'll do some tests in a few days when this weather has blown through. Sounds to me like I will have to buy another diode, it just puzzles me that it may have failed without any apparent cause. Thanks to everyone who's commented.

<hr width=100% size=1>

 

[halcyon]

They do not have to have a obvious cause, they can just die.
Do not assume it has failed though till you can prove it, it may be a diode failure in the alternator. But if you have 12 volt on the blocking diode alternator terminal, with the alternator not connected, then it's the blocking diode.

Brian

<hr width=100% size=1>

 

[andyball]

you can use a multimeter on ohms function either digital or analogue but expect to get a reading of some hundreds of ohms in forward direction and infinity in the other


All the digital meters I've used don't have enough voltage across the probes on ohms scale to turn on a diode, many modern analogue meters will do the same......

<hr width=100% size=1>

 

[VicS]

That's the reason I said earlier in this discussion that you must use the diode test range rather than one of the resistance ranges.

<hr width=100% size=1><font color=purple>Ne te confundant illegitimi.</font color=purple>

 

[dick_james]

Very strange! In my thirty-odd years experience I have known plenty of semi conductor devices fail short-circuit. Certainly no myth.

<hr width=100% size=1>