Splitter Diode

[charles_reed]

I've got a 150amp 3-way splitting diode fitted.

For 15 years, despite getting quite hot it's worked fine, now it just gets warm.

Has anyone had one of these fail and what was the mode of failure.

In my case the Adverc just puts on its green light and produces no charge at all, on the system voltmeter, and the on board alternator puts out 13.1 volts, but fails to put any real charge in after about 25'.

The alternator is new and is producing <15.3 volts all the batteries are less than 12 months old, amounting to a nominal 330ah.
The switch mode charger, wired completely separately, is charging the batteries just fine, going to float at 14.1v and maintaining 13.6v.

 

[pvb]

I had a 2-way diode splitter fail on my parents' motor cruiser. The leg charging the domestic batteries failed open circuit. The Adverc sense lead was on the domestic bank. As the sense lead couldn't "see" the charging voltage, the Adverc just increased the charge voltage (probably to around 16+) and fried the starter battery.

I think today's trend towards using VSRs rather than diodes is a good development, although in fairness diode splitters don't often fail.

 

[VicS]

I would suspect that all three ways of the splitter have failed open circuit. Or you have a bad connection between the alternator output and the diode splitter.

( Why three ways? presumabably three separate battery banks ??)

The Adverc should be sensing one of the batteries via the red wire but since nothing is getting through the diode splitter it is cranking the alternator up to more than 15.3 volts ( you say <15.3 .. did you mean to say more than 15.3, ) in an attempt to get the right reading on the sensing connection.

Easy enough to check the diode splitter with a multimeter with a diode test range.
Disconnect it and each channel should read infinite resistance one way and about 700 ish the other way.

( Ohms ranges of digital meters do not necessarily work for diode testing)

 

[halcyon]

on the system voltmeter, and the on board alternator puts out 13.1 volts, but fails to put any real charge in after about 25'.

Can expand on what 13.1 volt, is it battery voltage? in which case a bad connection from alternator to splitter is possible.

Loosing 2 or 3 diodes at the same is very rare, more likely internal link to diodes failed.

Brian

 

[Paddingtonbear]

Looks like another visit to Chris Gibson's Smartgauge site

 

[charles_reed]

13.1v is system volts at the feed bus, using the NMEA 2000 voltmeter, which I've found extremely accurate. Battery volts vary according to charge level in batteries (12.0-12.9, resting). I have 3 batteries, 1 of 2 is always reserved for emergency, unless on switch-mode charger.
The straight test of the diode on/off won't meet the needs as the failure appears to be progressive as the unit works, the drop in volts, on start up is about 0.4, from input to output. Fortunately the diode is a large, remote-mounted unit.
I've not heard of a splitter diode failing in this manner, hence the post. It's no help being single-handed working the boat up N through the Dodecanese - lots of obstructions and varying wind-strengths from F4-6, so I can't be below playing electrician.
There are two elements to the problem - the Adverc not working, probably down to mis-wiring by the electrician in Ag Nikoloas and the inability of the on-board controller to put an adequate charge into batteries that are clearly not fully charged.
Not life-or-death - the PV panels give me enough top up for 3-4 days under sail and there are lots of stopovers on the way with shorepower.

 

[david_bagshaw]

I've got a 150amp 3-way splitting diode fitted.

For 15 years, despite getting quite hot it's worked fine, now it just gets warm.

Has anyone had one of these fail and what was the mode of failure.

In my case the Adverc just puts on its green light and produces no charge at all, on the system voltmeter, and the on board alternator puts out 13.1 volts, but fails to put any real charge in after about 25'.

The alternator is new and is producing <15.3 volts all the batteries are less than 12 months old, amounting to a nominal 330ah.
The switch mode charger, wired completely separately, is charging the batteries just fine, going to float at 14.1v and maintaining 13.6v.

I have had one fail in exactly the same way.

I was told by a reliable engineer it is not uncommon for them to develop a greater resistance, which lead to a lower charge current. batteries fully charged over a longer time, so there was no apparent voltage problem.

 

[charles_reed]

More grist to the mill.

The unit is a Sure Power Industries Inc. Multi battery Isolator Model #1203

Resistance reading are as follows:-

1. 10.7K ohms
2. 9.8K ohms
3. 9.8K ohms
Seem a trifle high to me.

There is a Sure Power Inc. of Portland, Maine, who are UPS producers and Sure Power Industries Inc, of Tualatin. Oregon are listed as a tier 3 supplier in Motor Industry Handbook. Unfortunately their website address appears to be NU.

Anyone have any figures as to what the resistance should be?

 

[halcyon]

1. 10.7K ohms
2. 9.8K ohms
3. 9.8K ohms
Seem a trifle high to me.

You need to check them with the probes the other way round as well, you should have a few mega ohm resistance. You are only checking that you have flow one way, and blocking the other, not what value you have.

Simple answer if you can get to unit, is to take the terminal of split charge output, and bolt on the input stud from the alternator. Just by-pass diode and charge that bank. if reserve banks are oakay, just swop over the service bank. The other banks will be safe if the diodes are working, and blocking current flow.

Brian

 

[VicS]

Anyone have any figures as to what the resistance should be?

Each diode should have a very low resistance when measured in one direction and very high when measured with the probes reversed.

For all practical purposes zero resistance one way and infinite resistance the other way.

Its best to use the diode test range, if you have one, on a digital meter because the ohms ranges of digital meters often don't work for diode testing.

These readings were all with it completely disconnected?
If so based on them alone I'd say the splitter is defunct !

 

[VicS]

Anyone have any figures as to what the resistance should be?

As Halcyon says each diode should have a very low resistance when measured in one direction and very high when measured with the probes reversed.

For all practical purposes zero resistance one way and infinite resistance the other way.

BUT it is best to use the diode test range, if you have one, on a digital meter because the ohms ranges of digital meters often don't work for diode testing.

If these readings were all with it completely disconnected based on them alone I'd say the splitter is defunct !

 

[old_salt]

You don't need a meter to test diodes just a test lamp set up made up from a lantern battery with with a 2/5w bulb or a LED with probe leads.
An LED if you have one will give a more sensitive indication if there is any reverse flow in the diode on test.
Hook up the diode to the test rig and you will light the lamp one way reverse it and it won't, if it does then the diode is duff.
You don't have to worry about getting the connections the correct way round the reversal of the probes will do the necessary test.
You can go further and put it to a more serious test, then use 12v from a main battery with a heaver load.
I have a simple set up with an old car head lamp bulb wired so as both elements can be used if required 85w or 100w on it's own or the 85w+100w
= a 185w test lamp load it will sort most problems of dry joints, iffy open circuits and bad connections. Quite often without disconnecting anything.
also useful to do a complete discharge on the older type rechargeable battery's that can get a hung charge in hand held items.

 

[Billjratt]

I've had this problem.
Resistive measurement can be confusing using either the ohms range or the diode check range when the diodes are cold.
I believe they may be breaking down due to age/stress (high charge current)
You are right to believe temperature means working (the alt should be too hot to touch if it's busy)
The measurement you want is the VOLTAGE DROP across the diodes when they are in fail* mode. It would normally be 0.6 to 1.0 volts or so. If the alt is putting out 15.3v the batteries should be getting > 14v.
*When I say fail mode, you may have to wait until the fault appears, depending on how bad the diodes are.
For the moment, you can jumper the batteries (all 3 banks) direct to the alternator, at the diode block, and maybe the Adverc will spring into life again (if the sense wire is connected).
Then when you stop the engine open the jumpers again.

 

[TrueBlue]

The alternator is new and is producing <15.3 volts all the batteries are less than 12 months old, amounting to a nominal 330ah.
Th QUOTE]

Wondering -
Is the New alternator exactly the same as the old, or one that "does the job"

Even within a model range there are sub models that require an excitation lead to be connected and others not. This is independent of any external sensing required.

You would expect the external field wire to the Adverc to override that - but in my recent experience it doesn't.

Where does the 15.3 volts come from? alternator not connected, perhaps?


I only raise the issue as most of the other possibilities have been covered.

 

[charles_reed]

Postscript - not the diode, but the wiring of the alternator.
Adverc now works 14.48v @ feed and 13.85v out.

The give-away was when I tried to connect battery feeds and alternator input and the engine warning buzzer came on.

 

[pvb]

Postscript - not the diode, but the wiring of the alternator.
Adverc now works 14.48v @ feed and 13.85v out.

The give-away was when I tried to connect battery feeds and alternator input and the engine warning buzzer came on.

Where's your Adverc sense lead connected? I'd expect you to have around 14.4v at the actual battery posts.

 

[charles_reed]

Not after going through the diode - we'll see how it goes in use. From memory (don't forget I put this in 19 years ago) the sensor is on #1 battery. For me 13.85 volts is a bonanza.

Regarding diode testing, as so often happens theory and practice don't fully meld.

In the end I did the tests using leads from the battery to check if it was working. The use of a digital meter produced a highly variable number of readings, dependent on how long the probes were connected (started high and dropped).

 

[pvb]

From memory (don't forget I put this in 19 years ago) the sensor is on #1 battery. For me 13.85 volts is a bonanza.

It may be a bonanza, but it certainly isn't enough! The Adverc sense lead should be on your domestic battery bank and, if the system's working properly, should result in around 14.4v at the batteries (and around 15.0v feeding into the diode splitter).

If you're only getting 13.85v out of the splitter, it's likely that the Adverc isn't working. You could try taking the multipin plug out of the Adverc and seeing whether this changes the voltage readings. If it doesn't, the Adverc isn't working (possibly the result of wiring faults, as the units themselves are rather over-engineered).

 

[ghostlymoron]

I suspected my 2 - way diode splitter was faulty. On checking with multimeter I found no output on one diode so replaced it with a VSR which works fine with no voltage drop.

 

[halcyon]

Not after going through the diode - we'll see how it goes in use. From memory (don't forget I put this in 19 years ago) the sensor is on #1 battery. For me 13.85 volts is a bonanza.

That's good, keep you out of trouble till you have time to sort out.

Brian