Odyssey Battery has switched Polarity - Diod Meltdown -fire

[Fortysix12]

655 TT on my X Super Cub, I'm on my Second 680 Odyssey, not quite 7 month old,the first one I retired to light duty and is working in my forklift.

Jumped in the cub to take an early morning cruise, turned on the master no master click -no power- no starter. Shut the master off and started smelling smoke. But master off and smoke continues. I raise the back seat to gain access to the battery system and I spot a flaming wire. But hey the master is off what up. Fortunately I was able to extinguish the flame but not until it had melted the conductor in half. I''m scratching my head wondering what has burned and what caused it and how extensive is the damage. Here's what I've discovered and who can believe it.
1. The recommended diode from the Cont Duty SOL coil to to ground had a melt down,ok that's easy but man is this dangerous
2. Checked voltage on the battery and it was down to below 12, oh low battery I will hook my my charger.
3. Sparks fly when I try to correctly hook up the charger like I was connecting them wrong. Whats' up with this,battery charger must be on the blink
4. Called an electrical engineer friend and started checking resistance and other things
5. Check voltage again and discovered neg reading when normally testing voltage level , is this possible
6. checked voltage the other direction and get a positive 12.4 volts.,yes, red on neg blk on pos and I get a positive reading on voltage

Plan on discussing all of this with Odyssey or Energy sys.

So how in the world did this happen. I haven't a clue yet. but Here's what other discoveries I have made that experimental builders might want to consider

The cause of the diode failure was a reverse polarity so when i turned on my master power ran the other direction and burned it up, but it caused a fire. I understand that when I diod fails it can be hot real hot and its an avalanche type failure . A possible alternative is to fuse the diode assemble just in case of a diode falure.
The other thing is that this fire was right under the back seat.Had it not been so obvious or happened in the air it would of been a very serious problem Nothing but foam rubber and plywood straight up over my solenoid bank and battery. Not sure what I'm going to do about this.

The only thing I can think that may of caused all this is I charged the battery two weeks ago after leaving the master on and I charged the battery in reverse but everything I have read says this is impossible in addition I have smart charger that knows when you have hooked it up backwards so I'm baffled. Charging that day was a normal 2 amp feed for 12 hours or so. Completely normal.



Keep you posted once I figure out what has happened.

 

[8GCBC]

Seems like something positive touched something negative? Shorted to the airframe?

 

[180Marty]

Just found this.

The second possibility is reversing polarity after the activation process. This is also rare, as it requires a sequence of errors to be present after the installation of the battery. The only way for this to happen would be to completely discharge the battery, either by leaving the key on, or by an unnoticed dead short that completely dissipated the charge over a few days. After that happened it would appear to be a dead battery.
Remember, a completely discharged battery is nothing more than an empty vessel. In order to gain a negative charge, it would then necessitate being hooked up backwards, and charged that way. So the real question here is: how can a battery reverse polarity after it has been installed? That same previously discharged battery would then be vulnerable to reverse charging, either by connecting the battery charger backwards, or by a charging system that reversed polarity (very rare, but still possible).

http://www.batterystuff.com/blog/battery-myth-can-a-battery-reverse-polarity.html

 

[mike mcs repair]

hmmm

I think #6 caused #1... (the diode was only made to catch the small backEMF spark....

hmm.... never heard of reverse charging a dead battery.. but just google "charge a dead battery backwards" sounds like it...

maybe that auto sensing charger had nothing to sense..

 

[qsmx440]

If your not totally sure of your volt meter be sure and test a car battery with it and see it is reading the correct way on it.

 

[sjohnson]

I wouldn't worry about putting a fuse to protect the snubber diode. Diodes will almost always (engineers almost never use absolutes) blow open, and there is so little mass that the fire risk is negligible. In other words, the diode acts as its own fuse. They fail in normal operation, too, and we often don't notice.

 

[PerryB]

If the battery is dead, flat dead, you CAN charge them backwards. I've seen it done. This almost has to be what happened.

 

[8GCBC]

Will a "smart" charger allow a reversal of poles? There will always be as slight rise in potential between poles even after a complete dead short with a screwdriver on a "dead" battery. Unless there is an internal short (scary). :evil:

 

[fobjob]

Back in the old nicad days there was a phenomenon of cell reversal. The battery had to be mostly discharged, then when it was charged normally
and correctly the weakest cell would reverse on charge and let all the factory smoke out at some point......usually during the charge sequence...

 

[Fortysix12]

Will a "smart" charger allow a reversal of poles? There will always be as slight rise in potential between poles even after a complete dead short with a screwdriver on a "dead" battery. Unless there is an internal short (scary). :evil:

This is scary. I bring this to the group for the thought that if it can happen to me it can happen to anyone using an Odyssey. The point that I wanted to emphasize is what happened to the diode. It was the cause of my little wire fire and all because it was waiting for a reversed current when I threw the master and shorted to ground. I will fuse my next diode if for no other reason to protect against catastrophic failure even under normal conditions. I've not heard back from odyssey as of yesterday.

 

[Paul Heinrich]

Which side of the diode would you put a breaker if either side could be "hot"?

 

[qsmx440]

Which side of the diode would you put a breaker if either side could be "hot"?

As long as it's all insulated it wouldn't matter since the current is the same everywhere in the diode circuit but preferably you would run it coil terminal -> fuse -> diode -> ground. I would use a fuse and not a CB and I can imagine everything inside a Bakelite inline fuse holder.

 

[Paul Heinrich]

Yes, but if the polarity gets reversed, isn't the fuselage then "hot"? If so, wouldn't the fuse would be on the wrong side of the diode?

Getting back to the original issue, did you figure out what happened to start the fire?

 

[irishfield]

I noted today..... I pulled my electronic charger off one airplane and within a minute had it moved to my boat to charge the battery there. On connection of the + cable it gave off a GOOD spark upon connection (and was not plugged in) so suspect part of the normal sense system of these new fangled chargers, so I wouldn't read too much into your spark at connection.

Reading you charged a completely flat battery a couple of weeks ago. Smart charges generally can't even start a charge without sensing a voltage to start with. So if it was totally dead.. you connected backwards and left it that way it may have slowly polarized the plates and then charge it BACKWARDS. If you said a normal charger I would say DEFINITELY possible... with a smart charger certainly possible.

 

[mike mcs repair]

This is scary. I bring this to the group for the thought that if it can happen to me it can happen to anyone using an Odyssey. ....

that's not really a fair statement.... after looking it up, as posted bellow, it can happen to any dead battery like these(lead/acid/gel)

the small diode did what they do..... I got some monsters that won't pop

 

[qsmx440]

Yes, but if the polarity gets reversed, isn't the fuselage then "hot"? If so, wouldn't the fuse would be on the wrong side of the diode?

Getting back to the original issue, did you figure out what happened to start the fire?

Notice I said "if everything is insulated". Your just trying to stop the diode from starting a fire. One of the laws in a DC circuit is the current is the same everywhere in a strictly series circuit so anywhere the fuse is located in line it will protect the circuit from a short/fire (If I remember correctly a guy named Kerchhoff was involved) . Think old fashioned Christmas tree light string. For you under 50 somethings a bunch of low voltage bulbs were wired in series to run on 110vac. When one burned out they all quit and you had to start at one end with a good bulb and try it in each position going towards the far end until you found the bad one. You just don't get great electrical lessons like that anymore.

 

[Fortysix12]

I talked with Odyssey battery today. They are sending me a technical discussion on how and why it happened. Here's a bit of what he said.

I didn't connect the charger wrong. With master left on say a day or longer the resistance coil in the solenoid depleted the battery to less than .1 volts.
Second, the battery determines polarity during the charge not the battery charger. So something to do with lead batteries, they plates flip polarity and cause a domino effect. They said this in not uncommon in cases where a battery is completely dead.

I'm going to make at least two changes in my electrical system. Fuse all diodes, and install a proper size breaker on the buss feed from the Cont Duty Solenoid to the fuse buss, and a fuse link on the master switch line to the solenoid.


I will post his discussion when I receive it. Did you know that Odyssey battery started as the Optima spiral wound battery, a division of Gates Rubber company.
Odyssy type batteries are used in the F15 and F16.,F22. and lots of other ground vehicles and airplanes.

 

[Cub Builder]

This is fascinating. I'll be looking forward to reading this as I have left the master on and run my battery down twice now, although not completely dead. Perhaps I was close to disaster, but not quite dead enough to cause it to happen.

-Cubbuilder

 

[qsmx440]

Perhaps a new procedure should include checking for some battery activity first (turn master on and check for 10+ volts), if low first charge with a small charger until the battery has some charge then switch to the smart charger.

Now if it wasn't for the .3 volt drop through the diode I would suggest an inline larger diode or bank of diodes to ensure the smart charger only charges in the correct direction. I can't figure out why the manufacturers would make a DEFECTIVE charging circuit like being described. I guess they are being built for the low information voter that can't figure out which lead goes to the positive.

 

[Aviator]

... With master left on say a day or longer the resistance coil in the solenoid depleted the battery to less than .1 volts...

Something to keep in mind is that the dropout voltage is 4.5V (at least on Lamar's 12VDC relays). So even if your master switch is ON, the relay will (should) release well before the battery gets down anywhere near .1V. So the leak must have been somewhere else(?)
http://www.lamartech.com/solenoids---contactors.html

 

[Cub Builder]

Something to keep in mind is that the dropout voltage is 4.5V (at least on Lamar's 12VDC relays). So even if your master switch is ON, the relay will (should) release well before the battery gets down anywhere near .1V. So the leak must have been somewhere else(?)
http://www.lamartech.com/solenoids---contactors.html

While the relay drops out, the magnetic coil in the relay is still active (one side to the battery and the other side to ground through the master switch), so will still completely deplete the battery. The coil doesn't have enough power to pull the relay closed, but it is still draining the battery.

-CubBuilder

 

[mvivion]

Ya know.....the BEST way to prevent this is to not leave your master switch on overnight....:roll:8) I know, smart ass post, but wait...

For years (actually since the last time I left the master on overnight) I have ALWAYS left the strobe lights or rotating beacon switch in the ON position ALWAYS except in clouds at night, of course.

So, I get back from flying and I'm running late, so I rush through post flight duties, whether tie down or hangaring.....and as I walk away, HOPEFULLY I will notice those strobes flashing in my face, even in daylight.

Leaving that strobe switch on doesn't hurt a thing, and a bonus is that as soon as you engage the master switch, anyone around the airplane will (should) see the strobes/beacon come on and realize SOMEthing is about to happen.

It's worked for me for thirty years, and in a bunch of different airplanes.

FWIW

MTV

 

[mike mcs repair]

Another project that I have thought about and never followed through on, would be master switch minder or such, that would shut master off if engine is not running for say 10 minutes... Have it sense a plead wire like the electronics international tach does... Make it out of a simple PIC processor or such and a solid state really to be put in master switch wire..... And when it shuts off master it also looses power so it's not a constant draw.....

 

[180Marty]

Both Mikes, good ideas. mcs, a Murphy switch like used on an unattended diesel engine that shuts the engine down when oil pressure drops or temperature rises could work also. A friend of mine used to have a 4020 John Deere on a corn dryer with those gauges.

 

[8GCBC]

I use a checklist. Highly recommend.

 

[Fortysix12]

Mike,
I'm with you on the strobe lights. I leave them on always for this very reason, but this once for what ever reason they were off and for what ever reason I didn't use the 3M's at shut down. Mixture Mags Master. Actually I think I was showing someone the avionics and just plane forgot to kill the master.
The Stobe lights/rotator beacon have saved my bacon more than once over the last 30 years.

 

[qsmx440]

Another project that I have thought about and never followed through on, would be master switch minder or such, that would shut master off if engine is not running for say 10 minutes... Have it sense a plead wire like the electronics international tach does... Make it out of a simple PIC processor or such and a solid state really to be put in master switch wire..... And when it shuts off master it also looses power so it's not a constant draw.....

Hmmm. Maybe something like the devices in this article?
http://www.hendonpub.com/resources/article_archive/results/details?id=1420

 

[Aviator]

While the relay drops out, the magnetic coil in the relay is still active (one side to the battery and the other side to ground through the master switch), so will still completely deplete the battery. The coil doesn't have enough power to pull the relay closed, but it is still draining the battery.

-CubBuilder

Good point. I stand corrected. Thanks. Having reread the original post, it's clear the diode was between 12VDC and ground, NOT 12VDC and the + terminal of the coil. Diodes are sometimes stuck there to completely isolate the relay if it drops out due to low voltage. If the voltage drops below, say, 4.5V and the diode's threshold is 4.5V or more, no current flows to the coil. Their down side is that if they fail the relay might open (depending on the mode it failed in).

 

[Carey Gray]

Cubs aren't supposed to have batteries.