Electrical smell from Genset

[Juliet 15]

We're out in the San Juan Islands (it's beautiful - but cloudy..)

For the 3rd time in a row, I get a strong electrical when I run my Onan 15k genset. I've looked the thing over several times without seeing anything. It runs fine, everything is working normally, but the "hot electrical" smell is unmistakable. The other genset doesn't do it, and this Gen never did it before till about 3 trips ago.

I can't locate the specific source of the smell. Maybe it's the inverter/charger.

Anyone ever had a similar experience?

 

[High Wire]

Pull the generator terminal leads cover and any other cover on the generator end and look for burned wires and such with a bright flashlight. Check all terminal strip screws for tightness. Gen shutdown of course. Follow your nose. It could also be on the engine DC harness or run solenoid.

 

[Arc]

Capicators and bridge rectifiers can smell before Kaboom. The other possibility is the winding is shorting to ground. Depending on where the short is would still allow you to make power but not for long.

 

[Juliet 15]

Thx, good info. I spent time following my nose and found it's my Magnum 2000 inverter. It's working but the big #4 cables are warm (not hot). Pending inverter/charger failure maybe?

 

[boomerang]

An IR thermometer would be a good tool in locating your issue. Just the other day I had one of the ships I maintain complain of the generator popping a breaker. I opened the panel and found out immediately where the problem was by hitting the lines with the thermometer.

 

[diver dave]

12v @2kw= 166 Amps. I have the same inverter with 2/0 cable. Is yours a 12v system?

The tool of choice here is a hand held IR thermometer. The input bolted connections would top the list to measure. I have the same inverter; its been good to date.

 

[boomerang]

Guess which line was bad. ?

 

[diver dave]

The hot one!

Manufacturers most always list torque values for electrical bolted connections. Most of the larger stuff has some impressively high torque settings.
The more recent NEC code demands electricians use a torque tool.

 

[Art]

Check the switch: Gen Set to Shore Power on/off breaker. Had one go bad years ago. Hot electric odor. Found it had a melted inside when I replaced... should have tripped, but didn't.

 

[boomerang]

After tightening all 3 lines ,the temperatures came to a reasonable 115-120 degrees the next day. As cheap as they are ,an IR thermometer is a great tool to keep onboard.

 

[catalinajack]

After tightening all 3 lines ,the temperatures came to a reasonable 115-120 degrees the next day. As cheap as they are ,an IR thermometer is a great tool to keep onboard.

Yup, resistance creates heat. This is what I found when I removed the cover to my Westerbeke 7.7kw control box. That hokey connection you see - electrical tape removed - created so much resistance that the 60-amp breaker box two feet away had become too hot to touch. Fortunately, the breaker tripped. I had to replace the two wires leading to the brushes. Downright frightening it was.

 

[Juliet 15]

Bought an IR thermometer and will try it again today.

My inverter/charger is 32v, so probably the reason for the smaller primary leads.

Once I run it I'll post the results/temps.

 

[twistedtree]

Nice job tracking this down, and an excellent reminder to never ignore something that isn't right. This is the sort of thing that burns boats to the waterline.

 

[Art]

Note on IR guns. I've found: When batt gets quite low the temp readings can become sporadically inaccurate, misleading.

 

[diver dave]

Note on IR guns. I've found: When batt gets quite low the temp readings can become sporadically inaccurate, misleading.

Mine gives 10 deg difference between the black cooling hose and green paint on the same hose. Must be emissivity effects, but not sure what is closer to actual.

 

[twistedtree]

Mine gives 10 deg difference between the black cooling hose and green paint on the same hose. Must be emissivity effects, but not sure what is closer to actual.

The surface does seem to make a difference in the reported temp, sometimes massively so. I think IR guns are good for comparative reading, but not so great for absolute temp measurement.

It’s also easy to get lured into thinking that you are measuring the dot when it can actually be measuring a large area around the dot. Best reading are at point blank range.

 

[High Wire]

If you really want to track down something overheated fast borrow or rent an IR camera.

 

[Juliet 15]

Found it. See pics below. All cables were in the 80-90 degree range except this one. Where the two cables come together and cross what I believe is a current limiter this one was hot.


Once OT cools down I'll disassemble and correct it.

Thanks to everyone. My takeaway: if it smells like hot electrical, it is. Find it now.

 

[twistedtree]

That's probably a fuse, or perhaps a shunt for measuring the current. If the bolt connection isn't loose, which would make it a likely cause for the heat, I would go ahead and re-terminate both of those cable ends. Seeing the ends with tape rather than heat shrink leaves me suspicious of the lug installation.

 

[Juliet 15]

So it was a fuse under there (see pic). Not sure what happened, but I'm guessing the fuse didn't do its job. The terminal of the inverter-side cable got warm, but the terminal of the battery side got hot enough to melt. The hot side goes to the batteries, the warm side goes to the inverter.

The electrical smell only happened when running off the Gen, not shore power. Im surmising the Gen crams a lot of power thru there to charge the batteries, tho why shore power wouldn't do the same I don't know.

In any case, im not sure why the fuse didn't blow, instead, allowing the obvious overheat. Maybe the fuse was incorrectly sized.

 

[diver dave]

So it was a fuse under there (see pic). Not sure what happened, but I'm guessing the fuse didn't do its job. The terminal of the inverter-side cable got warm, but the terminal of the battery side got hot enough to melt. The hot side goes to the batteries, the warm side goes to the inverter.



The electrical smell only happened when running off the Gen, not shore power. Im surmising the Gen crams a lot of power thru there to charge the batteries, tho why shore power wouldn't do the same I don't know.



In any case, im not sure why the fuse didn't blow, instead, allowing the obvious overheat. Maybe the fuse was incorrectly sized.

Appears the bolted connection had high resistance. Notice the melt damage is not near the fuse link. My guess; low torque

The 400A rated fuse is 2x or more oversize for the 32V ??

 

[Comodave]

The fuse looks like a 400 amp one if I am reading it correctly. Take a lot to blow that fuse if it is.

 

[diver dave]

Would be a good time to upgrade to a T fuse system. This type separates the fuse bolt from the cable bolt. They also have much improved ability to clear large faults.
A 2kW 32V input should be using on the order of 100 to 150A fuse.

 

[Juliet 15]

Some clarifications:
1. It's my 12v bank that's affected, not the 32v. I previously said it was the 32v bank
2. The cables between the 12v battery bank and the Magnum 2000 (2kw) inverter/charger are 2/0 gauge.
3. At absorb charge (85 amps) one cable is reaching 350F at the connection to the fuse.

- I'm no EE; is a 400 amp fuse appropriate for this?
It's a bank of 4 6v Trojans connected to create a 12v bank

- Once it cools down I'll disassemble, clean, and retorque it.
If it's still hot after that then it must be the terminal-to-cable connection. Right?

- Someone felt a fuse was needed there. But the Magnum has 3 circuit breakers integral to the case of the inverter.
Is this fuse necessary at all?
I'm tempted to go straight from the inverter to the battery and cut out this fuse connection.

I appreciate any experiences youve had, or input.

Thx.

 

[Comodave]

Yes, the fuse is required to protect the wiring. Lookup the amp rating for your cables, my chart shows 300 amps. Then as long as the charger is below that rating, and it should be, then size the fuse appropriately. But don’t get rid of the fuse.

 

[High Wire]

The overheat is due the high resistance at the bolted/nut connection. Not the fuse portion inside the window or the crimped cable lug.
High resistance with high current causes heat. The fuse won’t blow because the current is not at the rated value. Btw 400a is above what the manual specified.
Disassemble both connections, clean all metal contact points bright and shiny with a wire brush, and tighten with an appropriate wrench.

 

[boomerang]

The overheat is due the high resistance at the bolted/nut connection. Not the fuse portion inside the window or the crimped cable lug.
High resistance with high current causes heat. The fuse won’t blow because the current is not at the rated value. Btw 400a is above what the manual specified.
Disassemble both connections, clean all metal contact points bright and shiny with a wire brush, and tighten with an appropriate wrench.

The picture I posted earlier of the lug with the 400° reading was indeed due to the crimped lug being bad. It failed after 8-9 years of constant high loads & varying engine room temperatures, I guess.
I agree with TT's previous comment of being leery of the taped lug crimp.

 

[High Wire]

The picture I posted earlier of the lug with the 400° reading was indeed due to the crimped lug being bad. It failed after 8-9 years of constant high loads & varying engine room temperatures, I guess.
I agree with TT's previous comment of being leery of the taped lug crimp.

I don’t doubt that. J15s photo shows heat discoloration from the right terminal to the point the window is yellowed too. If the crimp was bad enough to cause all of that heat then I would expect to see something on the heat shrink. Granted HS is designed to get hot.

 

[twistedtree]

For what it’s worth, now that we have eyes on the faulty connection, the discoloration does appear to be concentrated around the bolt connection. I agree that if the problem were the crimped lug, you would see heating effects on the tape and wire insulation.

I’d be real interested to hear how tight that nut is when you take it apart. Hopefully you will find it’s not tight, confirming the source of the problem.

All this really demonstrates the value in periodically inspecting high current connection while under load. If available, a thermal camera can make this much easier.

 

[Juliet 15]

To close the loop on this I did the disassembly and cleaning. The heat was centered on the bolt, and it was definitely not torqued tight. I'm going to have to assume, for now, that was the issue. I cleaned up all the parts and attempted to reassemble them but the plastic around the bolt in question had been damaged by the heat enough that it broke when I began applying torque to that nut.

New parts will be her tomorrow and I'll carefully reassemble everything, torque them well, then test it out under load. If it all works fine and there's no heat I won't come back to this.

If it still heats up and pints to another problem I'll fill y'all in.

Thx for the responses and input.