6k inverter putting current through mounting bolts

[Nilao]

What would cause current to flow through the mounting bolts? It is getting hot enough to melt plastic there.

Edited September 24, 2022 by Nilao
Add info

[Nilao]

It is grounded by a 4/0 wire and the most watts it has seen is from my 1500 watt microwave.

[Sid Genetry Solar]

This would likely have to be a grounding issue with the external system; I do not see any way that the inverter itself could be energizing the case.

The only internal connection to the chassis should be just a little 12AWG wire connected to AC Output Neutral through a 10nF decoupling cap (Rev. B and C).  This won't pass much of any power; it's there for UL requirements that the chassis should basically be "centered" to the AC output potential.

You can try looking through the side grilles to make sure that none of the transformer primaries are touching the chassis--but even if they were, they could not put a potential through the chassis mounting bolts (as the transformer is isolated between primary->secondary).

[Sid Genetry Solar]

Now, if you can rule out external grounding as being a fault, and the underside of the inverter chassis is very hot (but not the sides of the case), there's a chance the transformer bolt insulator is broken--or something's amiss in that general area.

Another possible issue would be if the inverter is bolted down to a metal surface, and the head of the transformer bolt (on the underside of the inverter) is in contact with said metal.  That will create a very high heat zone across the underside of the inverter; it can be very dangerous (and worst part is that it doesn't increase no-load draw very much!) 

 

Worth noting that on the redesigned GS inverters, I moved the insulator to the inside of the inverter...completely solving this potential issue.

[Nilao]

Is the case bonded to the negative terminal? I'm thinking it's possible my cab to chassis ground may be faulty and it's using the inverter as a ground path.

 

It is on a rubber mat with half inch cotton batt insulation underneath.  Only place the case is in contact with metal is the self tapping bolts.

[dickson]

Is the case bonded to the negative terminal?

The battery dc  negative should not touch the case .   The AC ground  is bonded to the  case .    RV chassis ground  is not ground for the  inverter .    Be careful  of RV  grounding   and most RV  has GFCI  to be safe .    

[Sid Genetry Solar]

56 minutes ago, Nilao said:

Is the case bonded to the negative terminal?

No.  Like I said above, the ONLY internal connection to the chassis SHOULD be to AC Output Neutral through a 10nF (0.01uF) capacitor.

I don't see how DC Negative could get shorted to the case, as the negative wires go from the battery terminals to between the heatsinks.

 

Check and see if the underside of the case (or as close as you can get to the foot) is hot with the inverter at idle.  That'd let us know pretty quick if the transformer bolt is at fault; worth noting, we are not aware of a single center transformer bolt isolation failure on a GS inverter--but that's not to say it can't happen.

[AquaticsLive]

14 minutes ago, dickson said:

The battery dc  negative should not touch the case

I was thinking of the same thing there may be a shunt involved as well.  Maybe pull the mounting screws out and check the case to ground with a meter to see if its a high voltage or low voltage issue. 

[Nilao]

Seems to be getting some ac bleed onto the case. I took the mou ting bolts out and went to lift the case up to feel underneath of it. All I can relate it to is the feeling of a low voltage current flowing though my hand. Enough to tingle but not enough to hurt.

[Sid Genetry Solar]

2 minutes ago, Nilao said:

Seems to be getting some ac bleed onto the case. I took the mou ting bolts out and went to lift the case up to feel underneath of it. All I can relate it to is the feeling of a low voltage current flowing though my hand. Enough to tingle but not enough to hurt.

Which there will be.  You'll feel the 10nF decoupling cap...

...I think you need to break out a meter and start measuring potentials (try both AC and DC) to figure out what's going on.

I would expect the following readings from the case to the AC output terminals (assuming your unit is 240v split phase output; if not, half the numbers here):

• L1: ~120vAC
• N: 0vAC
• L2: ~120vAC

And to the battery terminals, I would similarly expect about 120vAC to either.

Any notable DC potential (>3v) would indicate other issues--but this is way too low to be felt.

[Nilao]

L1 to case: 140.7VAC steady

L2 to case: 143.4VAC climbing to 159.8 before dropping to 143.4 and repeating.

N to case: 20VAC dropping to 17 before jumping back to 20 and repeating.

Similar readings to the negative terminal with the exception of L2 which is 105VAC to the negative terminal.

[Sid Genetry Solar]

Check and see if the underside of the case (or as close as you can get to the foot) is hot with the inverter at idle.  That'd let us know pretty quick if the transformer bolt is at fault; worth noting, we are not aware of a single center transformer bolt isolation failure on a GS inverter--but that's not to say it can't happen.

You can also measure AC voltage from the transformer bolt to the case; it should be ~1vAC.

[Steve]

Better option on the center bolt

Weld the nut to the strap and have the bolt be for shipping only

It won't be needed for support on an installed unit

Has no electrical function

Prevents critical air flow through the center of the transformer (the hottest part)

[Sid Genetry Solar]

9 hours ago, Steve said:

Better option on the center bolt

Weld the nut to the strap and have the bolt be for shipping only

It won't be needed for support on an installed unit

Has no electrical function

Prevents critical air flow through the center of the transformer (the hottest part)

That's an idea for sure...but there's 2 challenges presented by the "fully GS designed" upcoming units:

• The transformer bolt is now a carriage bolt.  Removal of said bolt would require opening the case, as there's no external head on the bolt anymore!
• Fans now blow from the side, not from the top.
  • additionally, I personally experimented with having a fan blow air directly through the center of a toroid transformer--and it did not increase the wattage capability by any significant amount.  This could be explained by the much smaller surface area of the inside of the toroid vs the outside: the more surface area, the easier it is to dissipate heat.

[Steve]

The outside will offer allot more surface area but the center will be much hotter

Really it needs to have flow on both

It may not effect it's capability all that much but it does help and isn't that hard to do

Consider a strap that has a ring in the center with 4 legs coming off it

It clamps across the top of the transformer and tight to the sides but not the center

The center bolt design acts as a thermos by  having an enclosed air space

It also eliminates the loop effect of the clamp (mitigated by the bolt insulator

[AquaticsLive]

The issue I had was that the transformer rubber pad PJ put under it was not a large enough diameter I pulled it out and raised up the coil and adding a 3D printed holder with air passageways under it for cooling and insulation.  The rubber pad on the bottom was not a large enough diameter so the coils were touching the case in a few spots not a good thing in a vehicle glad I caught it before it rubbed through the enamel. 

[NotMario]

8 hours ago, Steve said:

The outside will offer allot more surface area but the center will be much hotter

Really it needs to have flow on both

I did wonder why the big round flange was used rather than something more breathable.
With the straps, you could almost skip the flange and just use the bolt in the center holding the straps down. With narrow enough straps, good airflow could result, and ideally wide enough to not cut into the enamel.
Maybe they haven't found an economical enough way to rely on the straps to this degree.

[Sid Genetry Solar]

2 hours ago, NotMario said:

I did wonder why the big round flange was used rather than something more breathable.

I have no idea why...and then adding a big rubber pad to that.  And in some of the larger PJ inverters, you'll see a triple-cross bracket...

...when the real issue is that the 2-cross bracket isn't tight to the transformer.  The moment that transformer starts to move in shipping, it's game over.

 

2 hours ago, NotMario said:

With the straps, you could almost skip the flange and just use the bolt in the center holding the straps down. With narrow enough straps, good airflow could result, and ideally wide enough to not cut into the enamel.

That's GS spec: no flange, no big rubber mat on top, etc.  I've narrowed out the straps on the GS12 spec--and likely could narrow out the straps on the new GS6 spec if necessary.  Biggest difference on the GS spec is that I'm purposely making the transformer straps very difficult to put on--so they hold the transformer tightly as possible.

As a result, if my memory serves me right, we have not had a single incident of inverter shipping damage caused by a transformer bracket failure.

 

The transformer bracket mounting is not close to sufficient without a center bolt--remove that bolt, and the inverters will be internally demolished during shipping.  Probably a main reason that most other inverter companies go with an E-core transformer: those things are SUPER easy to mount!

[NotMario]

That's funny, i've looked inside the GS 100 times, and i never noticed that. Indeed, there is no flange. Just a strap with the center bolt holding it true.

I made this change on my U-Power when i had it apart. lol Not sure if i'd trust shipping it this way, i didn't know how much force i could apply on the strap.

[Sid Genetry Solar]

3 hours ago, NotMario said:

i didn't know how much force i could apply on the strap.

Load it down as tight as you dare: the strap is stainless steel, same as the chassis.  I don't think you'll be able to break it 😉.

The biggest challenge is the distance between the chassis bolt mounts and the first bend (on the PJ straps)...that large gap WILL bend under strain, and result in an even-looser transformer bracket.  Adding insult to injury is that the PJ spec has huge gaps between the strap and the transformer...rendering it basically useless.

[Nilao]

Finally got around to checking that bolt. It is slightly warmer than body temp all around the bolt and measures 1.0vac to the case. 

 

The mounting bolts holding the transformer to my rubber pad have gotten hot enough to melt the 1/8" aluminum floor under the mat. They are no longer doing any securement of the inverter other than acting as locator pins.

[Sid Genetry Solar]

39 minutes ago, Nilao said:

and measures 1.0vac to the case. 

This is exactly as expected (per transformer winding spec).  Worth noting that the transformer is all the way around the bolt, so it will warm the case as well.

If the bolt was shorted to the case, it'd cause significant heat in the bolt and case--and also would not measure 1.0v from the bolt to the case.

By "significant heat", I mean....you'll smell the heat within 10 minutes of power-up, and touching any affected metal will burn your finger.

 

I still do not think the inverter is responsible for the melted rubber pad/damaged aluminum floor.  If the inverter was generating that sort of heat, it'd have melted down long ago (as the transformer wires are aluminum!)

 

Your earlier test results of AC output potentials vs the chassis, and DC input potentials vs the chassis....doesn't turn up anything out of the ordinary.

 

Unless somehow the inverter is being used as a DC ground path, I don't see anything amiss.

[Nilao]

It is possible the cab is using the inverter as a ground path. I'll have to check voltage drop on my negative grounding studs on the cab. It sits on rubber isolators so the only path to ground is those studs....and now the inverter.

Edited October 19, 2022 by Nilao

[dochubert]

On 10/2/2022 at 7:57 AM, Nilao said:

L1 to case: 140.7VAC steady

L2 to case: 143.4VAC climbing to 159.8 before dropping to 143.4 and repeating.

N to case: 20VAC dropping to 17 before jumping back to 20 and repeating.

I'm probably missing something, but these readings worry me.  I'm surprised the inverter is still running with this going on.  I look forward to learning what the problem is in this setup, and hope you solve it soon! (before something blows up!)

[Sid Genetry Solar]

10 hours ago, dochubert said:

I'm probably missing something, but these readings worry me.  I'm surprised the inverter is still running with this going on.  I look forward to learning what the problem is in this setup, and hope you solve it soon! (before something blows up!)

I don't see why said readings are a concern.

Internally, there's a 10nF decoupling cap from Neutral to the case.  This should hold the chassis fairly close to the Neutral potential--which it is (~20vAC).  This should reasonably center the case between the two Line outputs--and it is.

Beyond this, if there's external grounding power being routed through the inverter chassis (intentionally or inadvertently), THAT could cause the issues being described here.  But at least so far, it does not appear that the inverter itself is responsible.