I didn't listen !

[coldspot]

I was being stupid

DanG warned me but I had to try it anyway.

I hooked up Inverter to battery bank and then fed the 120 V AC

into cabins AC in power pole that we just plug into any generator

and that powers the 120 VAC in the cabin.

Well, it must have a neutral to ground somewhere.

It smoked the "IRF3205"'s in the inverter

7 out of 8

All but one has this number

IRF3205

435. k
     
2. L 6G
   
   

The other one has this number

IRF3205

504. k
     
6. Z 5T
   
   

alldatasheet.com

tells me that the second number is a date code

and third number is a lot code

"  Power MOSFET(Vdss=55V, Rds(on)=8.0mohm, Id=110A��)"

is what they call it

I should be able to find some of these and try to repair this one

I've looked thru my collection but only a couple of IRF830

So I'll have to buy them.

Any other warnings out there?

I'll listen better from now on.

There are a couple of resistors pretty close to where the mosfets cooked

I better at least check them for proper values to see if they are ok

I've been looking at some "AIMS" inverters at invertersrus

and will now order one, really liked the 1500 with AC thru plus battery charger but

not for the cabin as it is now.

maybe just the 1250 thats allways on sale

Boy that sure bummed me out this weekend.

learning as I go

[kurt]

lol that grounded nuetral will get you every time.

[TomW]

coldspot;

Well, anyone who claims they never fouled up is full of bull sausage. I destroyed a couple very expensive pieces of avionics gear in my career in that field.

As long as you learn from it the experience is not all bad.

Carry on and don't let it get you down.

I bet from now on you are careful and check for how the wiring is configured before powering up. Nothing like burning your hand on the stove to teach you about "hot"!

Cheers

TomW

[stevesteve]

When I was younger, a friend and I were testing a simple R/C circuit in a physics lesson. Circuit had been running for a couple of minutes off a NiFe cell when we turned on the school's brand new oscilliscope. <<BANG>> Magic smoke and we though "OMG we've killed the new 'scope".

It turned out we had wired a large electrolytic cap backwards and blown up ~$1 of component, not the expensive shiny bits.

If you are still safe and haven't burnt down the place then hopefully it can be a learning experience and you can fix the kit.

[dinges]

"I didn't listen !"

Nobody ever does....

Hopefully you can get it working again and it's only the FETs that were blown.

It's easy enough to check whether your ground is connected to neutral in your system. It's just that 1) you have to know that this might be an issue 2) you have to check whether ground and neutral are connected somewhere in the system.

Apparently you knew 1) but didn't 2...

Glad to see someone's not too ashamed to post his boo-boos here. Hopefully it has made a few other readers alert of this potential problem.

And yes, I tend to be stubborn at times too. But, lessons learnt the hard way tend to stick very well...

Peter.

[Ungrounded Lightning Rod]

Well, it must have a neutral to ground somewhere.

If your wiring is to code (or was before you started screwing around with it B-) ) your ground and neutral will be "bonded" in exactly ONE place.  Probably the main breaker panel.  (This is to keep lightning currents out of the neutral wiring between multiple bonds and load currents out of the protective ground wiring.)

You need a disconnect or transfer switch for the line before plugging into your alternator.  If you use a three-phase disconnect you can use two circuits for cutting the hot wires and the third for cutting the neutral bond.

Rembmer that you DON'T want to unbond if you are using a generator.  You can parallel a single-phase disconnect across the bond-breaker contact and label it "alternator (off) / generator (on)".

Also remember that you want to use a DISCONNECT (switch), not a BREAKER, for cutting the bond.  The LAST thing you want is for a breaker to cut off the bond due to overcurrent - right when you need it due to a faut.  B-(

[coldspot]

ULR-

"started screwing around with it"

Did not, haven't touched anything but the power-IN pole.

"plugging into your alternator"

No, we are working with Inverters here. lol

The place is solar powered.

 (But I will add wind gen to boost the power,

They love the idea of a dump load as far as dumping to ground and maybe helping

to drive moles n other ground digging rat type pests away. lol!

I've yet to research about using ground as a dump load but think I've read something

about it before.)

NOT my cabin, girlfriends, she and her brothers built it.

Not on grid.

"wiring is to code" NEC

Thanks for the switch idea, that, I might be able to talk her into.

I was just to excited about not having the generator running.

(Have been looking at to many modern Inverters on-line and forgot

about this one being not that new, {04-2005} and ment for

auto/boat use)

AIMS, 1500 watt with AC Thru n battery charger, was what got me

thinking less than I should, I'm now thinking that that one also

would need a un-bonding switch.

Now that I'm fixing this inverter, I also took apart one that her brother killed

a few years back and it needs 4

motorola TMOS E-FET

MTP50N06E's

But so far that number isn't being found at any places that sell parts

the IRF3205's I need are easy to find but only would like to place one

order and am still searching.

Thanks to all!!!!!!!

Happy 4th of JULY !!!!!!!!



l8r

[Nando]

This MosFet [(Mouser part number) 844-IRF3205] is made by IRF and MOUSER.com has them in stock around $3.10 each ( USA dollars)

Better make sure you change the Resistors, they MUST be bad.

Depending on the circuit, you may have other things bad, LIKE Fuses, MOV's,  current detector and protectors plus the MosFet drivers and WHAT else ??

Nando

[coldspot]

nando-

what number did you find for me?

IRF3205

MTP50N06E

[Nando]

IRF3205 As indicated in my message, the first digits is Mouser's part added to IRF3205

MTP50N06E is 60 V 50 amps and you need 100 + Amps

Nando

[coldspot]

Thanks nando

but

That part I found for about $1.60 each

The other part is for a different inverter as I posted

While doing one I fished out an other one that smoked

in about a half hour when new

I found the problem on it to be crappy soldering from the factory

I'm pretty sure thats all it needs is the 4 motorola

MTP50N06E's

and yes they are only 50A, 60 V

it's a smaller inverter and older

500Watt 800Max

While the first one is a 1000 watt 2000Max

and made last year

I'm also pretty sure that all it's going to need is the

MOSFET's

maybe the resistors

I was quick to realize that I was letting out the very important

magic smoke and pulled the battery clips quickly.

lol

I was freaking out about this all being in HER battery box with the

solar charge controllers and didn't even tell her about it untill today.

This all happened last saturday. She wondered why I was not enjoying the trip.

I didn't want to try to replace the stuff in that box! So I figure that

I'll be setting up my own next weekend.

Her stuff is all

Vitel Inc.

two VREG-JL1's with a VREG-NXS also

Stevens now owns that brand

I'm still waiting to hear back from them on a NXS manual

DL.ed the JL1 manual

I don't see how a couple of 20 watt units can control a 200 watt system?

O'well, my charge controller is only 25A @ 12 VDC,(300Watt).

I was going to order the Amp expantion for it but now might just get a

different inverter, even if I can fix these.

I have a 75W, 400W and a 500W. with waiting for parts

a 500W and a 1000W. Still think I need a 1500 but

maybe I should just get a pure sine one next.

Still trying to figure out what the digital Vs pure is?

thanks

[ghurd]

"Hopefully you can get it working again and it's only the FETs that were blown."

I'll second that.

I don't recall one that had blown ONLY the fets. Ever.

Its usually a chain reaction where the fets show the obvious damage.

The cost of repair parts can cost more than a new inverter.

G-

[TomW]

Its usually a chain reaction where the fets show the obvious damage.



Yep, Just because the Fire Emitting Transistors blew up does not mean that is the only failure and often it is just a symptom. Cascade failure often happens in that type circuit.

Good luck.

Cheers.

TomW

[oztules]

Coldspot,

Whenever I've fixed units that have blown the fets, the little driver transistors have been subjected to the full 340v dc (240v system .. the dc volts will depend on what waveform they are trying to establish. square wave may be as low as 240v, modified wave may be >300v etc.) and have gone awol as well... so don't be surprised if they are gone too. I also use a high value resistor between the hv dc and the input to the h bridge to hold the current down to milliamps until I achieve a well shaped waveform coming out, then drive it straight. This keeps further mosfet kills down to a minimum.

...............oztules

[coldspot]

ghurd-

oztules-

because I understand less than I'd like about what

oztules posted.

"The cost of repair parts can cost more than a new inverter".

yep, I see that even ordering on-line the cost would put me at

half of what I paid for this one.

I'll just order a new one and go from there.

I might some day get the parts and try to fix the older one and see if it still

wants to smoke or not, (parts 4 it cheaper than newer one).

Thanks for the help, you people and this forum are the reason I

still have a internet connection!!!

[DanG]

Is there anything more memorable than bzzzt >pop< ?

you want maybe next time I should call Vinnie on you?

It looks like you have a nice $50 heatsink for your next mill's rectifier circuit!

I forget where but once upon a time far far ago even I have slipped a jumper between white and ground repairing a worn out two-wire wall outlet by installing a three-prong outlet in a house without true ground circuits. Something about leaky bearings and 120V tingle trying to use shop tools. The moral of the story is suspect some 'joe whenever' has been there before you - even if you lift the ground bond, walk the circuits out, open work boxes, pull recepticals and visually confirm before you commit to re-energizing someone else's work.

Reminds me of when I bought my inverter, he had it in the center of a garage on a temporary batt bank and hooked two heat guns on high and three halogen work floods and was going over minutia details with the loads running when his construction size kerosene heater kicked on and all I saw a very large surplus of orange glow, heat and noise approaching from behind the inverter... Did a nice little two-step dance figuring it out much to the amusement of the seller.

Sorry your weekend fizzled.

~

[dlenox]

TomW,

From what I read then in the main panel if a neutral to ground bond is used should it be separated before connecting an inverter?

Other than visual inspection is there some way to test with meter to see if neutral is bonded to ground somewhere?

Dan Lenox

[TomW]

Dan;

Yes. Powered down you can check continuity between them. Thats with an ohmmeter between the neutral and the ground.

I am no electrician and certainly no expert. I am also leary of passing out this kind of advice to folks when I am unsure of their abilities. Mistakes can kill or cause property damage. So, if in doubt find qualified help. It is not difficult but must be done properly to be safe.

Generally the neutral and ground are only bonded together in one spot [the breaker panel]. I never trust anything is wired properly because there is no way to know who messed with it in the past. If it is bonded in the breaker panel only and you remove the ground where it is connected to the neutral connection you should not see continuity between neutral and ground anywhere in the system. Check this to be sure. Good luck with it.

Cheers.

TomW

[dlenox]

TomW,

Ok I understand now neutral to ground only at a single location.  But what would happen in the following instance:

    Neutral to ground bonding in main circuit panel,

    One or more outlet(s) that also has ground and neutral bonded?

Is in necessary to have to check all outlets, etc. to ensure that someone did not wire them up improperly?  An ohmmeter would not detect this back at the circuit panel.

I will be using a master electrician work with me on the connections, thanks.

Dan Lenox

[TomW]

Dlenox;

After you separate them in the breaker panel then test continuity between those wires. If it is showing continuity you have one or more circuits with them tied together. Debug that by checking each circuit until you find the culprit. Pretty simple really.

Tell your electrician to look for shorts between ground and neutral. This should be something he will understand, I would hope.

Cheers.

TomW

[dlenox]

TomW,

Thanks that makes perfect sense.

I did not think about breaking the ground/neutral in the circuit panel to test the rest of the system.

Sometime the obvious isn't.

So I guess that my assumption that an outlet that bonds ground/neutral could potentially take out an inverter holds true?

Dan