Problem with welder and Inverters

[ChrisOlson]

I know this is a crazy question but do you happen to have a link or maybe specs to the transformers within that inverter or those inverters?

I don't.  But I know there's three big transformers in there.  I seen 'em when we had the covers off installing our inverters.

That 50 amps on the inverter output could very well be 250 amps at 48vdc on the input

It's more than that.  I had 63 amps on the Fluke (with the house loads on).  So it's more like 700+ amps with dual inverters (on 24 volt), 350+ amps on 48V.  You get a voltage dip at the bus when they do that.  So even if the bank is at 25 volts, the voltage dip will probably pull the bus down to 23.5 (24 volt system) when that 700 amp surge hits.

This is what I've been saying - if your bank can't maintain the voltage above 22 input to the inverter (with these particular inverters) they go into a cascade and kick out because they "see" overload.  Already BTDT before I replaced my battery bank last April.  When we first put these inverters in on April 4 last year the dealer told me my battery bank wasn't going to cut it.  I thought, like DanB, that I had enough battery.  But the dealer was right.  $10,000 worth of batteries later I can run 'em at Full Dawg.

And I got a huge four bar DC power bus designed for 2,000 amps.  If your power bus is just a couple bolts screwed into a piece of plywood I have my doubts you got enough parallel feeds to it to run two of these inverters at Max Dawg.

After we got our new battery bank even a dead short between the "hots" in the main panel won't shut them down - it just trips the mains in the inverters.

What I gathered is that DanB has an old forklift battery that he plays musical cells in when ever a cell goes tits up.  And he's feeding two inverters with one series string of those cells.  I got six series/parallel strings @ 1,310 amps EACH feeding mine.  And the voltage of the system don't make any difference because for a given battery bank every cell in the bank has to deliver the same amps for the same watts, regardless of operating system voltage.  The only difference is the cables from the bus to the inverters, which have to be bigger on 24 volt.  Otherwise my sW Plus 4024's got the same surge capacity as his SW5548's.

If his new cord "fixed it", then it's all fine and will probably fill his needs.  But I got an education on the fact that it takes power to make power if you want to be able to use big inverters to their full capacity.  For us, where we got 240 volt water heating, electric range, electric clothes dryer, and everything my wife wanted, we use them at full capacity on a regular basis.  When I picked 'em out the RE dealer where I got them told me that if we want to do all that there's only one inverter you can buy that's going to do it and stand up to the task day in and day out.  But you can't run 'em on 1,200 amp-hours worth of batteries.
--
Chris

Edit: I would bet you a shiny new nickel that if DanB would take the slave inverter out of the system and just run his welder with one inverter with a step-up transformer, that his problem would go away and one inverter would start the welder just fine.  It really sounds to me like he got these inverters, bolted them to the wall, and forgot to install the rest of the system it takes to run dual inverters on split phase power (batteries, leg balancing transformer, and a four bar bus that splits the output of the bank for each inverter).

[ChrisOlson]

I ask this because we could calculate or ' read ' the Z rating(s) of that or those transformers.

I just thought of another issue that may shed some light.  The 60 amps or so that I measured with the Fluke might be bogus.  These inverters in question have soft start in them and they seem to limit the starting amps of a load to around that 60 amp mark, no matter what you try to start with them.  I think this is by design because the main breaker on the inverter is 60 amp.  They will deliver more amps on short circuit for a few seconds until the breaker trips.

BUT, if I try to use my welder with the standby gen it kicks the breaker out on the gen every single time.  The standy gen has a 40 amp breaker.  If the gen is feeding the inverters so I got both gen and inverter power then it's fine.  But if I got my big transfer switch flipped so the inverters are out of the system, the generator won't power the welder up.  It will start anything else but not the welder.  And 50 amps does not exceed the surge limit of the generator.

So it could very well be that if you have unlimited power to the welder with no soft start, it might draw a huge surge of power when the transformer first powers up.  The gen sort of has "soft start" because it lugs down when a big surge load is applied to it.  But that causes the voltage and freq to drop so it cascades into a short circuit situation faster than the inverters do.

I don't know much about this transformer stuff.  But that's just an observation I've made in the past for the experts who do know how these transformers work.

Edit:  Dug out my inverter manual to see what it says in the specs:
Short Circuit Amps 360
100 millisecond surge 78
11 second surge 60
Continuous @ 25° C 33

Since I still got the Fluke hanging on the wires from last night I tried it again.  We got almost no loads on in the house except for 2 kW load from the water heater.  So I turned that off and the inverters went into search mode.  I wrote this down on a yellow piece of paper every time I tried it.

When I flipped the breaker to the shop the Fluke showed 63.1 as soon as the inverters woke up, then back to 3.2.  I turned it off and flipped it again.  The second time it only showed a peak of 12.5 and then back to 3.3.  If I wait for a few minutes and try it again, then it says 61.9 and back to 3.3

So I flipped the water heater back on.  Now the Fluke showed 8.0 amps.  When I flip the breaker on to the shop it went to 63.7 and back to 11.2 running amps.  A second quick flip of the breaker got 22.4 and back to 11.1.

So it looks to me like if you do it twice in a row, it draws less the second time.  If you wait for a bit it draws more.  I think that soft start BS they got in these inverters is why it won't go over 60 something.  But I don't understand why it draws less power when you flip the breaker fast.
--
Chris

[DanB]

Interesting thread.  My problem was definitely solved with a short extension cord anyhow.  (before that... my inverters and my welder sit about 3' apart from each other and the wire run is very short/stiff - and teh battery is also wired up with short/thick runs, there is not much resistance in my system).

The 'error' I was getting before was over current.  I've never even come close to low voltage (my battery isn't that bad Chris ~ I think both me and my battery are moody and think the worst of each other every now and then but it's been doing fine - I never drop below 48V when welding).

My welder is rated to draw no more than about 19 amps @240V (it's a miller 175).

At any rate, everything is working like a charm now and I'm happy!!!

[ChrisOlson]

DanB - read above what I just did.  This one has aroused my curiosity as to why these big transformers do this.  It's weird because what I'm seeing is there's definitely a "soft start" in there that limits the surge amps.  I don't know how that works but I remember seeing some big SCR's there, and I've seen those in electric motor soft starters.  I wonder if it works the same way as those non-mechanical soft starters for big three phase motors?

So OK - if you're confident you have enough battery power to run those machines at full surge and continuous output, maybe you got a bad soft start board in one inverter?  You could unhook the stacking cable and try each one individually with a transformer on the output to maybe isolate which one has the problem?

Edit:  I tried another test just for the heck of it.

I went out to the shop and turned on the welder, drill press, lathe, 3,600 rpm bench grinder, and all the lights.  Then I went back to the utility room and flipped on the breaker that feeds the shop.  With an already 2 kW water heating load on, plus the 'fridge running in the house, the inverters sounded like somebody hit them with a maul, then they buzzed really loud for about three seconds.  This time the Fluke said 76.6 amps peak!  But they brought everything to life.

So I did yet another test.  I flipped the breaker off, went back out the shop and hooked the ground clamp to the electrode holder and put the welder on 225 amps.  I went back to the utility room and flipped the breaker on.  There was a big "thud" from the inverters, loud buzz, and it tripped the 60 amp mains  The fluke said 99.0 amps peak!  But the inverters themselves still did not error out with an overload.

If there was no breaker on the inverter output, that last test would've for sure caused the overload error.  I don't know if this helps Watt figure out the "Z" or whatever it is.  But regardless, them are some healthy frickin' inverters.  My standby generator won't run all the stuff I had turned on in the first test, much less bring it all to life at the same time   
--
Chris

[joestue]

I went out to the shop and turned on the welder, drill press, lathe, 3,600 rpm bench grinder, and all the lights.  Then I went back to the utility room and flipped on the breaker that feeds the shop.  With an already 2 kW water heating load on, plus the 'fridge running in the house, the inverters sounded like somebody hit them with a maul, then they buzzed really loud for about three seconds. 

3 seconds? yikes....

Let me explain it this way:

When you turn on a 240vac 60hz transformer, somewhere between zero and 1 volt seconds (240vac * 4.15mSec) has to disappear; +/- to any magnetic flux left in the core when it was last turned off.

To see this in action connect another large, (or any size toroidal) transformer to the inverter and use your Fluke meter to watch its amps when you flip on the welder.

If the volt seconds disappear because the inverter let some dc voltage appear at its terminals, you will see the transformer that was previously connected go into saturation when you flip on the welder.

If however the voltage output of the inverter collapses, and the volt second(s) disappear into the resistance of the welder, and any added resistance such as an extension cord, then you will not see any abnormal current drawn by any of the other loads.
Since you've got 100+ feet between the welder and the inverters, you can run that test on both ends, and you'll almost certainly see different results.

other than that, i really can't think of any way to see what is going on without an oscope.

[Watt]

DanB - read above what I just did.  This one has aroused my curiosity as to why these big transformers do this.  It's weird because what I'm seeing is there's definitely a "soft start" in there that limits the surge amps.  I don't know how that works but I remember seeing some big SCR's there, and I've seen those in electric motor soft starters.  I wonder if it works the same way as those non-mechanical soft starters for big three phase motors?

So OK - if you're confident you have enough battery power to run those machines at full surge and continuous output, maybe you got a bad soft start board in one inverter?  You could unhook the stacking cable and try each one individually with a transformer on the output to maybe isolate which one has the problem?

Edit:  I tried another test just for the heck of it.

I went out to the shop and turned on the welder, drill press, lathe, 3,600 rpm bench grinder, and all the lights.  Then I went back to the utility room and flipped on the breaker that feeds the shop.  With an already 2 kW water heating load on, plus the 'fridge running in the house, the inverters sounded like somebody hit them with a maul, then they buzzed really loud for about three seconds.  This time the Fluke said 76.6 amps peak!  But they brought everything to life.

So I did yet another test.  I flipped the breaker off, went back out the shop and hooked the ground clamp to the electrode holder and put the welder on 225 amps.  I went back to the utility room and flipped the breaker on.  There was a big "thud" from the inverters, loud buzz, and it tripped the 60 amp mains  The fluke said 99.0 amps peak!  But the inverters themselves still did not error out with an overload.

If there was no breaker on the inverter output, that last test would've for sure caused the overload error.  I don't know if this helps Watt figure out the "Z" or whatever it is.  But regardless, them are some healthy frickin' inverters.  My standby generator won't run all the stuff I had turned on in the first test, much less bring it all to life at the same time   
--
Chris

Chris, I will play around with the scope and a transformer tonight and show you some snapshots of what I was getting at.  As far as the impedance of your transformer, a bit more would have to be known about material, size and wiring of your transformers in your inverters.  Also, the soft start option would have to be taken into account.  But at any rate, some scope pics with help explain what I " think " I was getting at. 

[ghurd]

So it looks to me like if you do it twice in a row, it draws less the second time.  If you wait for a bit it draws more.  I think that soft start BS they got in these inverters is why it won't go over 60 something.  But I don't understand why it draws less power when you flip the breaker fast.

Because of a long wire, and a big pile of fat caps.
G-

[joestue]

So it looks to me like if you do it twice in a row, it draws less the second time.  If you wait for a bit it draws more.  I think that soft start BS they got in these inverters is why it won't go over 60 something.  But I don't understand why it draws less power when you flip the breaker fast.

Because of a long wire, and a big pile of fat caps.
G-

i'm going to need to consult my 8 ball on that one...

you're rolling the dice every time you flip the breaker.
will the contacts energize between 0 and 90 degrees or between 90 and 180? (or 180 and 270, or 270 and 360)
(the current will also depend on when in the cycle you turned it off)

i had to flip my welder on and off some 30 times before catching a good shot of the peak current spike on camera.