12 volt dc to power vfd controllered 3 phase motor 3

[oceanman]

We have set up a system to use a 12 volt battery to run a 3/4 horsepower 208 volts three phase motor. The system as currently laid out is as follows: A 12 volt lithium ion battery pack is used to supply power to a 2500 watt modified sign wave inverter which provides 120 volt AC power to a VFD drive which then powers and controls the speed of a 208 three phase 3/4 HP AC motor. In the lab, the unit worked well for a while, then the inverter failed. We got a new inverter, and it too failed in a fairly short while. We suspect that it was a bad decision to have used a modified sign wave inverter. Our next proposed step is to switch to a true sign wave interter, but we are not 100% sure that this will make any difference. (We know that we could switch to a DC motor controller and DC motor, but we do not wish to do this, as we are trying to simulate a production system which will be powered by AC line power.) I would appreciate any opinions on this issue, especially if anyone thinks that the true sign wave interter will actually provide the fix we are looking for. I am open to any other ideas too. Thanks.

 

[itsmoked]

oceanman: Ow! That hurts sorry to hear of your fuse like VFDs. I am surprized that that inverter is damaging them. I would expect a junker squarewave inverter would feed a bridge/cap front end without a problem. I would think the rectifiers would even be happier than with a sinewave.

Do you have any idea what is failing? Is it the front end of the VFD? Or the whole thing just goes dark? Or looks fine lights up etc. and the motor just won't move?

You are obviously driving a 3ph front end with single phase. Are you oversizing the inverter to take this into consideration? Have you tried other power input connections to see if you just blew one set of diodes. Another words, hook the power to your dead inverter using every possible connection pattern. If a different pattern works, (even briefly), this will tell you that your setup is abusing the front end of the VFDs.

Remember that using single phase causes 100% of the power to come in thru %66 of the front end. This is why I keep looking at the front end.

 

[BigLe]

how did you configure the battery pack? did you load the motor while it failed? did you measure the battery voltage before, during and after your inverter failed? and by 'fail', is there a failure code given by the inverter?

 

[oceanman]

Hi BigLe,

Thanks for the comments. I can try to respond to your questions, as follows:

Battery Pack:

It was a lithium ion battery pack, three cells connected in series, for a nominal voltage of 12.6 volts. It was 100 amp hours.

Failure Mode

The motor ran OK the first time used. Upon shut down and restart of the inverter, the output voltage went down to 55 volts. It should have been 110 or 120 volts. The battery voltage was measured in all cases and it was still in the vicinity of 12.6 volts, so the battery pack is still good.

The inverter indicates an overload occurred.

I hope this helps, and I would appreciate your comments as to what happened.

Mark

 

[itsmoked]

Oh wow, I read that one wrong.... Sorry I thought your VFD choked!

Oceanman: Your motor is going to want to draw about 3500W on start up. Usually on inverters they rate them something like 1500WATTS!!! (1200W continuously in fine print) A generator would be rated the same way, but unlike an inverter if you overload a generator you may get a lower voltage during the overload, but an inverter will either turn off, blow a fuze or burn out a semiconductor and very quickly too! What i'm trying to say is an electronic inverter as compared to a generator would be damaged very quickly if its maximum rating is exceeded. Read that rating as "if exceeded will be damaged".

You either need a substantually larger inverter for that size motor, to take the starting load, OR you need to configure your VFD to start the motor slowly and ramp up thereby limiting the starting current.

Can your demo system be started slowly? By slowly I mean can the motor go from zero to running speed over 2-4 seconds? That would probably limit the current required to the 20Amps your inverter can provide.

I am having one other problem with your set up. You state that you are running a 208V motor with a VFD that you are only supplying with 120V power. I didn't realize VFD existed that also stepped up voltage!

I would think you need to supply the VFD with 208V. Maybe I'm missing somthing?[else ] You may be able to run a transformer with your inverter to get you to 208V, but it would have to be a sinewave inverter. It might be better to get an inverter rated at 220V output.

 

[oceanman]

Thanks, Itsmoked. I appreciate the comment about the motor start up. We had the VFD set for a "soft start" to elminate this problem. I will try a test, plugging the VFD into a wall outlet power source to test start up load using various motor start parameters. We are using an E500 series VFD from Mitsubishi, which can take in 120 volts and put out 208 3 phase.

Oceanman

 

[itsmoked]

oceanman: I see it's the "W" option. That's interesting. They have included even the kitchen sink in that family now. I used several of those for down hole 3ph well pumps, running off of 1ph 240V and thru 400 ft of extension cord!! They ran well..er okay!

If your inverter now only puts out 55V unloaded, you have cooked one half/side of the output drivers. This is from excess current draw. It may be repairable by replacing the output FETs. The only fix will be to keep the current draw down to the 2500W average probably no more than about 20 Amps. If you have a scope you could monitor the peak current drawn while running from the wall. Use a low value shunt resistor and carefully watch your scope ground. Or use an isolated current probe. Then you can screw with the VFD settings until you stay below that threshhold.

Are you running a pump that is pushing water?

 

[BigLe]

Did you start the motor with load? Can you try to start the motor with the same drive configuration but don't load it yet, if there is still an overload error, would you double check the winding of the motors and its grounding? also for the drive, i assume it's a v/hz type one, would you double check the setup of control parameters? If you have a DC source instead of battery, you probabaly want to use that to avoid damage to the battery pack.

 

[ScottyUK]

Capacitive loads are hard work for any inverter. For a 'modified sine' - equally accurately it is a modified squarewave! - it is harder because the gently rising sine spreads the charging pulse for the DC link capacitors over a couple of milliseconds. The modified sine has a flat top voltage, so the capacitors try to charge almost instantaneously when the voltage reaches peak value. This will probably be the overcurrent which kills your inverter. A sine output inverter may be a little better, but beware of the types that use a resonant circuit in the output to generate the sine. They do not normally behave too well on non-linear loads.

Possible alternative:

Use a DC/DC converter 12V-170V and directly feed the DC link of the VFD. You will still have to contend with the sizing issue identified by itsmoked, but the pure DC supply to the drive should make things a little easier because the converter feeds the DC link continuously, not in a couple of pulses every 20ms. 170V output may be tricky to find, but 160V is pretty standard and most units have an output voltage trim.

Possible alternative:

If this is a test, it may be economical to rent (or even buy) a small diesel generator to provide power. It will deal with the motor inrush much better than an electronic system.


Good luck.


----------------------------------

If we learn from our mistakes,
I'm getting a great education!

 

[BigLe]

good point.. just noticed it was 'sign wave'..

 

[OperaHouse]

Maybe it's how you hooked it up! You stated it worked in the lab but not on the site. I suspect in the bench test there was one thing you didn't do. To comply with "codes", did you bond the "neutral" to the battery earth? A great majority of these small inverters must have the 110VAC output totally floating otherise you get exactly the result you got.

 

[itsmoked]

Yo Scotty, he can run that sweet VFD right outta a wall socket.

Oceanman: Scotty makes a good point.. You are essentially going to a large electrical headache to make some kinda AC when the VFD would be happy with DC.

http://zahninc.com/su1.html

Another thought that may sound crazy but I would examine is batteries.

(142) AA Li-Ion. Several cars have been powered by lots of cheap(cuz they're mass produced)AA batteries.

Just a thought.

 

[oceanman]

Dear All,

Thanks for all the comments. We went ahead and ordered the pure sign wave inverter, and should have it in fairly shortly. I will try to provide more information to your additional questions on Monday concerning some specific questions. Yes, we are trying to pump water, but the failures occured under essentially low to no load on the water pump. Yes, I can plug the VFD into a wall outlet for testing, so I do not need the generator. I like the idea of feeding in DC. Itsmoked, are you proposing lots of AA batteries in series to generate the (295 volts?) needed to feed the inverter? Or, did you have something else in mind for the AA's? I am not 100 percent certain of how you propose to use them. Thanks.

Oceanman

 

[itsmoked]

Yeah ocean I was thinking about Scotty's 170V comment. But if I look at the E500 Data sheet it is a bit confusing. They note NOTE 7 That you need about 285V as related to the 200V input inverter. But with respect to your 100V input inverter they say you can feed it DC but don't actually say what voltage...would it be 185V? I would guess this but maybe you can't?

Yes I was thinking of a bunch of batteries in series. AA's would provide enough power for probably 1HR of use. I don't know what your physical layout requires so they might or might not be a solution for you. If you can take your power source to the demo sites on a little handtruck like cart it might work. They might also be too much trouble to charge, (strange voltage). Generally if you buy a bunch like 300 you get them at wholesale with a whopping big discount. That's why the car guys tried them,(350v).

 

[oceanman]

I thought that I would provide an update, which may put my issue to rest We installed the true sign wave inverter and it did not fail. We ran the motor for a few hours off of the battery pack without problems.

Itsmoked, do you have a web or other reference for the 350 volt car project you are referring to?

Thanks to all

Oceanman

 

[Warpspeed]

The problem to me appears to be the compatibility between the inverter and the rectifier system in the VFD.

The VFD probably uses a voltage doubler to generate a dc link voltage of around just over 300 volts. This is probably two large diodes and two large electrolytic capacitors.

The problem with this is that the poor inverter sees very high current spikes only right at the very crest of the voltage waveform to charge those capacitors. It will most likely be none too happy about that.

Probably the best way would be to use a square wave inverter with a 170 volt peak waveform, as has already been suggested. But the current spikes will still be huge. Going to a true sine wave inverter is probably going to make the problem even worse.

Your best bet may be to monitor the ac current waveform on an oscilloscope, and add some series inductance between the inverter and VFD. That will spread out the conduction angle of the rectifiers a bit and reduce the peak current.

But the choke will need to be rated not to saturate at the peak current, which will probably be several tens of amps.

The very best solution would be a properly designed high frequency dc to dc inverter to supply the proper voltage regulated dc link voltage to the VFD, and forget about intermediate 60 Hz power altogether.

 

[oceanman]

Warpspeed,

So far, so good with the true sign wave inverter. It has not failed yet. I appreciate your other comments. Do you know who makes a compact dc to dc converter that will do as you suggest. Most of the ones I have seen are either the wrong voltage, or are really large. For now, I will keep my fingers crossed that the true sign wave inverter will not fail.

Regards,

Oceanman

 

[Warpspeed]

Glad it is working out o/k, but I would still look at the current waveform on an oscilloscope.

I cannot really suggest anything suitable in the way of a dc to dc inverter. I used to design switch-mode power supplies professionally many years ago, and I would just build my own. But that is of no help to you unfortunately.

 

[itsmoked]

Glad to hear its working.. Thanks for the update. What model batteries are you using? They sound impressive.

I would've guessed the sinewave inverter would be even more unhappy.. But I suspect Warpspeed's voltage doubler assessment is right on.

The batteries in the car....ugh. I hoped you wouldn't ask.
As I looked for the site before mentioning it. I think that particular car/site was bought by a related company selling similiar things and they wanted that information...um supressed.

This is the car type they did it in.

http://www.acpropulsion.com/PDF files/tzero_brochure.pdf

Note this thing has a 200HP electric motor.

Here's a quicktime movie of it blowing off Ferrari.

http://www.acpropulsion.com/tzero_pages/Ferrari_race/tzero_ferrari.mov

This is the company that seems to own the rights now.

http://www.acpropulsion.com