Saturday, March 9, 2013

Summary and Conclusion: Compatibility Issue Between MS2000 inverter and Splendide 2100XC Washer/Dryer

OK, This has been an interesting road to travel down.  One I had totally not expected to have to travel, but I think I am at the end.  If you have stumbled into some of the other posts, or want to see more details, just look back over the last week or so.  But, with this post I am bringing things to a conclusion.

Short and Sweet:

  • The Splendide XC2100 Washer/Dryer and the Magnum MS2000/MS2012 inverters do not play well together.
  • Other small washer/dryers may also have an issue.
  • Problem is a bit of both side's fault - with the cause a well know issue in AC Power systems.
  • Problem can be corrected with addition of 50uF motor Run Cap 

Skip to 'The Solution' below for the quick answer, read on for the eye-glazing details.



==================================================================



Here is What I Know (Have seen with my own eyes):

  • We have an interaction between the Splendide XC2100 Washer/Dryer and the Magnum MS2000 inverter.  This manifests itself during the Wash portion of the cycle, and is indicated by a jerky motion on the drum, the LEDs flashing on and off, and the door-lock solenoid clicking.
  • This interaction can be easily reproduced, though seems to be somewhat dependent on the load being washed. 
  • We did not see this at any time over 2 years of operation using a Xantrex ProSine 2000 inverter.




Here is What I Have Been Told (And the sources seem to me as credible):

  • This interaction has been seen before with the Magnum Inverter and Splendide washer/dryers.
  • The interaction has been seen before with the Magnum Inverter and other brands of small washers (e.g. Bosch)
  • This issue is NOT seen using some other brands of inverters, specifically the before mentioned Xantrex Prosine 2000, as well as the GoPower line of Chinese stand-alone inverters.
  • Three T&T mailing list members have indicated no issues using the Victron line of Inverters
  • Three T&T mailing list members have indicated problems with small washer/dryers and MS2000/MS2012 inverters.  Two were solved via the Cap solution below, one did not - his was solved by changing out to a MS2812 inverter.
  • Four T&T mailing list members have indicated no issues using the Magnum inverter, all of them were using an MS inverter (true Sine-wave) other then the MS2000 or MS2012. (Ala, a larger Magnum inverter).



Here is What I Suspect (Likely true, but do not have any complete supportive information)

  • Many brands and models of these small washer/dryers have issues with the Magnum inverter as well as some other brands and models of Inverters, and even some small AC Generators.
  • We have powered the Splendide using our Honda 2000 generator with no issues, but only a handful of times.  (Next time I have the Honda gassed up I will put the scope to it and see what things look like)


Here is What I Conclude (Based on the above, plus other research and data collection)

  • The Splendide XC2100 has little to no (likely no) PF (Power Factor) correction in its design.  Likely other small size washer/dryers and stand alone washers have the same deficiency.
  • The Magnum MS2000 / MS2012 inverters are unable to properly support poor PF loads.





Background:   Real Power vs. Reactive Power.  In the Nice World (being DC world) all is simple.  Amps are Amps, Volts are Volts and Watts are Volts * Amps.  If something needs to get done, apply some Volts and send some Amps and energy (in the form of Watts) is delivered.  In the Complex World (AC World - and yes I am totally ignoring all the issues a DC system can have) Amps are Amps, and Volts are Volts, but Watts are not always Watts...  See, it all comes back to High School Math (why it is all the gnarly issues always come back to High School Math???)  In the AC world we need to not only look at Amps and Volts, but their phase relationship   Here is an example of a well behaved AC system, where the Volts (top trace) and the Amps (bottom trace) are all in step and symmetrical:


Nice well behaved.
Voltage and Current traces are in step and very Sine looking.
The above is an example of energy being delivered in a way referred to as Real Power by the Power Company folks.  In this example a space heater, a pure resistive load,  is being powered.  Now, look at this example:


Not so nice. . .
Notice wild lower (current) trace
The Voltage still looks like a Sine wave (as we want), but the Current trace (the lower trace) is all over the map.  What is going on here?  This is an example of a complex load being driven -- in this case the Splendide washer/dryer.  And it consists of two major components:
  • The spike aligned with the top of the Voltage Sinewave curve is from the internal 'electronics' power supply.
  • The wild squiggles on the left are from the variable speed induction motor driving the washer drum.

Of specific interest here is the wild squiggles.  Not that they are squiggles (though that is interesting on its own), but that they are not lined up with the peak of the Voltage waveform.  The squiggle current draw occurs before the peak of the voltage.  In fact, this is a prime example of an Inductive Load, it occurs 90o out of phase with the Voltage and is what is know as Reactive Power. It is the direct result of an inductive load being powered (the drum motor in this case), is well know in the Power Company world, and is a major headache for them:  http://en.wikipedia.org/wiki/AC_power

This out-of-phase current demand is summarized in a value called Power Factor.  If everything is aligned, as in the 1st photo, the PF is known as 1.  Anything less than 1 actually causes inefficiency and waste.  In fact, the ENERGY STAR specifications often call for corrected Power Factors and require the load waveforms to look more like that shown in the 1st picture.  But it costs money to add in PF correction.  And when in a house with the MASSIVE AC power distribution (think Bonneville Dam and such) system supporting you, poor PF is really never noticed.

So what?  Well, in my small world this became a major headache  - because when this less than ideal PF load is being served by my relatively small inverter, the inverter needs to try and keep up with things.  And it can not depend on 100,000's of tons of concrete and steel to do so.  In my case, when driving this complex load, the Magnum inverter had a hard time keeping up and this is what I got:

Magnum unable to supply current demands of the Splendide

See how the Voltage waveform now has a kind of hitch in it as it raises (top trace, left hand side)?  And notice how the current being delivered no longer had the wild squiggles on the left.   What is happening here is the Magnum inverter is not able to keep up with the demands of the Splendide washer/dryer.  It can not supply that large out-of-phase inductive current draw we noted in the 2nd photo, and it can not keep the voltage nice and clean.  It is struggling here - and as a result the washer/dryer starts misbehaving  because it is not getting the reactive power it needs, the motor movement becomes jerky, the lights start flashing, things start clicking.  It is a mess...




And a side note:  This is also why some people complain their battery charger only delivers 60A while running from the Generator, but can do 100A when plugged into shore power.  PF and the ability of the generator to support those less than ideal loads.  (Should point out some inverter/chargers do have PF correcting in them - including the Magnum unit!  But that is only when it is running as a charger. . . .)






Where is the 'Fault'?  I would say both.   The Splendide is sold into the Mobile market, and as such should take better care to design in some level of PF correction.  Being a 'Poor Neighbor'  when plugged into shore power is not noticeable due to the massive capabilities of the land-based AC power distribution system.  But being that same poor neighbor in our small isolated world is -- well -- just being a poor neighbor.  The Magnum line of inverters are from designers with a long history and experience in the field.  Given it is clear that it is possible to design inverters with the capability to better support poor PF loads - witness the Xantrex ProSine 2000, the inverter inside the Honda 2000, generator, reportably Victron inverters, and even the Chinese inverters, I would have expected the Magnum to be better able to handle this complex load.   This is perhaps my biggest surprise and, to be honest, disappointment.



What about OutBack inverters?   Well, I have no idea.  I had only one chance to look at an Outback on another boat, but just did not put the effort into figuring out how to move our boats near each other, plug in the Splendide into his boat and take measurements   And no one from the T&T world has commented on the combination   Also, the Splendide support folks do not seem to run across the Outback inverters often. (they mostly see Magnum and Xantrex, and mostly in high end Motor Homes!)



Resolution: Addressing and correcting inductive Reactive Power is a well know issue in AC Power Distribution systems.  As almost exclusively reactive loads tend to show inductive properties the addition of capacitance to counteract is very common (Capacitance and Inductance are on the opposite ends of current off-setting).  In fact, if one looks at AC Power distribution sub-stations you will see a wide array of Corrective Capacitors - they are typically semi-trailer sized units.  And power companies charge factories extra if they themselves do not take steps to clean up their PF , typically through the addition of capacitors.   But a key here is how to get exactly the right balance, as too much (over-correction) can bring its own issues.  Lots of $$$ go into that problem, and solutions range from switched banks of capacitors to using other approaches as well.  In the Big Boy space, this is a Big Boy topic.

For my issue on Viking Star, research has pointed to the addition of a 50uF capacitor to help address the Inductive part of the Splendide load. And here is what we get:

50uF Capacitor to the rescue!

Notice how the voltage waveform has lost is kink, and looks more like a nice Sine wave.  And notice also we are now able to supply the inductive current (that peak in the lower trace on the far left side of the screen) as needed by the washer motor.  What is happening here is the Capacitor is saving up energy and releasing it when needed by the motor.  Thereby relieving the Magnum of that responsibility.  BTW: the kind of flat-top on the voltage waveform is another issue, for another day :-)


But a word of caution:  because this problem only  presents itself while running the washer/dryer, and there is always a risk of over-correction,  I have placed the capacitor and the Splendide on an on/off switch.  So the washer and the cap are only 'connected' when needed.   And there is a  side benefit!  We had noticed the power saver sampling of our Xantrex inverter would wake up the Splendide - causing it to flash its LEDs.  It did this every 30 seconds or so.  Given our unit is located in the aft stateroom this is something we just grew to accept while trying to go to sleep.  But with this new ON/Off switch we have no blinking LEDs!

I have also added a slow-blow fuse to the 50uF cap.  Here I got a bit of controversy from some knowledgeable folks, some saying it is not really needed.  But in my independent research, over-correction is a real issue, and it manifests itself in extreme cases by the system entering a resonance state.  Given the Magnum is based in transformer technology (as opposed to pure PWM switching ala the Xantrex's ProSine) there IS a LC present between the 50uF cap and the Magnum inverter and hence a resonance state is possible.     I was cautioned by one reviewer that some of my wave form captures  had an early indication of such.  A system that enters such a state will show large currents going back and forth to the Cap.  Very large currents, in the real world that can cause issues like overheating of wires, tripping of breakers, etc.   So, I do suggest the addition of a fuse.  It should blow if a major resonance state presents itself.




==================================================================




The Solution:  Ok, enough.  Here is what I have for the solution.  Adding a 50uF Motor Run to counteract the inductive load from the Splendide.  Caution:  Do make sure to use a RUN cap, not a Start Cap as the Start ones are designed to be used only for a few seconds.

Solution to Magnum / Splendide incompatibility

Do note the part numbers shown are just reference -- any brand of Motor Run cap can be used, as well as fuse holders, etc.  Just make sure they are rated for AC voltage (370+ on the cap, and 240+ on the fuse and fuse holder - note those Blade fuses are only rated to 32v!).

An enhancement would be to add a small 120v panel light (e.g. Radio Shack 272-710) paralleled with the Fuse to indicate if the system had at some time entered resonance blowing the fuse.  Or if one notices the Splendide acting up again, check the fuse.  If this keeps up, try using a smaller value of Run Capacitor.  Maybe a 25uF.  (In my trials, 7uF did not fully address the problem 15uF seemed OK.  50uF cleaned it up, and is the value recommended by Magnum - so that is why I went with it).



Thank yous:  Nothing like this happens in a vacuum.  I want to thank T&T members for their input - online and off.  And specifically Mark W. for his expertise in these areas as well as sending me test equipment and a range of motor run caps.  It is specifically through my mailings with him that I have a good level of confidence in the assessment and the solution.

Both the Magnum and Splendide tech support units have been responsive and helpful.  When our Xantrex failed I specifically looked for another company based in their reputation for after sales support, and this experience has shown I made a good choice in Magnum.  And I spent over an hour talking with the Splendide support head - getting not only insight into this issue but hints for cleaning the XC2100 as well as how to reduce wrinkles   Both of these companies could have simply said 'Not Our Problem' (there was, to be honest,  some of that in their responses), but neither did.  They reviewed my data and gave insightful comments and in the case of Magnum- engineering input and review of the solution.

There you go.  More than you EVER wanted to know, but then - are not Boats just full of Learning Opportunities!




1 comment:

  1. Wow, how have I missed this all. TWO SPAMMer Posts, so I took the opportunity to retain their kind words, but edit out the embedded URLs:


    Trent Palmer - June 9, 2013 at 9:02 PM
    Hi,
    the blog about "Summary and Conclusion: Compatibility Issue Between MS2000 inverter and Splendide 2100XC Washer/Dryer " is very detailed, realistic and superb analysis, of this issue, very nice write up, Thanks.




    suki lee - August 19, 2013 at 2:48 AM
    Nice Post!! pretty informative..thanks for providing such a nice post.

    ReplyDelete