Questions on Xantrex SW4024 inverter?

[gr8njt]

I'm about to buy a new Xantrex SW4024 inverter charger.
I have read the "specs" and it will fit my projected use perfectly, however,
they're not as meaningful as actual field bumps and bruises.
Anybody with personal/actual experience in reference to:
a. Reliability of this unit. How often did you have problems if any?
b. Actual experiences on it's install/use particularly the SW4024 model.
c. Things you did not like with the SW4024 and why? Are they tolerable?
d. If you had to do it all over again, would you buy the SW4024 again?
....if NOT, what unit would you buy and why?
THANKS.

[H3Jim]

I've had mine installed now for almost 2 years.  nary a problem.  I have not had a lot hooked up to it, especially in the beginning.  It does everything it's supposed to do.  It might be higher capacity than I need, and I've paid for that excess capacity, but on the bright side, I know I can run anything I do need.

The matching capability is very nice.  I have a 150 extension cord running down to where my bus is parked, and I'm limited to about 7 amps through the cord before I start getting significant voltage drop. Whenever I run tools or anything else the inverter automatically makes up for the difference from the batteries, then recharges again when I stop.  This is true whenever visiting "amp challenged" relatives or an old  RV park with limited electrical capability.

I have had no issues powering anything AC from the unit, computers, tv, stereo, air conditioners etc.

So bottom line, it seemed like a lot of money for it, but it does it all well and I've never had to think about it since I installed it.  I have the remote panel and I like that a lot, although I'm in the process of adding solar and I expect I will be checking the solar controller now more than the inverter controller.  The solar converter has a shunt and a total amps used and total amps input input and available capability.

[gumpy]

a. Reliability of this unit. How often did you have problems if any?

No problems.  There are some quirks you need to be aware of with this unit.
1. It does not do neutral-ground bonding (please God, don't let this simple statement turn into a full blown discussion about neutral-ground bonding on this board!!)
2. It will sometimes cause GFCIs to trip. Cause unknown. Solution unknown. Sporadic. Sometimes it does, sometimes it doesn't.
Hasn't really been a problem, just a casual annoyance.
3. It is sensitive to frequency and voltage (both high and low). Difficult to use a cheapo contractor generator with it. Not impossible, but takes some finesse.

Very satisfied with it. The campground in Billings MT we were in took a direct lightning strike, which blew out 2 of the 3 transformers, ours included. My inverter quit and displayed an error indicator. Turned out to be an overcurrent. It tried to pick up the entire load I had on at the time, and it was too much for it so it shut down.

b. Actual experiences on it's install/use particularly the SW4024 model.

Very easy to install. However, you must account for the neutral-ground bonding (please, just let it go!!)

Takes a bit of practice to navigate the menu system. Takes even more to understand all the operating modes and capabilities and settings, but for the most part, the factory settings are pretty much right on.

c. Things you did not like with the SW4024 and why? Are they tolerable?

Would prefer if it had selectable priority on generator vs. shore cord. The standard SW4024 is priority shore cord. The MC version is priority generator.  Standard version offers more options (selling solar power, etc) than the MC version.

User interface is somewhat difficult to navigate. The control panel was not designed for ease of use but you get used to it.

The input current limit setting is really useful when you're plugging into a lower rated source. It allows you to set a maximum current limit that can be drawn off the power source, and will not exceed it. It will augment it, though, if necessary. That's also very nice for heavy short term loads.

d. If you had to do it all over again, would you buy the SW4024 again?
....if NOT, what unit would you buy and why?

Yes. Well worth the money.

I think I'd look at the new company that is producing a 4000 watt inverter with a similar name (4024). Can't recall the name of the new company, but it's supposed to be made up of ex-Xantrex engineers. I'm not sure their product is available, or equal in capability to the Trace sw4024. The Outback units are interesting, too, but I don't think they provide as much capability as the SW4024, yet.

[gr8njt]

H3Jim and Gumpy,

Both your experiences are priceless!!    Thank you.
I had compared the "specs" of the SW4024 with other brands of Pure Sine wave inverter/chargers and though it is more expensive than the next priced competition, I think it may be worth the money for all it's "bells and whistles".
Your experiences are mostly positive that it boosts my confidence to purchase one.

Anyway, Gumpy, you mentioned "neutral-ground bonding" problems. Do you think this is causing your GFCI to trip?
I never knew there was a full blow discussion about it so if you are kind enough to PM me your solution to the problem, (for some backstage tutoring) I'd greatly appreciate it.

[H3Jim]

Gumpy, I knew he couldn't do it! ( not ask about it), your plea's not withstanding...

And I admit I was a little irritated that Xantrex did not address neutral / ground bonding when they built / designed it.

See this thread for some discussion on the issue.  There were others - you might try a search for "neutral ground bonding"  Important to know and do, yet fairly arcane.

http://www.busconversions.com/bbs/index.php?topic=593.msg4609#msg4609

[gr8njt]

Gumpy, I knew he couldn't do it! ( not ask about it), your plea's not withstanding...

And I admit I was a little irritated that Xantrex did not address neutral / ground bonding when they built / designed it.

See this thread for some discussion on the issue.  There were others - you might try a search for "neutral ground bonding"  Important to know and do, yet fairly arcane.

http://www.busconversions.com/bbs/index.php?topic=593.msg4609#msg4609

LOL!!! Yep, I couldn't help but ask. It's like dangling a cawwot in fwont of a wabbit.
And thanks for the link.

[gumpy]

The reason Trace didn't address this when they first made the unit is because it was designed for home use, where the bonding is done within the distribution panel. Why they didn't correct it when they made the MC version, I can't say.

The SW4024 manual offers some insight on the problem, and their suggested solution.

I am currently in the process of updating the electrical panel page on my website. It will include a CAD drawing of my particular
solution. It's gonna take some time to get it up because it is going into more depth then I typically would. Suffice it to say, there are easier solutions than mine, and probably more elegant ones.

In the meantime, I suggest you do a search on this board and the other one and get familiar with the issue, and the controversy I alluded to.

And that's all I have to say about that.

[Sean]

OK, I'll add my voice here.  All the comments thus far have been good ones, and I agree.  Probably, though, I abuse my 4024 more than anyone.

Before I go further, a shout out to Craig:  I solved the GFCI tripping problem!  I know I promised to share when I did so, but I just got it working last week, and have not had time to write it up.  Details at the end of this post.

a. Reliability of this unit. How often did you have problems if any?

Extremely reliable.  We've never had a problem with the unit in three years of use, most of that full time, and much of that time in the boondocks.

b. Actual experiences on it's install/use particularly the SW4024 model.

Installation is tricky.  As others have already noted, the design history of this product, being from the fixed-installation, grid-tie, renewable-energy market, means it has some design deficiencies for mobile use.  As already mentioned, the two big ones are lack of a multi-pole transfer switch, and lack of an internal ground-neutral bonding relay.  This means that you will need to engage in unnatural acts to get the thing wired up where it is safe, compliant with code, and still providing the features you thought you were getting from the spec sheet.

Back when I bought mine and designed my system, I did several write-ups:

• My AC power system: http://www.ourodyssey.us/bus-e-ac.html
• The transfer switch we had to build to use all the 4024 features: http://www.ourodyssey.us/bus-e-ats.html
• Differences between regular and MC2 models: http://www.ourodyssey.us/bus-e-trace.html

It's arcane stuff, but you might get something out of it.

c. Things you did not like with the SW4024 and why? Are they tolerable?

• The user interface is ancient, non-intuitive, difficult, and generally a pain in the a--.  I've been using this thing for 30 months and I still trip over the menus, and forget to change certain settings.  80's-era technology.
• It generates more heat than I'd like, and more than I planned for, especially when charging.  I had to install an extra flushing fan in the compartment, on a thermostat.
• The fans are noisy, again particularly when charging.
• That darn bonding relay issue, and all the problems it created.
• It trips GFCI outlets to which it is connected (solved, see below).

Yes, all these things are tolerable.

d. If you had to do it all over again, would you buy the SW4024 again?

Absolutely.  There is nothing better on the market, although Outback (the ex-Trace guys discussed above) makes a comparable unit with a more modern user interface, at a higher price point.

Now, about the GFCI nuisance tripping:

I got a tip almost a year ago about how to solve this, but I was not able to test it until recently (well, OK, I was too lazy to take off all the interior trim I needed to remove to get to the darn thing, until I also needed to replace the aforementioned flushing fan).  As I always expected, it is the enormous internal capacitance of the input filters that cause the GFCI trips as soon as your cord-connected SW4024 is plugged in to a protected shore power outlet.  (There is also the matter of the much-maligned bond relay, but that's a separate issue -- taking the bond out altogether did not previously solve the trips.)

The simple solution to this, per Xantrex engineering, is to open the AC terminal access cover, and unplug the green ground jumper from the case ground to the left (rear) of the terminal compartment.  Later model units have a male/female spade-terminal arrangement to facilitate this, although older units may have a ring terminal on a post.  This disconnects the low side of the capacitors from the case (and thus input) ground.  The downside, per Xantrex, is that the unit will be "noisier" electrically.  I have tested this, and, sure enough, it works.  Presumably, Marathon and Vantare and other major volume buyers of SW4024-MC2 units have been privy to this for some time, and have just been quietly removing this jumper.

What I did on my unit is to install an SPST switch just outside the unit, and run the jumper out to the switch and back to the case ground.  This way, I can leave it connected for a less noisy unit 99% of the time, and disconnect it in those rare instances where the only power outlet I can find is a GFCI.

Now, of course, merely disconnecting this will not solve your whole problem if you have correctly implemented the ground-neutral bonding relay that we are all avoiding discussing 

That's because, in most cases, the relay "lifts" the bond only after shore power is connected, using the shore power itself to lift the relay.  The bond exists at the initial instant of connection to shore, and the GFCI will usually trip before enough magnetic flux can build in the bonding relay to lift it.  For now, I have been slipping a dollar bill under the contacts of my bond relay just until I'm fully connected; then shore power is holding the relay open and I get my dollar back.

I'm probably not going to re-engineer my transfer switch to fix this; instead, I will make a short adapter cable where the plug's ground tang is a couple MM shorter than the hot and neutral tangs, so that power will apply to the relay before shore ground makes connection.  (This solution is not to code and I can not recommend it.)

The only other solution I can conceive of to this conundrum is to have a completely solid-state device to deal with the bonding, that acts more quickly than the mandatory trip time of a GFCI device.

HTH,

-Sean

[pvcces]

Sean, the power to operate the transfer relay could come almost entirely from the DC side if a solid state device was used to monitor incoming AC, couldn't it? It seems to me that the monitor could draw less than 1 ma, and close the circuit between the DC side and the transfer relay coil, so that the amount of current the relay draws would not matter.

At least, this is what I plan on trying if I ever am confronted with the problem. I would like to get one of the power supplementing models; I just don't need one and have been reluctant to spend the money on one before I need it.

If you know some reason that this couldn't be made to work, I'm all ears.

Thanks.

Tom Caffrey

[Sean]

Tom,

That's right.  A sensitive solid-state device could monitor the shore input, and detect the presence of the shore cord without placing enough load on it to trip a GFI.  The sensor device could then signal the bonding relay to lift, and then a contactor to close, connecting the shore power to the rest of the system.  In fact, I believe this is exactly how some of the bonding systems work in the newer inverter-chargers made specifically for the RV market, as I know they have been designed to get around this problem.

-Sean

[gr8njt]

Back when I bought mine and designed my system, I did several write-ups:

• My AC power system: http://www.ourodyssey.us/bus-e-ac.html
• The transfer switch we had to build to use all the 4024 features: http://www.ourodyssey.us/bus-e-ats.html
• Differences between regular and MC2 models: http://www.ourodyssey.us/bus-e-trace.html

HTH,
-Sean

Sean,
The links to your write-ups are extremely informative and your answers to my questions are greatly appreciated.

[gumpy]

Sean,

Thanks for chiming in on this. I was hoping you would.

This is really interesting. I've always thought the problem was related to charging of some large filter capacitors within the invertor, but don't know enough about that kind of stuff to prove it. When I mentioned it on the board and to Trace techs, I was pretty much told it couldn't possibly be caused by that so I just assumed they were right. Hmmm.

When it warms up, I'll revisit this thread and take a look in my inverter to understand what you are talking about. Thought I had some photos of that area, but nothing with detail.

I haven't had a problem tripping GFCIs for quite awhile. In fact, I think I've only tripped two individual GFCIs since I started using it, and I don't use those particular outlets any longer. I don't recall having any problems on the entire trip to AK.

At the rally, there was a suggestion that it might be caused by rocking the plug as it's inserted, and a possible solution might be to plug the cord into the power outlet first and then plug the Twist-Lok plug into the side of the coach because it's less likely to mate one contact before the others than the pronged plug. Made some sense at the time I heard it.

[pvcces]

Craig, I believe those plugs are all designed to contact the safety ground pin before they contact any of the others.

Tom Caffrey

[gumpy]

Craig, I believe those plugs are all designed to contact the safety ground pin before they contact any of the others.

Tom Caffrey

I think you are correct. The safety ground is on the side of the plug and it makes contact before any of the internal pins.

I'm not convinced this is a solution to the problem, but there was a good explaination for plugging in the pole first (of course I can't
really remember what it was, but it made sense at the time  )