Battery switch deconstruction

[Sean]

A couple weeks ago I posted here that one of our main battery switches had failed, and pleaded for help in finding a new one.  One of our members, Bob (Taibob) had a couple of take-outs and offered to send me one, gratis (thanks, Bob!).

The switch he sent me is at least as old as the one that failed, and I noticed some issues with it when it arrived.  Rather than just stick it in, I thought I'd open it up and check it out first.  It had some significant hidden problems, and it occurs to me that many of us have these switches, and if they've been around the block a few times as this one was, they could be harboring the same issues.  I thought folks might be interested in having a look at the innards and reading about the failure modes of this particular type of switch.

My write-up, complete with photos, is here:
http://ourodyssey.blogspot.com/2011/07/we-hold-these-truths-to-be-self-evident.html#switch

Frankly, for a switch that costs, new, in the neighborhood of $200, these design issues should just not exist.  The wiper should be held down with spring tension throughout its range of motion, not just in the "on" position.  The case should be sealed against dirt and moisture intrusion.  The "off" detent should be more positive and internal to the case.  And, most importantly, the wiper should fully seat on the contacts well before the operating lever travel is exhausted, to account for wear over the years.

I am eager to open up my own failed switch to see exactly what the root cause of that problem was.  I am guessing that it, too, was wear-related.

Unfortunately, there is no convenient way to test one of these switches out-of-circuit, and with the riveted construction, they can not easily be inspected.  So it's hard for me to tell you what to look for on your own switches.  One indicator of internal problems is excessive play in the handle, noticeable only in the "off" position.  Another would be a switch that moves too easily to the "on" position; the spring clamps should provide lots of resistance.  And the handle is easily removed by undoing the center bolt, so you can inspect the key/slot arrangement for excessive slop -- if you see more than 3-5° of play in the handle without moving the key, the wiper is surely not making full contact.

A quick temporary fix for this last issue is to flip the handle over, defeating the stop, and rotate it further to the "on" position.  The wiper will stop traveling when it is fully seated in the clamps.  You can then flip the handle back and re-secure it so it is ready for the next "off" operation.  Do not invert the handle when operating to the "off" position, as the stops are essential in this direction to prevent wiper overtravel.

Hope this helps someone, and I will post any additional findings when I get my failed switch open at some point.

-Sean
http://OurOdyssey.BlogSpot.com

[David Anderson]

Sean,

I have the same switch and it failed 2 years ago.  I wired my cables together on one lug so now I still don't have a switch.  I agree. It always felt "soft" to me when I threw the handle.  My though always was, "that sure seems like a wimpy contact".  Well, it finally failed and was a bear to figure out.  It would show voltage on my meter, but it couldn't carry any amperage load to run the inverter.  What tipped me off was that when I engaged my house to bus alternator solenoid the inverter would carry loads.  When I disengaged the solenoid, the inverter would drop out.

Some type of spring loaded hard contact switch would be much better, but I haven't found one.

david

[Sean]

... I am eager to open up my own failed switch to see exactly what the root cause of that problem was.  I am guessing that it, too, was wear-related. ... I will post any additional findings when I get my failed switch open at some point.

Well, as promised, I did finally get my original switch open.

Again, my write-up with photos of the failure is on my blog, here:
http://ourodyssey.blogspot.com/2011/07/cavorting-with-some-dammed-french-broad.html#switch

My opinion of these switches stands.  They are a crappy design for their intended use.  I have two of them in service, on my house system (one for 24-volt and one for 12-volt loads).  I would be inclined to rip both out and replace them with something having a better design, but they are "built in" to my house battery bay in such a way as to make that quite a project.  Since we've only ever operated them perhaps half a dozen times in seven years, they will probably survive in our application once I cobble together a good switch out of the two broken ones.  But if we had to operate these on a routine basis, I would change to something else for sure.

What I did not write in the blog post is that, just before I removed the failed switch for good, I was measuring a drop of 2.5 volts across its terminals.  Mind you, this is with the batteries and all the major loads connected to the very same terminal; the only wires on the other terminal, because I couldn't get them to reach, were the connection from the Vanner and the Link-10 battery meter.  So while my inverter would report a voltage of, say, 28.6, the Link would report 26.1 -- quite a drop.  I verified the drop with a hand-held DMM as well.

My guess here is that ohmic heating caused the spalling of material from the contact pad.  A slight shift in the wiper position then caused the voltage drop, leading to far more ohmic heating, to the point where the bakelite ring that moves the wiper simply crumbled.  Once that happened I was no longer able to either turn the switch off, or move the wiper back to a less precarious position.

If you have one of these switches on your coach, I recommend that:

1. you never, ever operate it under load.  Make certain all loads are off at their individual switches or breakers before operating the battery switch.

2. make certain you operate the lever fully into the "on" position before placing a load on the switch.  The switch will carry current through some 30 degrees or so of travel, and you only want it to carry current at the very end of this travel.

3. you operate the switch seldom, such as when removing batteries, or doing maintenance on the alternator or starter.  If you need a switch that you can operate for each and every trip, I would strongly recommend you choose a different model.

FWIW, and hope this helps someone.

-Sean
http://OurOdyssey.BlogSpot.com

[Zeroclearance]

Sean, if you had to do it all over again,  which switch would you use for a high amperage battery bank.

I also have to ask, did you purchase the Deka AGM's..    I just picked up another set and the prices I got were alot better than what you mentioned.   If you want to call me I can share the info with you.

[David Anderson]

Who makes this switch?  From your pictures it is poorly engineered, and I'm being nice.  I'd be willing to write a letter to the company of origin describing my disgust.  I think Dick Wright needs to take these off his inventory.

I noticed after a couple of years of use (with very little switching) that oily goo started to leak out of mine.  That indicates heat and impending failure.

And like you suggested, I never, ever turned my bank on with loads.  Failure nonetheless, however.

David

[bevans6]

I think it's the stock switch on MCI's like mine, as well as many others.  I had to operate it under load once or twice, but I had a runaway starter motor with a failed solenoid, so lesser of two evils.  Knowing that you need to think about something is the first 90% of the battle won, so thanks for the head's up.

Brian

[Sean]

Sean, if you had to do it all over again,  which switch would you use for a high amperage battery bank.

I used these switches because I had them; they came with the bus and were among the very few items we re-used in the new conversion.  If I were to do something different, I would have to research it, something I have not done.  So I am sorry I do not have a pat answer for your question.  That said, I am usually impressed by items made by Blue Sea Systems, and I know they make a switch in this rating.  Honestly, though, I think I would rather use something spring-loaded, where the contact is made or broken rapidly, completely, and with a "snap" action.  I've used similar switches and even circuit breakers, known as "air quench" type, in the past, just not in this small a rating.

I also have to ask, did you purchase the Deka AGM's..    I just picked up another set and the prices I got were alot better than what you mentioned.   If you want to call me I can share the info with you.

I got a quote from DC Battery in Miami on the Dekas for $426 apiece, a great price.  However, we've decided to go with another set of Trojans at $488 each in Myrtle Beach.  Not a great price, but I know without reservation the Trojans will fit my racks and cables, whereas the Dekas are just a hair larger, raising questions about them fitting.  Also, Myrtle Beach is closer, and cooler.  If we had been closer to Miami, we'd do that instead.  They also quoted me $483 on Fullrivers, which is a fantastic price, but the terminals are different and we'd have some work to do to adapt the cables.

If you have a line on a better deal, let me know.

Who makes this switch?  From your pictures it is poorly engineered, and I'm being nice.  I'd be willing to write a letter to the company of origin describing my disgust.  I think Dick Wright needs to take these off his inventory.

The labels on mine read "Underwriters Safety Device Company, Chicago, Illinois" as described in this original post:
http://www.busconversions.com/bbs/index.php?topic=20461.0

However, I can't find them anywhere and think they are out of business.  I note the identical switches sold by Mohawk, as reported in that thread, have Mohawk's own label on them.  I have no idea how the ones currently sold by MCI are labeled.  Interestingly, the stamped metal label, riveted to the case, actually serves to retain the lever friction spring assembly.

I agree, they are poorly engineered, which is why I started this thread, to warn others.

I think it's the stock switch on MCI's like mine, as well as many others. ...

Many posts in that original thread would suggest you are correct, that this item is stock on many MCIs.  It also appears to be stock on a number of GMs and, apparently, Neoplans made in the US.

-Sean
http://OurOdyssey.BlogSpot.com