I agree Kadees are excellent couplers, and if I were starting fresh I'd go with the scale-size whisker type wherever possible.  At present I have a mixture of 5s and 58s, and I do find occasional problems with coupling, as in switching.  If two mismatched couplers are not closely aligned - and on a curve they won't be - the larger face of the #5 may tend to push the smaller coupler aside rather than opening it.  It happens relatively infrequently, and usually a slightly harder bump will do the trick, but it can be annoying.  #5<=>5 and 58<=>58 don't seem to have this problem.  So choose one size, and stick with it.

David

ps I have several magnetic uncouplers and have had no problems with uncoupling or "delayed" uncoupling with either size coupler

I've used Kadee #5, #58, and various specialty types (overset or underset shank to handle coupler height issues).

I've NEVER had a Kadee break (as in the knuckle broke off).

I do occasionally have the knuckle's coil springs pop-out. Not a bit deal. With a tiny bit of practice these can be replaced faster than it takes to go to the parts drawers and get a spare.

So what's the deal with #5s are more reliable than #58s?

It's nothing to do with mechanical failure (though if you're pulling 300 car trains the smaller amounts of material in a #58 might cause them to fail before a #5 would fail).

Instead it comes from the smaller size of the coupler being more persnickety during operation. In particular

• A #58 has a smaller gathering range than a #5. Gather range is how far the coupler can be off-center and still couple to another #58 when cars are pushed together. Because the couple jaws of a #5 spread wider than a #58 it has a wider gathering range, meaning that they are less likely to fail to couple when cars are pushed together. This of course isn't a serious problem. Reach in between the cars with your finger, uncoupling pick, small screwdriver, or whatever and shove the two couplers closer to their center position then retry coupling. Heck, it's even prototypical to have to manually center couplers.
• Because a #58 isn't as thick vertically, they require more precision when mounting to ensure they are at the correct height. Definitely use the Kadee height gauge to check this before putting a new car into service. Also be sure there isn't much vertical "wiggle" in the coupler installation. This allows the forces on the couplers when they are in a moving train to push the knuckles up or down. Use shims to minimize the vertical "wiggle".  Because a #5 is "thicker" vertically, it is less sensitive to these issues.
• There's another problem that can't be solved by precision mounting. That's your trackwork. If your trackwork has lumps, bumps, wobbles, or vertical kinks the coupler of car A may be at a different height than the coupler of car B. If the lumps, bumps, wobbles, etc are big enough, this can lead to the coupler of car A climbing high enough to pop out of car B's coupler knuckle and you can experience sporadic uncouplings. A #5 Kadee, with its greater vertical thickness is less likely to uncouple on lumpy track.

As someone pointed out earlier (the Prof I think). Kadee couplers work well. If you have operational concerns or are short of $$$ then the #5 will likely be better for you. If you like to take close up photos of your rolling stock the #58 (and their coupler brothers) are what you'd like. Note however, that if you're a prototype appearance fanatic, you might want to consider the Sergeant couplers. They have some of the best prototype appearance around. However expect more gathering range and sensitivity to lumpy track with them.

What about other couplers? Kadees are made from metal. There are other couplers, made of plastic out there that are generally lower cost. A plastic coupler is most likely going to be strong enough to run trains. The thing to watch out for is how the coupler knuckles are held closed. On Kadee's there is a metal coil spring that keeps the knuckles closed. NEVER use couplers with a small plastic tang to hold the knuckles closed. These tangs do not work reliably. When cars are left pushed together for a few minutes or longer, the tang will lose its springiness and you are likely to have your train break apart in multiple places when you start pulling it. Unless you are really tight with $$$ I'd recommend staying with Metal couplers. As I said earlier, I've never had a Kadee coupler break.

Cheers,

Charlie

The reliability is a simple statement of reality and no slight aimed at Kadee whatsoever..

A larger mating surface will always be more reliable than a smaller mating surface of the same design.

The #5 head is 0.148"

The #58 head is 0.135"

The difference in height between two cars [between two coupled cars] is not .011", but rather, 0.22" overall.

The moment of failure occurs once the centerline of the coupler passes the top/bottom of the head, hence the moment where it is easier for the coupler to go up or down vector versus hold the train in the forward/backward vector, unless there's no up/down play whatsoever.  With a #5, you have 0.22" to your margin of error before the same failure that occurs on a #58.

Sorry Benny, that's the problem with the internet. Your shorter statement made it seem that their scale couplers were not quite as good ... your full explanation makes sense, and is true.

And it's kind of what I and a few others were writing, that NMRA coupler height standards need to be adhered to when using #58/158's vs #5's. Otherwise that 0.22" can make a big difference.

...after considering this same question awhile back. Personally the appearnace is just not that big a deal to me. I like to think there are other things on my layout that will draw the eye rather than what size my couplers are.

The #5's are a bit more economical and a bit more forgiving. In other words, whats already been said by others.

Just make sure whatever you use that they are metal. The plastic shanked and jawed Kadee knockoffs don't work so well in my experience.

Michael

The main reasons the knockoffs don't perform, from what I found, is that 1) the plastic is too slippery, and 2), the plastic shank does not resist lateral deflection.

But they do work well in limited "use until bust" service.

"Use until bust"

Kind of reminds me of the world we live in today when things aren't really meant to last anymore. Use it till it breaks and then throw it away and buy another one.

Michael