Trying to select a Polyfuse to protect uncoupler magnets

Do an amp draw on the uncoupler then give it 1.5 times the max current. With a polyfuse you can prolly go at the tested current draw since it takes time for them to trip

Forum member PLCPROFhas does some testing on this. I think you should connect your multi-meter in line on a few and get some readings

From another thread:

As far as I am concerned the best way to do this is to run the uncoupling tracks on fixed voltage and put an undersized resettable fuse in the power feed to the uncouplers. I posted detail on this some time ago.

The principle of operation is simple. You size the resettable fuse so it will definitely trip a few seconds after the uncoupler is energized. If you release the button before that time the fuse never trips. If you hold the button too long the fuse trips, protecting the uncoupler. The fuse then resets a few seconds after the button is released. If the button is never released the fuse never resets. No timers, no relays.

The Fastrack uncoupler draws about 1.8A at 14 volts. If you use a resettable fuse that trips at, say, 1 ampere you will get probably 2-3 seconds of actuation. I use a 1.5A trip fuse, that gives me about 15 seconds. You will need to choose the exact fuse based  the voltage you use to run your uncouplers and the operation time you want. The fuses are cheap, pick up an assortment in the .8 to 1.5 amp trip range or thereabouts and experiment with them.

Note that operating cars use much more current than the uncouplers. If you are using operating tracks this solution may not work for you.

The same solution could also protect switch machines, although I don't know what size fuse would be appropriate.

I agree with Andrew's observation regarding the PTC.  If you expect 2 amps of current for operation, I'd consider a rating around 1.3-1.5A for the trip.  That will give you plenty of uncoupling time.

This is one of many tables, this one from Littlefuse.  Note that the 2.5A trip PTC takes up to 7 seconds to trip on a 200% overload.

Thanks guys, those are some pretty good hints. Track voltage is fixed at about 17V, but it has always been my plan to bring in a separate feed of a lower voltage, say 14V. Sounds like I should get some values between 1 and 1.5 amps and try them.

The above info is good and better for calculating your own specific values, but Here's a list of some that different forum members have used for different things. There is one listed by PLCProf for uncoupler tracks, but I believe he was using Fastrack. Might also be good for reference for other stuff. 

I'd be curious to know what the difference is between the Fastrack version and the classic 027 units, in electrical terms. Is Fastrack more, less, or about the same?

Carl - Can you post the link to the topic where pulled that text from? It looks vaguely familiar, as if I read it back when it was current, but it also seems like it was a while ago.

The classic 027 postwar magnets vary quite a bit in their current draw.  I tested a some of mine and found 1.5 to 2.8 amps in a sample of 6 magnets.  I have installed a few PTC's and they work well.  I get about 4 or 5 seconds of use from a button push before the coil turns off.  It resets in 20 or 30 seconds.  My operators managed to melt 3 or 4 magnets over the course of about 200 operating sessions in the last 15 years before I installed PTC's.  There are over 70 postwar magnets on the layout.  It wasn't a huge problem, but annoying.  Nobody likes a hot electrical smell during a op session!

Bob, it's interesting to know that there is so much variation between the 027 magnets. I had hoped to use a single PTC to protect five units with only one firing at a time. Now it looks like I'll be needing to use one PTC per magnet. Good to know that it works, since you've had so few failures over all of those years in service.

Gary, my operations guru (an HO guy), is somewhat skeptical of the Lionel style uncouplers. I've noticed that the newer cars that I've recently placed on the layout seem to work rather well, even among the various manufacturers. It's the older pieces that may cause problems. I'd like to re-truck my Weaver fleet, which would potentially clear up a big chunk of the problem.

I think if you pick from the lower end of the range of magnet currents, you'll be fine.  It only takes a second or so to do the uncoupling task, you have a lot of leeway.

An alternative approach is to use a current detector, that cuts power after current has been detected for a preselected time.  If placed in the feed for all controllers or all AIU uncouplers, only one is needed.  I cobbled one together for a test, and now I can't find the tech sheets that came with the cheap Chinese components that I used.  It's in my pile of almost-done projects.

Sounds like a plan John. Without measuring all 70+ units, I think a 1A PTC should do to keep me out of trouble. I'll place one on each as I connect them to the layout, but I'll have to retrofit the ones I already done. Not the end of the world. It's a lot easier than replacing fried units.

This became more important when I was testing the Adafruit / relay units, and found that they can get stuck on from time to time.

RJR posted:

An alternative approach is to use a current detector, that cuts power after current has been detected for a preselected time.  If placed in the feed for all controllers or all AIU uncouplers, only one is needed.  I cobbled one together for a test, and now I can't find the tech sheets that came with the cheap Chinese components that I used.  It's in my pile of almost-done projects.

I had thought about something like that. I would probably want to do a bunch of those, like 6 or 7, if I went that route. The beauty of the PTC is it's just a single component, and they're fairly cheap at 60¢ each. Bottom line is you get the same effect with the simplicity of the single component.

Elliot, I can't quarrel with that--the price and simplicity is overriding.  I am surprised to hear of the variation in magnet draw among Lionel magnets.  I do note that the UCS and the 027 magnets differ in shape, size and power.

I just placed my order, and there was a huge price break at 100 pieces. They went down to 39.6¢ each. That was nice.

Big_Boy_4005 posted:

I'd be curious to know what the difference is between the Fastrack version and the classic 027 units, in electrical terms. Is Fastrack more, less, or about the same?

Carl - Can you post the link to the topic where pulled that text from? It looks vaguely familiar, as if I read it back when it was current, but it also seems like it was a while ago.

Elliot,

It was the "momentary switch for Fastrack uncouplers" thread

Thanks Carl. Maybe I didn't read that one because it said Fastrack in the title. Some good stuff in there, but the relay method is better saved for a different application. I think the Prof's idea works best for this uncoupler business.

Eliot, could you share which part #, item # you purchased.... I have about 25 Gargraves uncouplers on my 3 RS layout using momentary pushbutton controls, but if I can protect them for less than a $1.00 each, I am very interested in getting the right component.   

GRJ, I find that Digikey doesn't have any polyfuses in the 1.5A range that are rated for >16 volts.  The one you mention above is discontinued.  I await with bated breath Eliot's reply.

How hard did you look?  Here's 59 in stock that have trip values between 1.4A and 1.8A, all rated at 30 volts or more.

Digikey PTC Search

Here's a search for just 30V parts, 17 of those.

Digikey 30V PTC Search

I wonder if there's some confusion on the terminology.  The PTC Ampere rating is the Hold-current value.  In other words a 1A PTC will operate at 1A (continues to operate, does not "blow").  This is kind-of sort-of backwards from a traditional fuse where the rating is the Trip-current value...that is, a 1A fuse will trip at 1A (no longer operates, fuse "blows").

Note that GRJ has been careful to specify he is discussing the Trip-current value.  For protecting an Uncoupler coil, you choose a PTC value well below (say half or less) the Trip-current value.  Stated differently, you choose the PTC Amp rating to be well below the known operating current of the coil; again, this is kind-of backwards thinking from choosing the Amp rating of a traditional fuse.

If this is all understood, great.  Some of the comments suggest otherwise. 

Obviously, I put in the wrong search parameters.  Mrs. RJR always says I should just call her whenever I need to find something.

Thanks, GRJ

Stan, your post and mine crossed.  I guess the issue is to decide what trip rating will give you a chance to get the uncoupling done before killing power.  I'm trying to come up to speed on these PTCs.  It appears that some current must be flowing, at least periodically, for so long as the device is tripped, to keep it activated; correct?

OK guys, here's the link to DigiKey's entire PTC list.  Up at the top of the page there are a set of filters. Select the current max value you want, then "apply filter". The list will go way down from 5430 items to a much more manageable number. You can sort by price and some other criteria. Watch out for minimum quantities, some you need to buy by the thousands. There are price breaks at specific "even" quantities.

This is the one I bought.

Big_Boy_4005 posted:

This is the one I bought.

QUICK!  Go cancel that order!  Call today!  You can't use those, a number of the parameters are all wrong!

This has a very low trip current, it won't do anything in the circuit.

Here's a good choice: https://www.digikey.com/produc...070-2HFCT-ND/5052874

And another: https://www.digikey.com/produc...RHEF070HF-ND/5029792

John, you saved my bacon on that one. Correct part ordered, piece of cake. The ladies at DigiKey are very nice.

Glad you were able to intercept it, you would have done a lot of work for nothing if you wired all those in.

Even with the ones you have on order now, I'd test them with your couplers to make SURE all the assumptions we're making are really true.

Thanks John, will do.

FWIW, I just tested AC amps draw on some uncoupling tracks.  Applied 17 VAC sine wave.  UCS gave me 2.6 amps, 2 Lionel 0-27  tracks gave me 2.55, and 2 Gargraves gave me 4.1.

Wouldn't the two GRJ just noted open too quickly>  I read the specs as saying---I could be misreading---that at 1.4 amps, they open in 3.2 seconds.  I think 4-5 seconds would be needed for normal operations.

Read it again.  Trip rating is 1.4A and time to trip max is 3.2 seconds at 3.5A

I've posted some curves before about PTC trip times, as I recall at 135% of their trip rating, it took a very long time to trip.

Note also you don't want or need 17 VAC, 12VAC will probably be plenty, and that will cut the current down.  The object of the exercise is to uncouple the cars, not suck the couplers off the truck.

RJR posted:

...

It appears that some current must be flowing, at least periodically, for so long as the device is tripped, to keep it activated; correct?

 

RJR posted:

... I think 4-5 seconds would be needed for normal operations.

Right.  Unlike a traditional fuse which "blows" and interrupts all current, a PTC flows a reduced amount of current which keeps it warm (so to speak) and in the activated/tripped state.  In other words, it goes from a low-value resistor to a high-value resistor as long as the fault condition persists.  As shown in the table , this warming power can be quite substantial (1.4 Watts).  Release the pushbutton switch to break the coil current, and the "warming" power goes to zero, the PTC cools down, and its resistance drops over time to live another day.

Perhaps stating the obvious, but these are NOT precision devices! Note how the initial resistance value varies almost 2:1!  And since it's this resistance which effectively warms up when current flows through it, what does that suggest about how accurately/precisely you can predict how fast it warms up and hence when it will trips?!

And when you see a characteristic like the resistance of the PTC 1 Hour later, do you really expect this to be a device that you can depend on to trip in 4-5 seconds vs. 7-8 seconds vs. whatever seconds?!   I realize you were just throwing out some numbers so I apologize if it appears I'm picking on you - I'm not.

For uncoupling magnet protection they are, arguably, not being used for their intended purpose.  If you want to limit uncoupler operation to 4-5 seconds, use a timer module (about $1-2) which has been discussed in other OGR threads.  Trying to manipulate, massage, or work the datasheet curves of a PTC to get 4-5 seconds of operating time is a fool's errand; that's just my opinion.  As I see it, if you're trying to protect the magnet coil from overheating/melting you should bond a thermal cutoff switch to the magnet coil itself.  So when the temperature of the coil itself reaches whatever degrees, it cuts off power to the coil until it cools down.  Thermal cutoff (TCO) switches are, say, 50 cents.  But very inconvenient for an installed layout to pull out UCS track sections so using a proxy (i.e., a PTC) as a thermal protection device is a practical solution.

You can get PTC devices on eBay for quite attractive prices (free shipping from Asia).  For example here's what I bought for Realtrax UCS which apparently have higher current than the Lionel/Gargaves being discussed.  As suggested in earlier posts, I think experimentation with a few different values on your configuration is a good idea.

There is considerable difference between the pulling power of these electromagnets.  Gargraves seem to be less than UCS.  I have some that, even with higher voltage, work spottily.