AIU doubts, questions, and concerns...

Dave,

Since you have a copy of my book (thank you!), the short answer is that the AIU will most certainly handle up to 10 FasTrack switches, one per AIU SW port. Refer to pages 72 (bottom) - 73 (top) of The DCS O Gauge Companion 2nd Edition for words and pictures regarding connecting FasTrack switches to an AIU.

Regarding the power-handling limit of the AIU's ACC ports for track power control, refer to pages 87-88 in the book. As noted, each AIU ACC port can handle up to 4 amps sustained draw.

Each DCS engine will draw approx. 1.5 - 2 amps, with smoke on, at a moderate speed.

This and a whole lot more is all in MTH’s “The DCS O Gauge Companion 2nd Edition", available for purchase as an eBook or a printed book at MTH's web store!

The wiring you describe for switches is a standard approach.  The AIU makes a momentary contact for 2-3 seconds when you press the button on the remote, and then opens. You can hear the clicks.  While I haven't opened any of mine, it acts as if there are 20 relays on the 10 switch circuits.

Hello God...err...Barry...reminds me of a Jim Carry movie, huh?

OK, so you've positively answered my question regarding ACC ports, current, and single-engine train current (or more specifically, burning the crap out of them). The answer is, not a problem.

But, there's still the issue that if the AIU has SPDT switches, it's like running a Lionel Fastrak layout with a train running on an outer loop, while manually holding the remote switch in the wrong direction.

The problem? The manual switch will force the hypothetical train in the opposite direction (for the sake of this discussion), but as soon as the train hits the split activation track of the switch, there will be a conflict as follows:

1. The manual switch held by the operator will force the train to switch

2. The train, hitting the activation of the switch will try to force it opposite

I'm not that stupid to force a conflicting condition, but I find myself concerned if the AIU will essentially create such a situation.

Without an AIU, I cannot know whether it has unique contact configurations (that I would not reasonably expect but would appreciate), that would create such an electrical conflict.

While waiting for other input, I will attempt the conflict above; I will force a MTH Diesel engine, approaching from the straight direction, to hit a Lionel switch held (manually, by hand..by me) in the turn direction.

This would conceivably recreate the scenario I've expressed in this post, in that if the MTH AIU is truly a SPDT switch, it could (or most likely be) problematic for Lionel Fastrack users.

Conversely, so many users use the above configuration, I honestly suspect that there is more information available that is not made public (for example, like the switch contact outputs are actually 3-state; straight/none/turn).

That would explain why so many people are happy, but wouldn't explain my confusion as to whether the AIU has:

a) 3-position relays: (Mom-Off-Mom)

b) Two relays for each switch output, which could easily accomplish my question.

Dave, did you read my post above?  The AIU does not send any power to the switch after those 2-3 seconds.  What you describe is not a problem.  I have some 45-50 switches, some wired for non-derailing operation, and have no problems.

Without the benefit of physical possession (ship fast "Silver City Hobby"!), your explanation makes perfect sense. However, implementing this (in the AIU) is not as simple as your answer suggests (though I accept your answer).

1. I don't know off any SPDT relay, of reasonable economical cost, could provide the tri-state positions you suggest the AIU provides. I'll concede you are correct, but only seeking an understanding as tri-state relays are nearly impossible to find on the current market.

2. Implementing two relays for the "Switch" ports in the AIU would most certainly be more cost effective than specialized tri-state relays. Since everything I've read says that there are only 20 relays in the AIU (10 for AUX, and 10 for Switches), it doesn't make sense (and is scientifically impossible).

Clarification please?

Thank you!

Was composing my reply when you posted yours (sorry!).

Will pause momentarily to allow all replies to be fed into the forum for review. In the meantime, I will NOT cause the switching conflict I suggested I would above.

However, if what you say is accurate, it appears that the MTH AIU has either:

- Tri-state relays: {on}-off-{on}    A very unusual relay to find

- Two relays for each switch output...a very acceptable and valid solution, but it doesn't match the count of the number of relays in various documentation

Either of the two satisfy my concerns, and upon confirmation, would close this topic.

Two relays per switch output. 

Note that for the Accessories, the terminals are labeled COM, NO, NC like what you'd expect for an SPDT and implemented (wired directly to) a single SPDT relay.

But for the Switches, the terminals are labeled COM, Straight, Curve implying (to me anyway) that it's a different animal. 

If you like technical, when you get the AIU, open it up and look at the components and you'll see the two relays per switch with the NC contacts unused for each relay.  The {mom} timing is implemented in software...that is the relay coils are connected to digital chips with the timing performed upstream.

So to paraphrase (since I ordered the AIU but not yet in my possession):

- There are 10 "Accessory" relays

- There are 10 "Switch" relays, for one direction (let's assume straight for this post)

- There are 10 "Switch" relays, for the other direction (we'll assume turn for this post)

That's a total of 30 relays inside the box...certainly reasonable and enough space plus control electronics. {The issue of momentary timing under software control is agreed.}

Can anyone confirm 30 relays inside their AIU?

Dave

While I have not opened my TIU, I can confirm it acts like there are 2 relays for each switch connection.

Since that is the least expensive and simplest solution Occams Razor says that's the answer.

I can confirm the AIU relay connections throw the switch for a short period of time then return to the open state. All the switches I've ever used work on this style input.

I haven't had an AIU open for some time. but I'm sure there aren't 30 relays in there.  I'm not sure how all the switching is done, but they do it with no more than 20 relays.

GRJ, can I double down my account with you? 

30 relays and my debt is cleared. 20 relays and I'll owe you double.

I'd love to take the bet, but upon further reflection, maybe I'll pass.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

I don't know why I thought there was less than 30 in there, but that does make the most sense.  It's also apparently the way it's done.

Dave R et al.,

I truly appreciate this thread. I've been wondering about this topic for a long time!  Now it makes sense!

Ed

GRJ, I can't quite read the text in your photo.  What does it say on the relays besides Omron? 

I can't quite read the relays either, and I already put it back together.  I know last time I checked, they were 5A relays, but for some reason I didn't realize that there were 30 in there, I thought there were less.

There's a previous thread with closeup pictures of the relays that show the lettering on the relays.  This thread has a bunch of pictures: https://ogrforum.ogaugerr.com/t...os-of-control-panels

Here's one picture from that thread.

Thanks, GRJ.  I see the coils are 12 VDC and the contacts can switch up to 5 amps @ 250 VAC.

Remember, if you happen to be switching DC, the ratings may be less, especially voltage ratings.

True, but I would assume that a relay that can switch 5 A @ 250 VAC can switch that amperage at any DC voltage one might REASONABLY be using on a layout, say =<18VDC.  I suspect it is not too often that, with 3-rail O-gauge, one is controlling DC with an AIU.

Dave,

    I used the AIU for years and had no problems with any of my switches, I wanted to eliminate the AIU switch wiring so I went Command Control with my new Lionel FasTrack switches, it's definitely something to think about in the long run.

PCRR/Dave

The issue with DC is that when the contacts open, there is never a zero crossing to quench any arc, so the ratings are a lot more conservative with DC than AC.

A good question, whether one uses AC or DC, is whether when there is a heavy inductive load on a relay, one should use spark suppression.  I'm thinking of those old NJInt'l switch machines, which made a good spark.  I had wired mine so that the power was cut by the contacts on the switch motor rather than by the relay, but I can see that eventually contacts will go bad.  With DC, one can put a diode across the coil that generates the spark to short out the induced current (I did this many years ago when I ran RC boats), but this isn't feasible on AC.

A TVS across the contacts for AC should help a lot.

Being the relays are made by Omron, they probably have diodes across the coil. They supply relays for the automotive manufactures and they come that way for the automotive manufactures. They don't want that voltage spike backfeeding into your engine controller.

We're talking across the contacts, not the coil.

"With DC, one can put a diode across the coil that generates the spark to short out the induced current (I did this many years ago when I ran RC boats), but this isn't feasible on AC."

Again, we're talking a TVS diode across the contacts, NOT the coil.  The TVS will appear as an open circuit until it's breakdown voltage is reached.  The common values for model trains are 33 or 36 volts.

I understand the kickback protection for the coil, but were at the other end of the horse.

Riverrailfan, by "the coil that generates the spark" I meant the coils in the actuator that is driven by the relay, be they a magnet or the coils in a motor.  Sorry if I wasn't clear.

I doubt a TVS would suppress the spark that occurs on each activation.

On a PS2 upgrade of a postwar Lionel steamer , I retained the Lionel tender coupler but Lionel couplers MUST be driven through a relay rather than directly from the PS2 board.  Activating the couplers would cause the PS2 board to shut down or lock up.  Cure was a .02 ceramic disc from relay center contact to the NO contact.  Since AC was being fed to couplers, a simple diode wouldn't do the job.

I bought the dcs book also tonight and that was the first place I went does a aiu work with fastrack switches I also have 2 sc-1's and they don't so before I bought the aiu I researched the ? on weather they handled fastrack switches the once I bought the book I confirmed it. the lionel switch controller has a momentary switch in it to switch both directions but I think the anti derailing feature bypasses the direction yes if you hold it with your hand I doubt it would switch but other wise holding the controller it would still switch. since I bought everything tonight to go dcs . I have a lot of redoing I think. dremel the center rail in a lot of places and more feeders in lot of places

If I follow the books idea's on how to do it

There's no reason the AIU shouldn't operate anything bare contacts will operate, as that's what it has. 

The SC-1 uses small triacs to directly switch the power, so they're limited in what kinds of things they work with.  The SC-1 assumes a common ground, the SC-2 or AIU uses bare relay contacts so they're far more compatible with the full universe of switches and accessories.