Low DCS Signal on New Layout

I would try the obvious swap channels test method. If it follows the channel, then that channel might have a bad ACT244.

Again, no guarantees that is your problem, but it does the basic  is it your track and wiring vs is it a bad channel on the TIU testing.

https://ogrforum.com/...-aghr-and-mth?page=7

I watched your video again. Don’t see much that can go wrong as far as wiring as it’s just a simple loop with no turnouts. If your TIU checks out okay after swapping channels.

Did you test with the same engine that got solid 10’s ?   Did you test with any cars behind the engine or anything lighted or just the engine itself  ?    Certain lighted cars can degrade the signal.  Do you have toggle switches to shut down both loops ?  If so, you could have one that may not pass the signal well. It seems though your signal jumps around.
Are both outside rails connected together ?   Or did you isolate one to trigger accessories.

My own layout breaks all the rules of wiring DCS and still has solid 10’s. Probably way over wired.  Your loop just seems to simple to have any complicated wiring issues.

I'm with Vernon, first step is to swap the two TIU channels and see if this follows the track or TIU channel.

Thanks guys.

Got too far ahead of myself and forgot to try the basic stuff. First, I realized that the fixed 1 track that got all 10's was running a TMCC engine. Apparently you can't do a signal test with a TMCC engine as the test number never changes. Verified this by trying that engine on the other loop.

I did have light passenger cars on both loops and both loops do have an outside insulated rail section for accessory activation, but nothing is hooked up to them yet.

So, I took all the cars off and just ran a MTH PS2 engine. On both loops I only got a signal of 1 to 3. I tried a second MTH PS2 engine and got the same results. So both loops are low.

I wired the layout as follows. 180w bricks into TIU. On the TIU outputs I have two pairs of wires connected to each channel. One pair goes to a terminal block on one end of the layout the the other pair to a terminal block at the other end. From these blocks the power drops are run to the track. Attached is a basic diagram of the wiring from the bricks to the terminal blocks.

Could the issue be that I ran two pairs of wires from each TIU channel? If I only ran one pair of wires, I would have had another terminal block at the middle of the layout to split the wiring going to both ends of the layout, so I don't think it would be any different.

Yes. It’s best to run one paired wire from each TIU output to a terminal strip. Preferably somewhere in the middle of all the action. Then run from that strip out to the track itself. This will if nothing else create a shorter wire run. There was something in Barry’s book about splitting the signal one time. In your scenario it’s split at the TIU and then again at the terminal strip.

That basic diagram looks fine, from each of those TB's, you should come out with the star wiring to the segments of track.  Remember, there should only be one connection from the terminal block.  Here's a typical star wiring plan, note only the inside loop is wired in this example.  The red lines across the track denote where the center rail is separated with an insulating pin.  Note that a run comes from the common terminal block direct to each track segment, and the center rail is cut midway between the power drops.

The reason I did not run one pair of wires out to one centrally located terminal strip is because my wire runs to the track would vary greatly in length. The layout is a folded dogbone. I always thought it was important to keep the runs closer in length. I did try disconnection one pair of wires and see how the signal was on just half the layout. I still only averaged 3 to 5's.

John,

That is exactly what I have except I have a second pair of wires off each channel going to the second terminal strip at the other end of the layout.

I am also using barrier strips instead of terminal blocks. I did this on my last layout as well. I don't think that could be an issue.

I don’t think the length of the wire runs have to be that close in length.

The way it’s wired presently. Can you dissconnect all the blocks except one at the terminal strips. So only one block is powered up and re test. Sort of like creating a programming track. It should yield a 10. At least verify the TIU can put out a 10. I’m not sure as to your TIU version. I know some of the earlier version were greatly affected by adding Marty F’s magic light bulb fix.

It worked fine on your previous layout. Just curious as to how it was wired. I know the paired wiring is the recommended way to do it. I’ve seen plenty of layouts run just fine with the old school common buss wiring.

@NJCJOE posted:

John,

That is exactly what I have except I have a second pair of wires off each channel going to the second terminal strip at the other end of the layout.

I am also using barrier strips instead of terminal blocks. I did this on my last layout as well. I don't think that could be an issue.

I only have one tip.

When you have an issue like this, suspect everything until you prove it not to be. Stuff breaks. Joints fail. etc.

Dave,

I'll give that a try when I get a chance. The TIU is a Version I3a. These had signal issues from the factory. I sent mine back to MTH to be corrected. They marked it with a silver dot on the bottom.

The old layout was wired in a similar fashion except I had only one pair of wires going to a central terminal strip. The layout was smaller.

@NJCJOE posted:

Dave,

The old layout was wired in a similar fashion except I had only one pair of wires going to a central terminal strip. The layout was smaller.

That is the way it should be wired. It is a myth that the wiring needs to be equal form the terminal block.  It has been proven years ago.  Star wiring was a poor name for the wiring and it confuses everyone that the wiring needs to be equal like a star.  Radiating wiring would have been a better name in my opinion.

Well, I set up a 6' long test track right next to the TIU. I got 8's and 9's on the variable channel and 9' and 10's on the fixed channel. So, the TIU appears okay. I may have to redo the wiring so that there is only one split.

Actually, you should be getting solid 10's on a test track.  Try a "magic" light bulb, AKA any 18V small lamp, across the end of the track and see if that changes things.

I'll give that a shot John.

Tried the magic light bulb on the test track. Got the same reading, 8 and 9's. I think this TIU being a Version I3a is a little weak with the signal and then I compounded the situation with how I wired the layout.

Yep, I haven't gone the extra mile on my layout, I never did get around to chopping the rails between power drops, and I only went with a modified star wiring design.  I get 9's and 10's all over the layout and have no problem with plopping a locomotive anywhere on the track and adding it and starting it up.   I have around 500 feet of track running on four TIU channels, the folded dogbone mainline is 140 feet on one TIU channel.

Joe, you can live with 8’s and 9’s. This is a reach. You mentioned speaker wire. Most of mine was done with automotive wire with buss wiring. When I added on I did use speaker wire. Much like what you are doing. Not sure if it’s all created equal. What I did use was 14 gauge. It was very fine with a lot of strands. Seemed to work well. My wiring out of the TIU is 12 gauge. I know nothing about standard gauge. Watching your video it wasn’t what I expected as they ran slow and quiet. Pretty cool. Do they feature larger motors and draw more current than o gauge ? Just wondering is 14 gauge enough on long runs. The trains seem to run fine so you would think the signal would be fine to
How long is the wire run on your 6 ft. test track you set up and did you use the same wire that’s on the layout ?  If it’s very short. Could you easily add a much longer length just to see if the signal changes for the worst with a longer run. I know what was recommended at the time  was the wording  quality speaker wire. Not knowing much as far as what that meant I bought on the expensive side and luckily it worked well.

You are right, all speaker wire is not the same. The speaker wire I used is 100% copper high quality. I used the speaker wire on the last layout as well. Standard gauge does have bigger motors. My last layout used 16 ga. wire. I think the 14 ga. I used on this one is suffiscient.

The trains do run fine. I'm just having trouble loading new engines and the issue with duplicate softkey commands.

@NJCJOE posted:

You are right, all speaker wire is not the same. The speaker wire I used is 100% copper high quality. I used the speaker wire on the last layout as well. Standard gauge does have bigger motors. My last layout used 16 ga. wire. I think the 14 ga. I used on this one is suffiscient.

The trains do run fine. I'm just having trouble loading new engines and the issue with duplicate softkey commands.

I've used speaker wire with good results. Following Barry's list helped me. Changes to the layout or any equipment threw in a few variables.

A couple of times it wasn't the layout. It was things like PS2 batteries going dead, equipment killing the signal, A bad software release, etc.

I felt like I got pretty good at getting premium signal as I built new layouts. You have to keep an open mind about solutions while trouble shooting.

Test everything and rule out everything you can.

Sometimes you get an engine that is weaker at picking up the signal. I use that engine as a test bed so I know if that one works, my others will as well.

I just finished setting up DCS on my O Gauge layout.  Signals were horrible until I did these things.

1.  Changed my feed wiring from 16 to 14-gauge, stranded copper (not copper clad aluminum!).  This made a big difference.  I changed one feed and engine on that section of track jumped from 6 to 10.  I did not do anything else, so I know that was what improved it.

2.  Went from 4 drops for the whole layout (legacy was 100% fine with 4 drops) to 12 drops.  I did not use center rail breaks either even though Barry's book says you should, I would have to tear up all my ballasted track. But it works fine and no double signals (maybe because its solid 10's).  Before adding drops, you can verify the signal will improve by taking some jumper wires from the distribution block over to the section with the low signal.  If it jumps up, you need a drop.

3.  Turned off Legacy signal (signal jumped from 5 to 9 just turning off Legacy signal, no idea why).  This is an easy one to try first, turn off your TMCC signal just to see if that makes a difference.

What I did is I basically put an engine on the track, tested signal, ran it a few feet, tested again and kept doing that around the whole layout.  That told me where I needed another drop.

I now have 10's across entire layout no matter where the engine is.  For my layout it needs tons of feeds for DCS signal.

And just to clarify, my track is super tight.  I have had solid 18v measurements every single spot on my layout for over a decade and no Legacy issues ever.

DCS is just way too finicky!

@Sean's Train Depot posted:

3.  Turned off Legacy signal (signal jumped from 5 to 9 just turning off Legacy signal, no idea why).  This is an easy one to try first, turn off your TMCC signal just to see if that makes a difference.

Yep, it's a known issue that DCS can be affected by the Legacy signal.  The proper wiring for DCS will help alleviate this issue.

@Sean's Train Depot posted:

DCS is just way too finicky!

No argument here.

Does anybody know definitively if a low DCS signal can cause duplicate softkey commands?

@NJCJOE posted:

Does anybody know definitively if a low DCS signal can cause duplicate softkey commands?

It was doing it on my O gauge layout when I had 4-5 signal strength.  When I added the feeds and got it up to 10, it no longer does it.  So, I would say it can definitely be a factor.

The duplicate commands are due to signal reflections on the track, the controller sees multiple commands.  This is what the various wiring "fixes" for DCS should correct.  Star wiring, DCS filters for pre-Rev.L TIU's, etc.

Joe, way back you mentioned you have an isolated outside rail to activate accessories but nothing hooked up yet. If you have wires attached to the rails not hooked up yet. Can you hook one to the terminal block or simply just jumper the outside rails together. Just to see if you get any improvement.

I was a early pioneer of DCS with a layout with buss wiring. Well wired 12 gauge buss and 14 gauge feeders. Every section of Gargraves had feeders. Way over wired according to what was recommended for DCS. 4 long blocks all with isolated outside rails to run a signal system. Removed the signal system and ran speaker wire feeders back to a terminal strip. Left the buss wring in place.  Went from 2’s and 4’s to 10’s. The signal system had 14v AC relays. Not sure if the relay itself caused the issues or just adding the outside rail and maybe the paired wiring fixed the problem. All my yard tracks not involved in the signal system all were fine with buss wiring once I got different areas on their own channels. I started the process by running 14 gauge speaker wire 20 feet to a isolated turntable track and nothing else just to see what a 10 looked like. Fixed one block at a time. Took about a month. I did experience a horn that wouldn’t shut off till it entered a block with a good signal.

Your layout just doesn’t seem that complicated. What you get on the programming track should transfer to the layout. Maybe just disconnect  your star wiring and work on one block at a time.

Thanks again guys.

True, the layout is fairly simple and this is why it's driving me crazy. I ran a few more tests. I took my 6' test track connected to the TIU with 12" of #14 speaker wire. I then tested all the channels,  both with and without the magic light bulb. The light bulb made no difference in the signal readings. I also used the same 180w brick for the testing.

Here were the results.

Variable 1 - 8-9's

Fixed 1 - 9-10's

Fixed 2 - 9-10's

Variable 2 - 8-9's

Here is a partial sample of the readings I got while running on one of the layout loops. This was on the Variable 1 channel which has been set to fixed.

8, 7, 4, 3, 2, 2, 1, 6, 3, 5, 2, 4, 2, 7, 4, 2, 3

As you can see, the readings are all over the place. The readings even vary within the same block.

Dave, I tried what you suggested and jumpered the insulated outside rail sections. It made no difference in the readings.

John, what are the filters you are referring to? I have been using DCS for a long time, but my last layout had very few issues so I never really paid attention to some of the fixes that came out.

I also realized that this TIU, which is a Version I3a that had the signal modification done by MTH, was not used on my last layout. I had a different TIU on that layout. This one was mostly saved as a backup and for trains around the Christmas tree so I may have never noticed if it still had a weak signal.

Do you have another TIU? If so swap it in and see if the issues persist.

Unfortunately, I don't.

Would it be possible to borrow a tiu from someone and see what happens with that on the layout? I am a non person with DCS, but given how simple your layout is, something doesn't add up. Kind of reminds me of a power supply that puts out great voltage until you put the least load on it.....a known good tiu would at least eliminate that from being the problem, and possibly save you a lot of running around.

Looking at the overall layout and your posted wiring plan. Looks like you have maybe 4 electrical blocks on each loop. I’m guessing each block is maybe 15 feet. You seem to get good numbers with a very short wire run on your test track. Can you place the TIU under the layout and just power up one block only using the feeder wires at the terminal strip. . Nothing else and see what you get. It should be similar to to your test track. If good. Place the TIU back in place and test just this one block again with the wires in place that were from the TIU output to the strip. The only thing that should be changing is the length of the wire run.

With 60 feet of track and four drops, I can't imagine it being a huge issue.  I have 140 feet in a folded dogbone with a single TIU channel feeding it, I get mostly 10's with an occasional 8-9 at times.  I haven't even gotten around to cutting the center track between drops to eliminate the signal issues, it hasn't been an issue, so I did more important things.

If 60 feet of track is giving you that much problem, I'd REALLY want to test the TIU independently and also test another TIU on the layout.

So, one loop is approximately 87' long and the other is about 82' long. Each is broken up into 8 blocks, so each block is roughly 10' long.

I don't see any reason this length of track should be an issue for a single channel of the TIU.  I'd think testing the output of the TIU might be a good place to go next.

@NJCJOE posted:

John, what are the filters you are referring to? I have been using DCS for a long time, but my last layout had very few issues so I never really paid attention to some of the fixes that came out.

I also realized that this TIU, which is a Version I3a that had the signal modification done by MTH, was not used on my last layout. I had a different TIU on that layout. This one was mostly saved as a backup and for trains around the Christmas tree so I may have never noticed if it still had a weak signal.

Deats filters | O Gauge Railroading On Line Forum (ogaugerr.com)

Just a follow up on this. I rewired the layout this morning to eliminate the double split of the DCS signal, which is how I originally wired it. I now have 10s everywhere, except for a drop to a 7 for a second. Also, no more double commands from the soft keys. Moral of the story, don't split the DCS signal more than once.

Joe, I’ve been looking for a follow up post to see where your at. Glad it’s working now and didn’t involve having to buy another TIU. It’s tough to move forward on a layout with other plans when it’s not running well. Sometimes it’s best to walk away for a bit and work on other projects.