PS3 Starting on Power to Track

make sure the wires don't lay on the torrid coil of the board.

PS3 board wire rerouting and board check with photos | O Gauge Railroading On Line Forum (ogaugerr.com)

MTH PS3 Steam - Positioning of Power Wires in the Tender.pdf (ogaugerr.com)

Also, Make sure there aren't too many engines on a single TIU channel at once all getting power. More than four or five? on a single channel and you may get this issue. What is your power source to the rails?

Maybe try them each one at a time, on their own with nothing else on the track.

How strong is your track signal?

Any TMCC/Legacy engines or other powered equipment on the same rails?

Has each been run recently to charge the caps?

Thank you for this information, very helpful.

A conventional reset might help ( one whistle, five bells ), along with rerouting the wires.

I do my upgrades and test them on my bench.  There is only one engine on the TIU on the bench, and I have about a 25 foot run of wire between the TIU and the track.  I also do my upgrades with pure DC on the TIU input channel.  This is how MTH does them in their tech center.  Note that when you're loading sound & chain files, you normally get a startup after the file loads, you can't stop that.

Once you've done a couple of power cycles after loading all the sound and chain files, they should not be coming up live with the application of power on the channel.  I've done a lot of PS/3 upgrades, and I don't have this issue, so I suspect it's something in your environment.

I do initial testing and running on test track.  I am using the commander for DC power and run the signal through a 15 foot wire coil to the track.  Engines will respond correctly in this situation.  It is when I move them to the track that I have issues.  I have one currently that will correctly respond 1 out of 7 - 10 attempts on my layout.

I have only one other engine live on the track, a Premier Yellowstone, and it is a pita to take on and off so I avoid that if possible.  I recently purchased and test ran a Bi Polar PS3 on my layout and it performed flawless.

I have two loops, I power each off variable power, and two additional blocks powered by the fixed channels.  They are my yard and my roundhouse.  I am using two ZW transformers for power and isolate the two blocks by powering down the handles.  I power switched, accessories and track on three switches on a barrier strip.  I turn on the switches, then accessories and lastly the track via TIU.

Then I power the main loop by using the remote and setting that to 18 volts.  The engine in question will turn on the headlight seconds later and then shortly after that power the sound system on.  I can then shut it down with the remote and re-power it  and it will function properly.  I only had the Bi Polar on the track recently but it sat through all the steps to power up the layout with out incident.

Any more thoughts or ideas.

Thanks

@Charly posted:

I have two loops, I power each off variable power, and two additional blocks powered by the fixed channels.  They are my yard and my roundhouse.  I am using two ZW transformers for power and isolate the two blocks by powering down the handles.

This is a RED FLAG!  Unless I don't understand what you're saying, this isn't going to work well.  You need to physically isolate the center rail feed from each TIU output.

Ummm, the O48 loop is variable 2 and the O72 loop is variable 1, the roundhouse and yard are fxd 1 and 2.  Each section of the layout is isolated from  each other.

Ok?

OK, it just sounded like they were not isolated.

So is there anything in my layout operation that would continually cause PS3 engines to start prematurely?

Nothing obvious without seeing more, it's obviously a DCS signal issue, but as to why, that's another story.

One thing that affects the DCS signal is anything that is connected to track power, signals, accessories, etc.

I do not connect any other item to track power, and I have all other engines in the roundhouse and off.  I do have two lighted cabooses.  But I should think they would be minimal in effect.

Try taking EVERYTHING off the tracks.  Some lighted cabooses can seriously degrade the DCS signal, it all depends on what lighting they're using.

@Charly posted:

... Any more thoughts or ideas.

Is it impractical or have you otherwise ruled out opening engines to examine the wire routing per Engineer-Joe's response?

Idea 1.  For the engine that properly starts only 1 out of 7-10, when you do gain control of the engine can you run a Track Signal Quality test?  This is the one where you want to get 9-10 rather than 1-2.  If you get a low or mid scale reading does it materially change if you move the engine to a different position in the block...such as right where TIU power enters the block (e.g., at a lock-on).

Idea 2.  Is it practical to swap the red/black power cables at the TIU output to a different output.  For example, if the block in question is being powered by Variable 1, can you swap the cables right at the TIU with Fixed 1; I figure it's only a couple inches apart so might not be too difficult to swap?  Thinking: there is some evidence that TIU channels can lose signaling performance in some environments.  Secondary thinking: even when operating in "fixed" mode, a Variable TIU channel has internal electronics between input and output.  OTOH, a Fixed TIU channel is essentially a straight shot from input to output.  So with the same reasoning discussed by others that DC power is the "best" environment for DCS signaling, it stands to reason that a Fixed TIU channel might have better DCS signaling than a Variable.  This latter thought may be in the noise...but you did ask for more thoughts!

@stan2004 posted:

Thinking: there is some evidence that TIU channels can lose signaling performance in some environments.

Actually Stan, there is a lot of evidence that the Rev. L TIU can indeed lose DCS signal strength in a number of environments!  I had one that dropped to 25% of it's normal signal strength just being used on my test bench.  I've repaired a host of TIU's for that very problem, it's actually quite common.  The issue is the 74ACT244 IC used in the application with the gates are paralleled, a transient spike can zap one or more of the four paralleled gates and reduce the signal strength.  Succeeding spikes apparently keep killing the gates one or two at a time until you don't have any signal.  Truthfully, I think the parallel gates are a mis-use of the 74ACT244, I don't see any evidence that is accommodated in the design of the chip.  MTH has even tried a couple of times to come up with a fix, and they've mis-fired both times.

@gunrunnerjohn posted:

Actually Stan, there is a lot of evidence that the Rev. L TIU can indeed lose DCS signal strength in a number of environments!  I had one that dropped to 25% of it's normal signal strength just being used on my test bench.  I've repaired a host of TIU's for that very problem, it's actually quite common.  The issue is the 74ACT244 IC used in the application with the gates are paralleled,...

OK.  This is getting in the techno-geek realm...

I'm not disagreeing with you!   However, the idea of paralleling the 74ACT244 has been used (to the best of my knowledge) since day 1 of the TIU.  What I'm thinking it must be something else about the design of the Rev. L that is the issue.  Heck, unless I missed it,  I/we don't know if Charly even has a Rev. L TIU?!

We're all entitled to our opinions, and I most definitely respect yours, but parallel'ing gates of logic IC chips like the 74ACT244 is a tried-and-true design technique.  I agree that it is something you probably won't find in a manufacturer datasheet...and surely something I figure they don't "teach" in college EE courses!

But to the matter at hand.  I am speculating, I repeat speculating, that Charly may in fact have 2 (or more?) things going on here.  1) the PS3 upgraded engines may have reduced/demoted DCS signaling performance, and 2) the TIU channel may have reduced/demoted DCS signaling performance.

I can understand to the reluctance to perform "surgery" on the engine to re-route harness wiring or whatever. Hence, if the only option is to mess with matters EXTERNAL to the PS3 engine then so be it.  IF (in capital letters) that is the case and the engine is not to be touched, then it seems to me the question is can the DCS "environment" be changed to work with the upgraded PS3 engine as is.

After some futzing around with the layout, we have determined that there is some signal anomaly in the main loop that the PS3 engines do not like.  They perform as expected on the test bench and on the second loop of my layout.  With that, we are going to endeavor a re wiring of the layout and try to determine what is bothering the engines.  The amazing part of all this is that all of my trains behave.  *shoulder shrug*  If we discover something, we'll let ya'll know.

Thanks

Cheryl

I find this occasionally happens to my engines when placed on the Variable TIU channels. It never happens when using the Fixed outputs.

When initially powered up, the variable channels will default to 0 volts, regardless of what they were left on when previously powered down. When changing the voltage after power up, a PS2 or PS3 engine sitting on those variable lines sometimes turns on by itself. I haven't been able to narrow down the cause any more than that. At least for me.

Could it have something to do with the engine thinking it is in conventional mode?

@VonFrank posted:

When initially powered up, the variable channels will default to 0 volts, regardless of what they were left on when previously powered down. When changing the voltage after power up, a PS2 or PS3 engine sitting on those variable lines sometimes turns on by itself. I haven't been able to narrow down the cause any more than that. At least for me.

Actually, you can set the variable channels in fixed mode and they come up full voltage.  Also, that's a good operating mode as to run in variable mode, all you have to do is select the track and vary the voltage, the channel switches to variable mode.  When it does the switch, again it outputs the watchdog signal.

Yes, that may be something?

I always set my channels to fixed as I run all in command mode. I have both version "L" TIUs and an old one.

( I swapped it out years ago and forget which version it is now?)

So..... I don't get to experience some of these bugs? Lucky me?....

(for a change).