LM-4S-G Erratic behavior/Intermittent "Cannot Connect"

[jordanhd87]

I have no answer to your problem, but think it's 99% your track power, intermittent voltage from the track probably effects the Rail Pro module to function as it should.  If it works on battery power 100% of the time then you have your answer, just my personal thoughts on this. I know what I would do if I had a track powered layout outside, change to battery power and be done with it, no track cleaning and no track power connections to deal with. 
trainman

If that is the case, then Ring has designed these modules poorly and should not market this unit as being able to run off of track power. I would not have bought RailPro if I knew it couldn't run reliably from track power.

The battery topic is a whole different animal. The way I run and the duration I run, I would have some serious concessions going battery power. I am well aware and well versed on keeping track connections clean and in good conductive shape. I have tools to clean the track thoroughly and easily. Neither one of these things are a drawback in my particular case. However, constantly having to charge batteries, have limited runtime, spend another $100+ per locomotive on batteries, (not to mention finding battery solutions for running rolling stock lighting) are all big drawbacks for me.

The USAT power supply should be here tomorrow. I'll test with it and see if it fixes my issues.

[Gibs]

Ok, just humor me and the rest of us, just hook up a battery temporary to your worst loco and see if it makes any difference. Its just that what you describe your issue as smacks of a power outage glitch, so lets just eliminate the easiest one to check first.

PS: I didn't see the posts above. Just a bit of FYI, Brass is in fact a terrible CONTACT conductor, why, because it will tarnish right before your eyes straight after giving it a good clean/polish. Internally brass is fine, but surface to surface not so flash. Its why its not used in any switching capacity in electrical equipment. More expensive Silver and Gold is used on contact surfaces because it will not Tarnish. Its one of the reasons guys into G Scale go Battery rather than track for power, since brass track is cheaper than whats needed, Stainless Steel Track for outside. Even Aluminium has the same problem as Brass but faster yet again.
Anyway try the new power supply but I'd still be keen to see if a battery powered loco has the same issues. If it doesn't then you will know 100% that its NOT power related but rather a Signal issue with either the Modules or the HC.
Is there any one who has a RailPro set up living close by, you could use their HC to eliminate that as well. They don't even have to be into G scale.

[gregeusa]

I have G scale track power outside. I do have stainless steel and I am DCC.

Long ago, DCC decoders would go nuts with the smallest glitch in power, resetting the microprocessor in most cases.

That was 10 years ago, modern decoders no longer have that issue, having internal 5 or 3.5 volt power supplies that are designed to keep the micro alive during short glitches.

The addition of "keep-alive" capacitor banks also help really bad power pickup issues, but I have about 40-60 locos and not one has a keep-alive.

Sounds like the RP unit is sensitive to power glitches, and forcing someone to battery is like turning the clock back 10 years in design. With the overall cost of the electronics, this is an issue that can be easily and economically solved and HAS BEEN solved with other systems.

(It still needs to be proved that the behavior here needs to be "proved" that it is coming from power glitches, maybe the voltage being higher is causing the issue, we had the similar thing in G scale where track voltage is often 24 volts and many DCC manufacturers that "came from" HO had issues with the higher voltage, like SoundTraxx)

Anyway, there is no reason to "force" someone to battery for a deficient design (to be proved).

Greg

[jordanhd87]

Greg - I completely agree. For what it's worth the LM-4S-G manual states maximum input voltage is 28v, and in parenthesis, absolute maximum 30v. I am well below those values @23.2 volts. Admittedly I do not have the equipment to test for transient voltage spikes, however my brother has access to an oscilloscope which we are going to look at the output on my current power supplies and see if we spot anything odd with the waveform. Also, the USAT supply is "out for delivery" so I should have that in my hands today for testing.

[jordanhd87]

Hi All,

Alright I got the USAT train power 10 supply and am bench testing my locomotives on it. RP fails to connect 75% of the time if I set the voltage output to 19-20 volts. If I go full power (25v) they were all connecting. Also, reading the voltage input on the handheld, it was right at 24v (to be expected with a small drop from lights and fans running). As I was resetting my full motor current values on all of my locomotives (being that the voltage is higher), after I got it set on my last locomotive I tilted it sideways and put it back on the track. It would not reconnect. I even disconnected the track leads from the power supply and connected them to a 24v battery, and it wouldnt even connect if powered from the battery. Waited 5 minutes, tried powering it again from the battery, and it connects. At this point I don't know what to do other than pull all my modules and send them to Tim.

[Gibs]

Greg
You really didn't get one bit of what I was saying........no surprise there.
Testing issues is a case of elimination. Always start with the easiest and most likely suspect. In this case using a battery to TEST a difference in Power type (Track V's On Board) was/is quick and easy to do and cheap.
Instead hes gone and brought an expensive new Power supply (his choice) that's checking....TRACK POWER only. In the end a battery was still used........I wonder why.
And what has he learnt at the end of the day, well its was more expensive to get a new power supply AND a battery to find out it has power issues ......... from track power as well as battery ......... Silly me.

Jordan
I'm wondering if the Module has an internal CB with an auto reset feature (that's slow to reset) and its this that keeps playing up. I'd suspect it to be a firmware issue if it is and may be a bug from the last update. This sorta thing does happen often with updates, breaks one thing while fixing something else.

Its Ring Ring Time....again.

PS: Was it just that one loco, the last one that played up, or are all the other RP equipped Loco's still doing the same as well.

[jordanhd87]

Gibs, it is all 5 of my Railpro locomotives that are having connectivity issues. Sometimes one will act up, and the others will test OK. Sometimes a different one will act up, and the previous one will test OK.

Anyways more info after further testing - out of a last ditch effort I have disconnected all of the cooling fans in 3 of the 5 locomotives to take out the possibility of EMI from the fans. They now connect to the handheld every single time. If I use my old bench supply, it still fails connection sometimes. If I use the USAT supply (even at the reduced 19-20 volt setting), it connects every single time on all three of these locomotives. It seems there may have been a dual fold issue - RP being sensitive to certain power supply outputs and well as EMI from cooling fans. I just hope the overheat issues are indeed fixed with v1.05 and I won't need the fans in the summertime. Anyways, they are now passing the test on the bench so I will disconect the fans from the other two and then outside I will go for the next round of tests (when the weather clears up).

The fans causing that much EMI does boggle my mind a bit, only because I know Robby at RLD puts a fan on everyone single one of his installs and he has not encountered this issue yet.

[Gibs]

NOW you tell me you have fans installed....lol. Best cause of RFI is a non suppressed electric motor - its after all one great big spinning magnetic field making thingy (technical term there).
"So why doesn't the on board Loco motor cause issues", because the power supplied to the drive motor is pulsed, Cooling Fan's, eh not so much. The usual cheap and cheerful fix would be to install an RF Choke onto ONE side of the power feed to the suspect electrical motor (fan). You can also use a condenser from an automobile Alternator (same thing really).
Not gonna say its a slam dunk fix, but again its quick, cheap and easy to do.

As to WHY the Module is having an issue with RF, well its also a very small low powered non shielded RF Tx. You could try shielding it, but there is no Antenna so........
If you did want to play around with that just use some alfoil, but then that would negate getting cool air from a cooling fan onto the module, so no RFI to begin with, bit like the Egg V Chicken.   lol

[jordanhd87]

Gibs - check my first post, I mentioned that I had brushless fans installed. That's OK though, I put a lot of details in that post (and subsequent posts), it's tough to keep up with them all!

Anyways, it seems I'm headed the right direction. Next tests will involve going outside. I will test on my existing power supply first, and if I have issues then I will test on the USAT power supply. Due to all of the headaches around this, Ring should mention "DO NOT INSTALL fans" in their install instructions. I even mentioned to Tim that I had them installed and while he didn't think it was the issue, he did recommend disconnecting them because he didn't think they were needed due to the software update.

[Gibs]

*Fingers Crossed*

Lets just hope its something this simple and stupid. (I'd believe stupid).

[G8B4Life]

Just a gentle reminder for all members to be mindful of what you write. Something that was written in this thread could very well be construed as insulting, which would then be flaming which is not tolerated.

That said, I'm surprised that a modern fan would give RFI grief, that's why I didn't call them out in my earlier post.  I have various fans here, I might have to try an experiment. Also, reading back though the posts, a 3 volt drop isn't good. Something is not right as the Crest is supposed to be regulated.

- Tim

[gregeusa]

Greg
You really didn't get one bit of what I was saying........no surprise there.
Testing issues is a case of elimination. Always start with the easiest and most likely suspect. In this case using a battery to TEST a difference in Power type (Track V's On Board) was/is quick and easy to do and cheap.
Instead hes gone and brought an expensive new Power supply (his choice) that's checking....TRACK POWER only. In the end a battery was still used........I wonder why.
And what has he learnt at the end of the day, well its was more expensive to get a new power supply AND a battery to find out it has power issues ......... from track power as well as battery ......... Silly me.

Jordan
I'm wondering if the Module has an internal CB with an auto reset feature (that's slow to reset) and its this that keeps playing up. I'd suspect it to be a firmware issue if it is and may be a bug from the last update. This sorta thing does happen often with updates, breaks one thing while fixing something else.

Its Ring Ring Time....again.

PS: Was it just that one loco, the last one that played up, or are all the other RP equipped Loco's still doing the same as well.

Gibs, You don't read so well, read the last sentence in my post...

I completely agree to test with a battery, and eliminate the power supply and pickup contacts.

Greg

[gregeusa]

Just a gentle reminder for all members to be mindful of what you write. Something that was written in this thread could very well be construed as insulting, which would then be flaming which is not tolerated.

That said, I'm surprised that a modern fan would give RFI grief, that's why I didn't call them out in my earlier post.  I have various fans here, I might have to try an experiment. Also, reading back though the posts, a 3 volt drop isn't good. Something is not right as the Crest is supposed to be regulated.

- Tim

I would be extremely surprised if the fan has brushes... it must be brushless. In any case, the RFI from the locomotive motor itself would be orders of magnitude greater (caused by the much higher current, and it indeed has a commutator and brushes)

By the way, RFI is not only radiated, but can be conducted, i.e. the high frequency noise is on the motor wires also. If you think about it, that is why inductors and capacitors are used, for conducted RFI... they do nothing for real radio waves emanating from a device.... metal shielding is the only thing that works for radiated RFI, conducted RFI needs chokes... (don't get me started in the magnetic fields ha ha)

Greg

[jordanhd87]

Hi Greg, yes I fully suspect the RFI I'm experiencing is most likely the "conductive" flavor, at least for the locomotives I have shell fans for. My shell fans are drawing from the track power inputs, go through a bridge rectifier (so the fans always turn the same direction), then to the fans. The power going into the RP module comes straight from the track inputs, pre-diode bridge rectifier. I think the shell fans are putting noise on the power input side of the RP module. While the above seems to be true for the 24v fans, some of my locos had 5v fans connected directly to the 5v outputs on the RP module. And even these ones were having communication issues. Explain that one to me...

[jordanhd87]

Well so far so good. I ran two locos tonight for about 2 hrs. No issues at all. Good connectivity (well, as good as it can be for RailPro). This was using my Crest switching power supply as well. The voltage drop was only about 1 to 1.5 volts while the trains were running. My initial reading of 3 may have been erroneous on my part. So far so good. More running sessions needed to see for certain and to gain my trust back.