Recent voltage drops.

[Frenchie]

HI !
My layout, the BNSF "Desert & Sierra Sub", Is now virtually completed. I chose to stick with DC and to use Blunami decoders for my locos. I already installed one
in an Atlas GE U36C.
Everything went well until three days ago where I faced sudden deconnections. The loco stopped, no more move nor sounds.
I just moved it a tad and it came back to life, showing "reconnecting" on my smartphone screen. Ok, no sweat until it happened again a couple times.
I tried with my wife's phone, not better. Following the on line counsels, I uninstalled the app and switched off my phone over night.
Next day I rebooted my phone and reinstalled the app. I also cleaned thoroughly the tracks, it was ok for a day and the problem came back.

I contacted Soundtraxx by mail. Very quick answer, thank's to them, It could very well be voltage drops in certain areas.

Track feeders are 22 AWG wires soldered to the rail joiners every 2 lengths of tracks. More where turnouts are present.
Extensions between feeders and main bus are 14 AWG solid wire. Main bus is stranded 12 AWG cable.

I DON'T use the power for accessories or anything else than running locos.

Originally the layout had one continuous big bus ( 4 mm2 = 12 AWG ), I cut it to make 4 power districts, each of them powered by MeanWell 6A, 13vdc
power supplies. In two districts the power supply feeds at the same time the upper and the lower level.

I think it could be the problem, too long a length of tracks, eventhough the problem occured not far from the connected power supply.
Well, I plan to double, wherever possible, the 22 AWG feeder wires, but more important I plan to modify the actual power district in 2 smaller ones, that is
1 for the upper level and 1 for the lower level, which means 2 power supplies instead of 1. I will add another dedicated main bus for the upper level.
The existing main bus will then feed "only" the lower level.

I'll do that in other areas, I should then have 7 to 8 districts instead of 4.

The other option is to leave it as it is, and just increase the power supplies voltage, I am not sure it would be a good idea...

What do you think ? any comments or suggestions are very welcome.

I attach two drawings showing, on one, the actual district wiring, and on the other, what I'd like to do.

I thank you beforehand.

 

[Canalligator]

You seem to be speculating that you have low track voltage;I don’t see any references to track voltage measurements. If you haven’t put your voltmeter on the tracks, you should do that first.

 

[Frenchie]

You seem to be speculating that you have low track voltage;I don’t see any references to track voltage measurements. If you haven’t put your voltmeter on the tracks, you should do that first.

Sorry I forgot that. I did check with my multimeter several areas and I didn't notice anything weird. But on my screen there is a direct reading of the actual voltage as you run the loco. I noticed variations during operation that could go from 13,5 v to 9 V.
And suddenly it would drop 5, 4 or 3 v and the loco loses connection.

 

[Canalligator]

Almost sounds like dirty track or wheels. Is that voltage a measurement made by the decoder and reported back to your phone? (I don’t know anything about Blunami.)

 

[Frenchie]

Almost sounds like dirty track or wheels. Is that voltage a measurement made by the decoder and reported back to your phone? (I don’t know anything about Blunami.)

Yes I am sure it is. Actually I don't see what else it could be. When I mentioned that to Soundtraxx, they didn't contradict me.

 

[flyboy2610]

Are the rail joiners soldered? Could be loose rail joiners causing a problem. Solder 2-3 lengths of track together and feed each section with its own wire.

 

[Frenchie]

Are the rail joiners soldered? Could be loose rail joiners causing a problem. Solder 2-3 lengths of track together and feed each section with its own wire.

As I said previously, I solder the track joiners every 2 lengths of track, and feeders are also soldered to those same soldered joiners.
This way these 2 track lengths are powered from the center.
The alternate joiners aren't soldered to allow for expansion. So even a slightly loose joiner wouldn't be a problem.
Peco code 83 joiners are very tight since they are also made for code 70. With painted and ballasted track, they don't move.
Naturally there are more feeders where turnouts are present.
So you suggest to drop feeders every foot and a half, track pieces being three feet
long ?

 

[kjd]

So it used to work ok and then stopped working? If so, something changed and just adding feeders won’t help. Can you measure current? You mentioned it happens even near power supply connections, that doesn’t sound like voltage drop due to wiring deficiency. How long did it work well? How long are the wire runs? Are the loco wheels clean?

 

[Frenchie]

So it used to work ok and then stopped working? If so, something changed and just adding feeders won’t help. Can you measure current? You mentioned it happens even near power supply connections, that doesn’t sound like voltage drop due to wiring deficiency. How long did it work well? How long are the wire runs? Are the loco wheels clean?

As I explained previously :
- Measured track voltage ok in the all district, upper and lower level.
- Voltage between 13 and 14VDC.
- Wire size ok. 22, 14 and 12 AWG.
- Track and feeders soldered every 2 lengths of track.
- Track and wheels cleaned thoroughly ( not each day ).
- Loco equiped with current keeper loading properly.

Well, loco stops suddenly, "disconnect" appears on screen. Then I just move it a tad by hand and connection is back.
I previously described the voltage fluctuations I read on the screen.

I sometimes run the loco for almost an hour without issues and it happens. Then it's ok again. Another day it will happen 2 or 3 times in a row in a very short period of time. Then it's ok again.
Like you I suspect to long a length of track.
There is probably about 80 feet of track in this district, that’s quite a resistor.
That’s why I want to make the change you can see on the drawings attached with my first message, that is cutting the district in 2 shorter ones.
For the entire layout I would then end up with
7 to 8 districts instead of 4.
Thank's for the helpful reactions.
Any more comments and ideas are welcome.

 

[bnsf971]

Does it happen more when the engine is at its furthest point from the power feeds? If yes, suspect the track/wiring. If it does it anywhere, suspect the engine.

 

[Frenchie]

Does it happen more when the engine is at its furthest point from the power feeds? If yes, suspect the track/wiring. If it does it anywhere, suspect the engine.

No, I didn't notice any change linked to the distance between the loco and the power supply.
But these voltage fluctuations I can read on my screen while the loco runs are weird to say the least.
I'm gonna follow your suggestion and check thoroughly the loco. It could very well be the issue since connection is back as soon as I slightly move the loco by hand.
I don't want to drastically add more districts, hence more busses and power supplies without being certain it would be the right decision. Thank's for the suggestion.

 

[bnsf971]

No, I didn't notice any change linked to the distance between the loco and the power supply.
But these voltage fluctuations I can read on my screen while the loco runs are weird to say the least.
I'm gonna follow your suggestion and check thoroughly the loco. It could very well be the issue since connection is back as soon as I slightly move the loco by hand.
I don't want to drastically add more districts, hence more busses and power supplies without being certain it would be the right decision. Thank's for the suggestion.

The voltage fluctuations are from the motor getting, then losing contact. When there's contact, the load of the motor will lower the voltage.

 

[Frenchie]

Sorry I am obviously no electrician.
Would that mean that for some reason the motor load could lower the voltage to the point that the loco stops for lack of power and disconnect ?
When I gently move it a tad it reconnects and the voltage is back to normal, and it stays so as long as the loco idles.
Would the motor be the problem ?
I never experimented any problem with this loco, an Atlas GE U36C.
Thank's for your comments.

 

[diesel]

If you're measuring voltage at the controller, a major drop in voltage doesn't mean circuits are opening (bad joints, dirty wheels, etc.). It means increased current is causing the voltage to drop. So things like wheels shorting out in turnout gaps... controller/power supplies are made to reduce voltage when the current is excessive, to protect the controller and/or wiring. In other words, it sounds to me like you have shorts not opens.

Or even, locomotives that are mechanically binding and putting excessive load on the motor. But if the problem clears on straight track (i.e. not on a crossover or turnout), then the normal variation of voltage due to loss of contact, may be greatly increased if the controller is for some reason unable to supply the normal amount of current. So you should check the controller wiring and also the wiring that powers the controller.

What was mentioned about measuring voltage in many places, is especially true, when troubleshooting. If you see roughly the same measurements everywhere, then you need to look back at the controller and its wiring.

People who are electricians, like to see numbers, too.

 

[diesel]

My loose rule of thumb for feeders is, not less than one every 1.0 meter (~3 feet), and at least one on any unbroken section of track (no matter how short).

 

[Frenchie]

Thank's for your reaction.
Since I use a Blunami decoder, I have no controller, my phone controls everything and send signals to the decoder using Bluetooth.

BUT I have DC power supplies, one for each district. 4 so far, they deliver 12/14VDC at 6A.
So I agree that there could be an issue somewhere and I'm gonna check that.

I also think that the loco motor could very well be the culprit.
As soon as possible I'll install a decoder/current keeper in an other loco and see. I'll then come back to you guys, hopefully with good news.

Attached, one of my power supply. Cables are 14 AWG.

 

[diesel]

Those supplies have excellent voltage regulation. So a comparison of voltage at the trouble track and back at the supply, could be a big clue to what is going on. A problem with poor conductivity like long tracks, bad joins, etc. will make them different - they would only be the same in the case where more than 6 Amps is being drawn from the supply. Some supplies have an overcurrent LED or beeper to indicate that situation. Actually, the most useful indicator you can have on a supply like that, is an ammeter.

 

[insomniville]

I also cleaned thoroughly the tracks, it was ok for a day and the problem came back.

This stands out to me. Have you tried cleaning the tracks and wheels again, and had it work for another day?

If so, you might read through this thread. The post that stands out to me talks about stuff called NO-OX-ID that is supposed to work miracles. There's a lot of info there, though.

FWIW.

 

[kjd]

It works fine sometimes and then stops working well? Touching the loco "fixes" it? I'd carefully examine the loco, I don't think it is the track.

 

[insomniville]

It works fine sometimes and then stops working well? Touching the loco "fixes" it? I'd carefully examine the loco, I don't think it is the track.

Touching the loco could also result in moving the loco very slightly past a bad spot, which is the impression I had of what he meant by that.

As for intermittency, there are variables like the exact speed of the locomotive and such that can affect whether a problem shows up. I would think that if it's moving fast enough, a short spot of dirty track won't be a problem. The slower it's moving, the more likely it is to react to every bad spot on the track.

On a more general note, my typical troubleshooting methods would result in taking said loco and running it on "known good" track, like maybe some spare Unitrack or something that is known to function properly. If it works there, and not on the layout, then you know it's something about the layout, otherwise it's the locomotive.

JMHO.