Major DCC issues.

[Scale7JB]

Hi collective. and @Oz7mm

Had some more major testing yesterday on Love Lane, track is getting really good now, with a 28 wagon train traversing through both crossovers quite happily and at a good speed too. I think the curves are too tight to propel suck a long train, so even i will admit defeat on that one.

We did have some very strange DCC issues yesterday though. Its almost as if the layout is haunted, and it 'Doesn't like Diesels'.

It all started with a JLTRT Class 37. As it travelled through a particular turnout, it caused that particular turnout and the following one on its set path to change themselves, causing a major derailment as the first few wagons decided to take the other line. Now this happened religiously. We tried this with a few steam locos and all fine. We then tried a Heljan Falcon, and Heljan CL26, LLC Class 15, all the same, causing the points to change. It genuinely had to be seen to be believed.

A few points to make are that

1) It only seemed to happen when the loco was pulling a heavy load (although the J39 was pulling the same load and didn't cause a problem).
2) If the train was travelling in a different power zone it didn't cause a problem.
3) We are using the Megapoint DCC servos/boards and frog juicers for the crossing see power supplies

Might this be caused by the loco asking for too much power?

Now, a second issue is that when a diesel is travelling from the operating section onto the scenic section (powered by a second booster just for the scenic section, it stops dead, and resets itself (sound goes off, and engine restarts etc) but this is happening without a heavy train, but again a steam loco sails through with no issues. Could this be a power load issue again ?

Any thoughts and suggestions would be welcomed in this frankly bizarre issue.

JB.

 

[richard carr]

JB

The first thing to say is that it is not a power load issue. Even the Heljan locos running light at modest speed only draw 300 mA.

Are the points being DCC controlled, if they are this needs to stop, I thought there was a separate control panel for the points ?

As for the transition between the booster zones, I will check the wiring this evening. I was running mine at the weekend and had no problems running locos (all diesels) in and out of the fiddle yard.
With the new booster that Robin got recently is it running at the higher track voltage ?

Thats all that springs to mind for now.

Richard

 

[Scale7JB]

Robin had installed a DCC controlled points board but this was disconnected from the network last night but didn't make any difference. However the module was NOT disconnected from the track power supply.

Somewhere there is still a connection though, as if a short occurs, some of the servos do reset themselves.

We concluded that it wasn't overpowering the system, although the Cl15 was stuttering, and the problem didn't seem to occur with a lighter train.

JB.

 

[Scale7JB]

I couldn't tell you if there was a voltage difference. I have been firmly concentrating on track rather than electronics so I just don't know enough about the system to comment.

JB.

 

[Yorkshire Dave]

I would go back to basics rather than fault find with everything connected. Apologies for the simplicity and if is already been explored.

Electrically disconnect each board from each other. Power up board 1 then test (using a diesel). Add board 2 then test again and so on until the fault occurs. Then you know the fault is in that board's wiring.

From the descriptions it sounds like the longer wheelbase diesels or rolling stock are inadvertently completing a circuit. I would re-check the point wiring (including frog polarity and switches). In addition I would also re-check all rolling stock wheels to make sure they are insulated from each other and the chassis/body.

 

[Scale7JB]

Thanks Dave, only issue is that the problem only occurs when the diesels are pulling heave trains. Light engine they don't seem to cause any problems..

JB.

 

[Kev T]

A couple of thoughts spring to mind.
Due to the higher current with Heljan diesels the problem may be caused by EMI. Try to move the offending servo leads as far away from track power as possible. If the leads need to cross the track power make sure they cross at 90degrees and try to keep them at least a couple of inches away from the track power.

2nd thought is that Dave Fenton (Mr Megapoints) is very helpful and will definitely help if you call him.

The problem with crossing power districts; are you absolutely positive that the phasing of the DCC is the same on both power districts?

kev

 

[Dog Star]

Seems to me that the crucial question here is what is initiating the movement of the switches... by that I mean:-

* an unexpected DCC control signal is reaching the controller board of the servo and causing the switches to move;
* a stray DC current is getting to the servo without any corresponding DCC signal to the controller board;
* an interruption of the power supply to the controller board causes the controller board to reset.

I shall be surprised if a correctly formed and spurious DCC signal is causing the controller board to take a rogue action.

The switches that move together - connected to the same controller board? to different controller boards? Do all of the switches connected to the controller board move unexpectedly (or just two)?

 

[Steph Dale]

It sounds to me as though your circuit bus(es) isn't properly terminated.

I'll have to find the values, but there should be a resistor/capacitor 'snubber' circuit on the end of each bus. It stops the DCC signal bouncing around the system and generating spurious commands.

Steph

 

[Nigel Cliffe]

Robin had installed a DCC controlled points board but this was disconnected from the network last night but didn't make any difference. However the module was NOT disconnected from the track power supply.

Somewhere there is still a connection though, as if a short occurs, some of the servos do reset themselves.
.

Is there any commonality in the power supply ? I've seen a voltage spike on a DC power supply cause Megapoints boards to reset. In that case it was some uncoupling electromagnets which were installed without a diode over them. On going to "off", the spike would cause the Megapoints boards sharing the same DC supply to fail. It is an 18Amp supply, so not short of "grunt". Fitted the diodes and problem was solved, and good illustration of why the diode is needed!

I'm not using frog juicers on the above layout, its relay boards instead (12v units bought assembled from Amazon/Ebay). Some switched directly from the turnout switches, on a separate line to that to the Megapoints boards, and some switched as slave boards to the Megapoints servo boards. Relays don't risk short circuit spikes on the track/DCC signal, unlike the frog-juicing method.


- Nigel

 

[Steph Dale]

Snubber circuit component values: 2.5w 150R resistor and 100nF / 0.1uF capacitor in series across each bus end. For a sake of a the few pence these things cost they're worth a try and likely to be successful.

Steph

 

[simond]

Servos are known to be sensitive to noise, some worse than others of course, but that does not account for both ends of a crossover changing - I’m sure that’s got more to do with supply side issues - either a brown-out, or noise.

The diesel resetting itself suggests that there is a polarity (phase) issue at the interface between the districts. Easy check would be to be running another loco on one or other district as your diesel crosses the boundary. If the booster shuts down, and resets, (the other loco will stop) that’s the likely issue. The diesel differs from the steam loco in the “live wheelbase”, assuming the bogies both pick up from both sides, ie not the “American” system, so the booster(s) would be experiencing short circuit issues for a more prolonged period. Are the locos fitted with keep-alives?

I tend to view Juicers as a very heavy hammer for a rather small nut, but I don’t think they’re relevant in this case.

Hope these thoughts prove helpful
Best
Simon

 

[Steph Dale]

Actually, Simon, if it's phase there's an easy test.

Connect the track bus(es) to an analogue supply and drive a diesel loco across the track that's causing problems. If it stops you can at least work out the locality of the problem. All you've got to do then is test whether it's a short- or open- circuit.

Steph

 

[simond]

Steph, yes, easy if you have an analog option, and that could check the power district boundary, but the Juicers won’t work.

I think having another loco crawling (and maybe stopping) will provide ample evidence!

Best
Simon

 

[Boyblunder]

Here is the latest news. Ran the same stock and locos all evening with no sign of the turnout problem recurring. Richard’s JLTRT 37 suffererd a motor problem from the weight of stock it was pulling so I think we gave it a good test. The scenic side did display issues on both up and down tracks with locos stopping on the power district boundary at one end, but not the other end. Also the PSX1 was indicating a short briefly while locos were in the middle of the scenic section and nowhere near the power district boundary. After restarting the DCC system everything worked fine. Tried reversing the DCC phase again, made no obvious difference. The intermittent short circuit problem stopped after Richard fixed the ‘37, but this may be a red herring because I also tightened the DCC bus connections. After this everything ran smoothly for the remainder of the evening.

David Rayner had a look earlier today and made the same suggestion as Kev that the DCC bus wires need to be further from the servo leads, I will do that tomorrow.

I’ll also follow Steph’s suggestion to terminate the busses. I have the components, just not bothered to fit them because we didn’t have a problem until yesterday.

The Megapoints boards, uncoupler electromagnets fitted with diodes by David Raynor, and DCC system all have separate power transformers (from inception to avoid issues just such as Johnathan experienced). But at the moment everything is running off one mains socket via various extension leads to replicate what we will have at Warley.

If it happens again after I’ve moved the DCC bus and fitted the terminations I’ll do as Yorkshire Dave suggests to isolate the fault, unfortunately it’s very big layout with loads of wires to test so we will maybe finish testing by Christmas?

Thanks for all the helpful suggestions,
Robin

 

[SimonT]

Is the need for the districts lead by the need to run slave DCC power supplies to the track? Aberbeeg started with two districts with a slave feeding the lower half. We got the odd stall on the junction of the two districts. I sat down and totalled the load throughout the sequence and realised that we only needed the one LZV100 command station. Might be worth a look?

Simon

 

[simond]

Robin,

your comment "Tried reversing the DCC phase again, made no obvious difference." is most puzzling.

If the DCC phases are not identical in adjacent districts, I would expect any loco (but particularly a long one equipped with keep-alive) to cause a short when crossing the boundary, although, potentially you could feed a whole district via a reverse loop module, or indeed, the booster could incorporate one, I suppose. The fact that it caused an issue at one end but not the other suggests that the first FY is reversed.

There are two ways to reverse the phase - you can reverse the connections of the control bus (C&D in Lenz) or you can reverse the track connections.

I can see no good reason to have a reversal between C&D - the command station C should be connected to all the boosters' C and likewise D to all the Ds, and therefore I suggst you check the track feeds. Again, being Lenz equipped, my mnemonic is K is blacK connected to the bacK. Your mileage may vary, of course!

Simon T's note suggests a test which may be easy to carry out - disconnect the track feed from the FY that is suspect, and connect it to the output of the scenic section booster - temporarily. See if the district boundary problem goes away. If it does, you have identified an issue with the booster or wiring on that FY.

The fact that an observed fault is no longer observed would worry me. It will, I fear, rear its ugly head when you least want it.

watching with interest
Simon (another one)

 

[Boyblunder]

Simon,
Be assured that the fault no longer observed is really worrying me!

We are using a NCE system with a CS02 control box and 10 amp PB110 booster feeding the 6 road storage yard and curves through a 4 zone Zoneshare protection box. The scenic section is powered by a PB5 booster feeding a PSX-1 circuit protector with 58' control bus cable from the CS02. This amount of power may seem like overkill but we had at least 10 steam locos and 5 diesels on the layout Wednesday, that's not up to the full compliment yet. When we were running the late, much lamented Steve Baldock's original Smoke & Steam 4MT we found the power consumption of that one loco triggered the PSX-1 set at 2.5 amps several times and we had hoped to have several of those on the layout at Warley. The set-up is the same as Richard Carr's where it is coping well with his massive storage yard and lots of big locos.

I 'm not an electronics engineer and don't have previous knowledge of these systems so we have just followed the NCE instructions and used "Wiring for DCC" as the bible, although both Steph's point about the bus termination with a snubber circuit and Kev's point about running servo wires close to DCC cables had not been implemented, the first because most of the zones were't quite long enough to suggest we needed them although I bought the components just in case and the second because in full panic mode to complete it in time I'd just slung the DCC bus feeds up willy-need ignoring the EMI issue.

Getting the phase right on the Scenic side is an issue that I don't think I fully understand yet. The NCE control bus terminals aren't labelled, not a problem I thought because the control bus is a pre-made RJ-H (4 wire) flat cable that can only be fitted one way round. The track outputs on the PB110 & PB5 are not labelled so I followed the NCE instructions that say "Wires from these terminals go to the track. If more than one booster is connected to the layout, be sure the left track terminal on all boosters is consistently wired to the same rail. This will ensure you have the same “phase” as you cross power district boundaries." Have to say I was culpable of crossing over the phase as it entered and left the PSX-1 before we added the PB-5 booster, a fault which Richard or Jonathan spotted months ago and we didn't have a problem after that until this week. Jonathan changed the polarity of the track feed from the PB5 on Tuesday and thought it was better.

It ran fine on Wednesday evening and Thursday morning. We dismantled most of the layout yesterday to make storage braces for the yard boards and to allow me to check all the wiring connections, fit the snubber circuits and tidy up the DCC bus so we won't be able to try it again for some days. I'll report back when its running again. Thanks everyone for all your suggestions.
Robin
Robin

 

[Boyblunder]

Latest news on the frustrating DCC issues. The track was put back together complete with snubbers last night. We immediately ran into 2 poor connection issues, one caused by my "rationalisation" of the bus wiring at the back of the layout which I discovered I hadn't finished on one zone (late night). The other caused I suspect by someone using corrosive flux for wiring joints, we are going to have a lot of trouble with that on some of the boards.

The good news is that there was no sign of the points changing under the influence of a loco running over the frogs. I'm hoping moving the servo wires as per Kev's advice and fitting the snubbers has eliminated this. The bad news was we were still suffering from visible shorts as locos crossed the boundaries into the PSX-1 protected zone. I hadn't used a DC loco to test it as suggested because I didn't have a DC loco or transformer handy and I'm not sure it would have worked with our equipment, but had very carefully checked all the wiring polarity with an ohmmeter. This time Jonathan used a steam loco with no stay-alive and it was very clearly happening at both ends, which was not clear with the giant stay-alive equipped diesels before and this had led to some of the head scratching. Followed advice on the NCE technical blog to check for reversed polarity with an AC voltmeter across the railgap between 2 zones, no volts shown so apparently polarity was not the problem. After more tracing of wires with no obvious errors found I tested the polarity into and out of the Zoneshare unit, bingo! The polarity reverses between the 2 way input connector and the 8 way output connector. I've scoured the Zoneshare manual, their website and other blogs but can't find any reference to this "feature". The dip switches are correctly set. After reversing the wiring from the PB5 to the PSX-1 it now all works OK, except the wiring looks incorrect compared to the NCE manual.

I think the no volts reading across the gap when the polarity was wrong may have been due to a faulty meter, its about 3 years old from Aldi and I have been getting funny readings lately so I'll try that test again with a better meter one day when we have more time.

I've emailed CVP Products to ask if the internal phase reversal is a standard feature or if we have an incorrectly wired unit, will let you know when they reply. In the meantime please all keep your fingers crossed that we have solved the problems and thanks very much for your suggestions.

 

[richard carr]

Robin

Thats great news


Richard