Hornby Elite - Normal behaviour?

[Brew Man]

With the new layout build underway I've been running a couple of trains round the two principle circuits. I notice that after a while the loco address disappears from the display on the Elite leaving only the clock visible. Everything still works OK, but it's a bit disconcerting. Is this normal behaviour? And what if I want to set another loco in motion, (haven't tried that yet) and I can't see what address I'm dialling in? I've got a vague recollection of this happening before but can't be sure.

Any help appreciated.

[96RAF]

Not something I have ever seen. The display on mine v1,45 stays bright with the last selected info. If I then select the other channel that data comes up and stays. The only time the display changes is if any of the main buttons either side of the display are invoked.

Unless there are favourites or named locos and/or points you want to keep then I would try a unit reset.

[Brew Man]

Yes I might try that. I did an initial test some time ago and everything passed OK. I've only got my locos' addresses configured at the moment so no big deal to set them up again. It's quite odd.

[Chrissaf]

I've had an Elite for many years, your observation is not something I have ever seen myself.

[Brew Man]

Maybe I can patent it grin. I'll try a reset and see how that goes. If you two haven't seen it before, it must be unusual.

[Chrissaf]

I forgot to mention that my Elite is still on version 1.44 firmware, not 1.45

[Brew Man]

Right. Rob is on ver 1.45 the same as me so unlikely that is the cause. Before I do a reset I will do another start-up test, though as I said, that was fine when I ran it before. I haven't timed it but it seems to take about 15 mins from switch on to when it loses the second line of the display

[ColinB]

Could it be something wrong with the display, as it warms up it starts to lose areas of the display. You could try seeing if you change menus whether that of the display gets written too, that would at least prove it is not software.

[96RAF]

Possibly a zebra connector problem. Look at my Elite Teardown page to see how to get at this.

Essentially a zebra connector comprises consecutive conducting and insulating strips of rubber in an assembled strip. This makes the strip non critical for placement twixt the Elite daughter mobo and the LCD screen.

It is possible to peel the strips off the screen and clean them up, allowing plenty of time for the cleaning fluid to evaporate before reassembly.

You could also try the Loco button at power up test to check the EEPROM is OK and each button key works.

[Brew Man]

@96RAF

I can't really see it being a connection problem, as powering down and up again immediately fixes the problem. Plus it behaves in the same way each time, i.e. no random parts of the display failing, always line 2. I intend to do an EEPROM check before anything else. After that, if a reset doesn't work I could try re-installing v1.45, though that's a bit of a long shot.

[Brew Man]

Update: First I ran an EEPROM check which passed with no issue. I then left it switched on but not connected to any track for one hour and then with the two locos I had previously been running, selected but still not connected for a further hour and the fault did not occur. So the problem appears to be load-related.

Later I will connect it to the layout and leave with nothing running for an hour, then for another hour with one of the locos running and finally with both running. If the fault occurs during any of those stages I will do a reset and go from there.

[Chrissaf]

If further diagnostic testing seems to infer a supply current issue (not voltage) then it is not unknown for lack of Elite currrent issues being attributable to a faulty 4 amp power pack.

[Brew Man]

The problem seems to be too predictable to be anything like that, though it's so peculiar that nothing can be ruled out. Not had chance to continue with the diagnostics yet today, so might have to wait until tomorrow.

[96RAF]

A dodgy zebra connector would tend to be more random. I would be looking at a loose ribbon cable, bad solder joint, whiskering or corrosion seeing as the Elite is fairly old.

If its any help I have the odd spare display that you could swap out for diagnostic purposes. Happy to post within UK f.o.c.

General opinion is pointing at a faulty display driver. A unit reset may fettle it or not. If its in the basic PIC operating firmware it would not be user fixable.

[Brew Man]

@96RAF

That's very kind of you, Rob, thank you. Since my last update I have left the Elite connected to the track with nothing running for almost 3 hours with no issue. I then ran both locos, (with trains coupled) and the problem kicked in after about 5 minutes. Next I ran one of the trains round the inner loop for almost an hour with no issue. Ran the second loco round the outer loop and the problem re-occurred. I must point out that the two loops are isolated from one another at the moment, apart from their common connection to the Elite. So the problem seems to be happening on the outer loop. Next I thought, could it be the power connector? It's funny what you wake up in the middle of the night thinking about. So I opened it up to check that I had snipped the capacitor out, which I had. However, with the latest batch of these connectors the capacitor is placed in such a way as to make it very difficult to remove so I just snipped one leg and put it back together. I noticed though that it might be possible for the snipped leg close to the body of the cap to come into contact with the conductor, so I swapped the whole thing for a DCC labelled one. Result - no difference. What I (think) I've noticed though is that it only appears to happen on the outer loop. I've got a train running round it, (again) now. I must point out that the fundamental difference between the two loops is, the outer is powered by the power clip and the inner is directly soldered to the track. I have come to the firm conclusion that those clips are rubbish and prone to very bad connections. If the current run proves issue-free I will ditch the clip and solder the power leads directly. These two loops are temporary just to get things going. Eventually all the track will be directly soldered via a bus with droppers. I've also done a reset in all this which still made no difference.

If all else fails, Rob I may well take you up on your kind offer. I'll check progress later and report back.

[96RAF]

Edit as shown by strike thru and italics.

It is looking likely that it is a power issue on the outer loop, maybe a high resistance increasing the current affecting the load on the controller, but why it should affect the display the way it does I have no idea. Good fault finding so far.

I will ponder the schematic again but not sure I can pull a rabbit out of that hat.

An RRAMP meter would be able to monitor voltage around the loop and provide current draw to compare with the inner loop, else using a scope to comparatively look at the DCC signal on both loops, but maybe I am getting too deep in the weeds.

Do you have anything else connected to the outer loop, such as acc decoder, current sensors, signals, etc.

I have bag fulls of those track clip connectors both DC and DCC so you can have a couple of those too. Do you have a power track you could slot in or maybe swap the inner soldered track piece across for the test.

[Fishmanoz]

Just correcting some logic here, but not helping with the problem.

You say the loops are not connected, except via the controller. What that actually says is the loops are connected to each other, in parallel via the controller.

Then Rob, you say a high resistance on the outer loop “increases” the current. Sorry, high resistance decreases current not increases it, all the way to nothing on the track means open circuit (infinite resistance) and no current.

As I said, doesn’t fix the problem but can think about it in a more informed way now.

Mod note - my post edited re your comment.

[Chrissaf]

I still feel my comment regarding the power supply is being overlooked. I have seen posts on the forum before where obscure Elite issues have been fixed by replacing the power supply. I appreciate that it is an expensive gamble, but at least a PSU purchased unnecessarily can still be reserved as a spare.

[96RAF]

Agreed Chris

There was a recent case involving several R8239 Booster Modules where the 4-amp unit was giving acceptable output when measured off load but was failing in service. The problem was manifest when a second booster unit was installed and it caused the loco lights to flash in that power district. Fortunately swapping boosters and PSUs around transferred the fault and the user was able to isolate the single root cause. A replacement PSU cured the users fault.

In the BrewMan case a marginal output under increased load of a possible dodgy (for whatever reason) outer loop may be tipping the scales.

[Brew Man]

I've thrown a red herring I'm afraid, chaps. Last evening I ran a train on the inner loop for an hour or so and the b*****d went and did it again. First time I've seen it on that loop. So, back to square one really.

@Fishmanoz Yes, you are quite right, they are connected in parallel. Just meant there are no other connections such as points between the two.

@chris I can quickly test the PSU, Chris as I believe I've got a 4A supply for my Select, albeit a laptop one I salvaged from an unused laptop. I'll give it a go later.

@ROB Pending the PSU check, if that fails to fix the problem I have to accept it as being an issue with the device itself, and probably a fault around the display driver as previously mentioned. In which case I will take you up on your offer of a replacement display to try. That is assuming that the driver chip lives on the back of the module.

I'll report back later after I've checked out the PSU.

Thanks for the interest, guys in helping me to get to the bottom of this.

EDIT: just to say I've been saying throughout this thread that the second line of the display disappears. It's actually the second and third. i.e. line two of the alpha-numeric and the direction arrows and speed indicator.

[Brew Man]

Do you have anything else connected to the outer loop, such as acc decoder, current sensors, signals, etc.

I have bag fulls of those track clip connectors both DC and DCC so you can have a couple of those too. Do you have a power track you could slot in or maybe swap the inner soldered track piece across for the test.

 

 

Just realised I hadn't addressed these Qs.

Nothing else at all is connected to either loop except the feeds to the controller.

I do have a power rail, two in fact but in the light of discovering the problem is affecting the inner loop too, it wouldn't really prove anything and with regard to the power clip, it was only meant as a temporary measure anyway, but thanks again for the offer.

[96RAF]

@ David

If you ref to the Elite Teardown here . . .

http://www.halton96th.org.uk/page12.html

. . . you will see the display is mounted to a daughter board (the one with the ribbon cable - check the clamp has not worked loose) by way of a metal frame. That board looks to mount the drivers. I can rob that whole assy from a dead Elite or I can provide a bare display as seen in this article about installing a backlight . . .

http://www.halton96th.org.uk/article8.html

If you do pull the Elite apart take off the heat sinks from Q1-Q4 and reapply heat sink paste and firmly screw them back on. Early Elites has those bent ali things but later ones had a black extrusion as adopted for other modules and controllers.

Choice is yours as to what you want me to send. I will make contact to get your address once you have panned out on the other fault finding routes. When the fault showed up from the inner loop was that the only one connected at the time or was the outer loop in circuit but not running a loco.

[Brew Man]

Hi Rob

I did have a brief look at that Teardown a week or so ago but will read it properly later. I seem to recall you mentioning octagonal(?) screwdriver heads needed to remove the screws. I'll need to get some of those. I'm sure Amazon will have them. I do have the star type used for removing tamper proof screws but I got the feeling the ones you mentioned were different.

Yes, the outer loop was also connected. I see where you're coming from, I'll try it again with the outer loop completed isolated. If the fault is still happening I will also try the PSU from my Select.

I really do appreciate your help in all this.

EDIT: Just been up to the loft to disconnect the outer loop and set the train in motion. I notice that the screw heads in the Elite are triangular, (only 5 out grinning). What are they called and I'll get a set.

EDIT 2: Had a look on Amazon and they're called Triangle head screwdrivers. Who'd have thought? joy

[Brew Man]

Well it's been going for just short of five hours now without a blip and it would have definitely kicked in by now if it was going to.

Next up I will run the same loco on the outer track isolated from the inner. It'll be tomorrow now. That will be interesting.

[96RAF]

Just a thought.

Although in your OP you say ‘the two principle circuits’ this does not specifically state that these circuits are totally isolated from each other.

I wonder if when you connect both to the Elite are you matching phase, i.e. inner rail feed wire to inner rail feed wire et seq. If not then there is a chance of both circuits being ‘shorted’ to each other say by a loco traversing a point for a common siding.

Running each circuit independently will prove them, but worth a phasing check when you wire them both in.