Select error codes & gives OL error with empty track

[walkingthedog]

I don't often read them either.

[Chrissaf]

It does become tiring to read long unbroken posts without any paragraph breaks. You start to read them, then get bored and then tend to scan the rest of the post for the salient points (no pun intended).

.

I find that the same text when formatted with paragraph gaps, is easier to absorb the details as one's mind gets little concentration breaks to consolodate the data contained within each separate paragraph.

.

F.A.O T&F,

Have a read of TIP 4 regarding paragraph breaks in my 'Forum How To' page.

[Yelrow]

There you are then, vindicated. Twas the paragraph  breaks, or lack of them.   Knew there must have been a reason, why i missed the extra points fitted bit.

[Fishmanoz]

What he actually says is " having enjoyed success with one, decided to expand the number of points to work in tandem in this fashion"

To me this means one of two possibilities, either:

 

-  not that he has increased the number of points but expanded the number of 8247 ports he has 2 points attached to, or

 

-  he has installed more crossovers between loops, so yes more points

 

But not sure either which way?

[Trackandfield]

Please dont accuse me of being ungrateful here, but having had my posts methodically dissected by the English language pundits and personal interpretation of terms, are we any nearer to finding a solution to the issue originally posted.

 

The number of points has not changed, merely their type -- from "normal" left/right to express (or high speed if you prefer) and 1 extra R8247 decoder has been added in order that one point can have one dedicated Logical address.

 

Hope this helps.

[walkingthedog]

Have you fitted Peco electrofrog points?

[Trackandfield]

Have you fitted Peco electrofrog points? - No only 2 Peco point motors under Hornby points identified as P3 and P4 in the 2nd image posted yesterday. These have worked from day 1, and continue to work even with the OL condition as they are wired from the 15v DC connection on the Select controller.

[Chrissaf]

Having now seen the layout photos. I confirm that there are no 'reverse loops' that have been built into the design.

.

If every point physically on the layout also existed on the pre-fault layout. That is to say, no new points have been added and no points have been physically replaced but the layout now has a short circuit. Then logically one of two things must have happened.

.

Either one of the points has been disturbed whilst changing point motor wiring, and the cross-over wiring fitted to the point in the factory under the frog area has come into contact. This is a point fault that has come up before on the forum and was documented in this previous post:

.

https://www.hornby.com/uk-en/forum/loop-siding-short-circuit-problem/?p=1/#post-231478

.

OR

.

One of the R8247's has gone faulty and is presenting a 'short' to the Select controller.

.

One easy way to try an identify the fault location is to:

.

• First disconnect all the R8247 'DCC In' connections, so that the Select controller is only connected to the track itself.
• Also disconnect the AUX output of the Select from the CDU.
• If the Select still shows a 'short' with only the track connected, then the fault is either with the track and has to realistically be a point fault as documented above OR the Select itself has a fault.
• So test for that by completely disconnecting the Select from everything and powering it up.
• If none of the tests above indicate a 'short' then reattach the AUX output to the CDU and see if the fault returns.
• If still no fault, reconnect each of the R8247 'DCC in' inputs in turn to see if one of the R8247s generates the short.

.

Doing the above tests should show what area of the layout has the fault and where further investigation is required.

[Trackandfield]

If every point physically on the layout also existed on the pre-fault layout. That is to say, no new points have been added and no points have been physically replaced but the layout now has a short circuit. Then logically one of two things must have happened.

 

 The 4 points indicated at lower left adjacent to the Level Crossing (although I do accept that only 3 are "marked" as TBA) have ALL been changed in the past three weeks or so from standard points to R8077 LH express points. The OL condition has only manifested itself SINCE wiring these 4 new points up to the decoder in multiples in order to make a path work, rather than one to one (a process I am now trying to reverse)

 

I am trying to avoid lifting the points at all costs, when all the evidence points to the Logical addressing 2 into 1, but am willing to consider all options. Reading the topic Chrissaf added above is certainly food for thought, but having tested each decoder individually via a test wire as the ONLY object connected, I believe I can rule out a faulty decoder at this stage

[96RAF]

Can you confirm what the firmware version of your Select is please. Check the first number seen at power up, which will be followed by 30 then 03. You should see 10, 11, 12, 13, 14, 15 or possibly 16.

 

I wonder if we have a Select firmware bug problem affecting a PAD (R8247).

 

The other thing is all the C ports on an R8247 are connected together internally to the unit. If you have any point motor wires crossed when pairing them then you have built in a short. If all point motors are the same make then make sure the colours match port to port snd motor to motor. The only exception is if one of a pair throws out of synch with the other motor then swap the + and - wires over on one motor. The common connection of all motors must go the the C terminal of the unit.

 

To add to Chris’s fault finding logic, disconnect/ reconnect the motor pairs one by one at each R8247 port until the the fault clears. You then have where your odd wiring is. Dig deeper from there checking the connections as I suggested earlier.

[Chrissaf]

...and 1 extra R8247 decoder has been added...

.

Could this 1 extra R8247 decoder have been configured in error for 'always on'. Very possible if it was purchased second-hand and configured such by the previous owner. Solenoid point motor coils have a very low DC resistance typically about 4 to 12 ohms subject to brand and model. If this low impedance (DC resistance) coil is attached to a R8247 port configured for 'always on' then it will look like a short circuit to the decoder.

.

The Select will not allow you to change the CVs associated with 'always on' unless it is first upgraded to version 1.6 firmware. In the absence of being able to read CV values (something you can't do with a Select, even with 1.6 firmware) then the only way of checking to see if a R8247 port is configured for 'always on' is to use a Multimeter on the Volts scale and take voltage readings on all the four ports between the C & + terminals and the C & - terminals preferably with no wiring attached to the decoder ports.

.

If no Multimeter is owned, then a car sidelight bulb will do instead, if no car bulb then a 1,000 ohm resistor in series with an LED, but note that the positive side of the LED would need to connect to the C terminal during the tests. If a bulb or LED stays lit all the time that it is connected, then that port is configured for 'always on' or if configured for 'pulse' operation it is faulty.

[Trackandfield]

Hi RAF96 - The firmware version is 15, and I will again check the individual connections tomorrow. 3 days lying on my back for 8 hours a day is starting to take it's toll, but many thanks.

[Trackandfield]

In reply to Chrissaf - no, all decoders were purchased new.

[Chrissaf]

In that case, the possible fault diagnostic options to check are dwindling fast I'm afraid.

[Trackandfield]

In that case, the possible fault diagnostic options to check are dwindling fast I'm afraid.

 

Beginning to accept that, but until all options are explored I remain hopeful !

[Trackandfield]

Its unlikely to be a dropper wire issue is it - can a dropper wire ever be too thick bearing in mind the width of the wires from the Hornby point motors are miniscule. I've tried to use telephone wire wherever possible with a colour co-ordinated collar for matching purposes at the decoder, but I do accept it's not all the same width. . .

[96RAF]

If a dropper wire is crossed then it will give a short circuit and resultant overload.

The gauge of a dropper is of no consequence in the context of this wuiry. You could wire your layout totally in barbed wire and it would work unless you have crossed connections.

 

If I remember correctly you say you have separate track and points buses. To rule out which is causing problems disconnect the points bus at the controller end and see if the track overloads the controller. Then switch connections I.e track disconnected but points connected. Does the short follow a bus, i.e. do we have a culprit bus. If so take the diagnostics forward from there based on what changes you have made to the layout. We can only go on what you tell us, not what we would know if we had modified the layout.

[walkingthedog]

Can you be sure none of the droppers are crossed. It can be very difficult when you are wiring circuits. The best way is to take a wagon  and add a length of wire the same colour as the droppers to each side of the wagon corresponding to the colour of the dropper that side. Then wheel the wagon all over the layout without lifting it off the track. The colour of the wire on the wagons side should always be the same colour as the droppers. 

[Yelrow]

What a good suggestion.  I used telephone cable, 8 core, here, so most of mine are the same colour. Fair to say though, when i was underneath, and could not be bothered to get up, grabbed what was to hand. Would have been ok, for about 90% though. Just as well, it wasnt the point decoders, as they got lengthened with what i had to hand, Needed a  sheet or two of paper to write them all down, just in case.

[Chrissaf]

Trackandfield, posted this text below @ 10:25. It has been copied and reformatted whilst the original post has been removed due to it all being displayed in a quote box.

---------------------------------------------------------------------------------------------------------------------------

RAF96 said

 

"If I remember correctly you say you have separate track and points buses."

 

This is technically correct but again, open to interpretation.

 

All 19 points are Hornby, 17 of which are dropped to the A & B sockets from the MAIN DCC track bus. The remaining 2 are fed via a CDU from the 15v DC AUX output of the Select controller and are therefore NOT logically addressed - these 2 are unaffected by the OL condition and will fire happily all day every day.

 

When all 5 decoders are wired to their respective points from the + - and C ONLY (and are not wired to the bus) the OL still exists. By renewing the wiring to the decoders one by one I hoped to eradicate any possible crossing over issues. I'm taking on board yesterdays comments and hoping somethig positive will come out of it.

[Trackandfield]

For the benefit of other Forum members, my issue under the above subject has been resolved by trial and error probably, but this is the bones of it.

 

As I said, the board is an 8' x 4' folding with a continuous bus. Where the boards join, the builder had looped the bus to allow for the fold.

 

When I replaced some of the points I modified the point motor wiring to fit the nearest decoder (OCD and all that) and imitated the builder in the same way. After threading the wiring through the quarter inch joint gap I used a tool to pull the wires down through the middle of the gap and it was this action that maybe caused a tear in the wire sheaths (maybe)

 

After rectifying those wires I changed the bus from continuous to a simplified T shape with droppers either side of where the bus now terminates either side of one joint.

 

Either way, the OL condition has gone, so to all who have contributed towards a solution, directly or otherwise, I thank you !