Problem with DCC Select or loco or both?

[Stephen-1211844]

Hi everyone, I'm a newbie (albeit 66yrs old) of two weeks. Two weeks ago I bought and got everything set up including some track and a Hornby loco and Hornby DCC Select and power clips etc, etc and I installed a Hornby decoder in a Hornby loco and all was running perfectly - so pleased. Then I consulted my local model shop about a second loco and thats when the problem started. I'm sorry this next bit is so long winded but its all I can think of but to tell the whole story.

The shop convinced me that a Bachmann loco with a Next 18 chip would be fine. And I saw it run fine on their test track (a DC set up). When I got home I installed the chip and allocated it a unique address on the Select, having taken my other one off the track, but it ran very poorly, to put it mildly, (jerky, taking ages to get up to speed and squeaking). I took it back for advice. They swapped it for an identical model and inserted a different make of chip and changed the address to '1' and the acceleration rate but using a different make of controller - a 'T' shaped thing.

When I got it home and tried it, it didn't run at all. I took it back. It ran OK again this time on their test DCC track with their controller, and this time, obviously it had the chip installed. I figured the chips just hadn't been compatible with my Select. They gave me a refund. So at this point I have had two identical locos and two different chips.

The problem is, immediately after all this, my original loco no longer runs perfectly as it did before. Ostensibly nothing has changed except the events above. It now makes a clicking sound, it doesn't run smoothly but shimmies slightly left and right, runs a little jerkily, and 50 percent power which was previously just right is now little more than a crawl. I have reset my Select, re-addressed the loco and adjusted the acceleration setting but it makes no difference. What to do?

Whatever they did somehow messed up or somehow irreversibly reprogrammed my once perfect set up? Nothing else happened that could have. All I can think of doing, is to replace the Select and trash the original loco (an expense that I would like to avoid). I have now ordered a second loco am afraid to put it on my system in case it gets 'messed up' too, by whatever occurred due to the sequence of events.

Once again, sorry this is so long winded. Many thanks, it sees strange, I know. Hope you can help.

[Chrissaf]

Can you please specify the track power connectors you are using by their R part numbers. If they are R602 or R8206 power track connectors then they are not suitable for DCC and can cause very odd behaviours to be observed if they are being used. This includes the R8201 link wires which are also not DCC friendly.

With regard the Select, what firmware revision level is it? If it is a brand new purchase from a reputable supplier, then I would expect it to be firmware version 20. The firmware number is the very first number to be very briefly displayed on the Select LCD screen as it is being powered up. Expect to see a number between 10 & 20. Knowing the firmware your Select is using, could affect what to suggest you do next to resolve your issue.

Finally, please state the Hornby loco model R number that makes the clicking noise and runs with a shimmy. If it is a loco that has coupling rods connecting the wheels together, it sounds like a wheel has moved on the axle and the coupling rods are now out of alignment.

[ColinB]

I have an Elite, but generally I can run virtually any decoder. Chrissaf is right, one of your issues may be the power clips. If you are using the normal Hornby ones, these contain a capacitor which upsets the DCC signal. I connect the wires straight to the track. I don't know a great deal about the Select, but the symptoms you describe, I don't think a Select could do, even with out of date software. The other issue you might have is the Dealer will have a straight piece of track, probably perfectly clean, you probably have a layout with bends and bits of dirty track. How you tried running each loco on a metre length of straight track. The clicking sound is probably either issues with the valve gear or possibly a broken gear, it is probably coincidence that it started happening when you bought the second loco. Also going round bends on your layout may have dislodged something. In my experience most of the issues you get with controllers is trying to program CV value and accessing function above the number 4. There seems to be some fake messages going around about Hornby controllers running certain chips. In my experience and I have tried very many of them, they all work with Hornby controllers, some have issues with programming CV values, but generally they will all run. I must admit I use Zimo decoders for most of my locos, unless they are sound, mainly because they have a higher current setting. I have had no issues with them at all, other than ones of my own making. I suppose the only other thing may be that you had to reprogram the address on the second loco, it may be that you inadvertently set the acceleration rate by mistake or your Select did.

[Stephen-1211844]

@Chrissaf

Many thanks for your reply. Here's the info you requested:

Track power connectors

I am connecting the outer loop with the inner loop with what I am sure is Link Wire R8201 with the copper coloured connector. One clip on each loop connected with wires a to a and b to b, on the same side of the layout that the Select power supply is connected (size of layout is all extension packs except F so Two loops). The main power supply connector with the two little green buttons on top linking the select with the track, judging by the catalogue, is R8241.

Firmware is 1.6.

Unit supplied by very reputable shop as brand new.

Loco R numbers:

The original one that I refer to as running perfectly initially, but which is now running with a shimmy and clicking is Hornby R3825, Peckett 614 Limited Edition. I've looked carefully and there is nothing obstructing the wheels. The second one added later that seemed to introduce the problems was a Bachmann.

Thanks again. Looking forward to hearing from you.

[Stephen-1211844]

@ColinB

Many thanks for your reply. I note you comments. Please see my reply to Chrissaf. I note you both mention the power clips. As for my track, I've inspected it very closely and seems to be OK and its elevated.

Cheers.

[RDS]

@Stephen

You start off by saying that your first Loco ran perfectly. Have you made any changes to your track layout since it was running perfectly?

Is it possible that the Loco was dropped, which is now causing it to click and run unevenly?

It is probably worthwhile just trying each Loco separately with just a short straight piece of track.

[Fishmanoz]

Stephen, for a start, it is most unlikely that anything has been caused by any decoder or controller or combination thereof that has led to any permanent damage. There will be a solution to your problem. And if there happens to be any mechanical damage, it will most likely be repairable.

First thing to check - Link wires - already mentioned in passing but absolutely essential you ensure any suppression capacitors in them are removed before doing anything else. Flip open the black covers adjacent to the contacts (both ends) and remove any components connected between the track connections by clipping off their legs. Chris has a diagram he may post to help (his diagram is in the Rookie Question thread currently immediately below this one). Having those capacitors in circuit distorts the DCC waveform and cause all sorts of problems in running, not to mention programming decoders to new addresses, accel adjustments etc.

Also make sure all connections are Track output A to Rail A and Track output B to rail B. You’ll know if you get this wrong because the Select will overload and trip out.

For the future when all of this is resolved and you start to extend your layout, DCC requires that all of the layout is live all of the time. You will likely find you need to install DCC Point Clips on your points to achieve this.

PS. The white arrow in blue box is not a Reply to Post icon. It just unnecessarily repeats the original post as part of your reply. Go to the text box at the bottom of the thread and post your reply there, just as I have done.

[Chrissaf]

Stephen,

The unnecessary re-quotes that Fishy [Fishmanoz] mentioned in his reply above have been edited out. You might benefit from browsing my 'How to use this forum' TIPs page. TIP number 5 relates to replies.

TIPs on using the forum. :: Hornby Hobbies

Now with regard your answers to my Q's.

The R8201 is not designed for use with DCC for the reasons that Fishy describes. In fact, with a DCC layout the need for any link wires is eliminated if R8232 DCC Point Clips are used. There are instances where the R8201 link wire is not wired A to A & B to B. This is covered in FIG 12 on Page 15 of my PDF tutorial. You mention Track Packs A to E indicating that you are using the Hornby TrakMat layout design. I have written a downloadable PDF tutorial for this layout. Chapter 6 covers the use of this layout when powering via DCC. The PDF can be downloaded from here:

Getting Started [sets and Track Extension Packs] :: Hornby Hobbies

The R8201 connectors can be modified for use with DCC using the method described by Fishy in his reply. All the connectors on your layout should be checked for the presence of any capacitors and removed when found. This includes both connectors of the R8201 link wire. The supporting graphic he mentions is reproduced below:

The capacitors in your link wire connectors are almost certainly a contributory reason for the DCC control issues you are reporting. For example, locos not responding and also decoder configuration errors. As Fishy says, the capacitors distort the DCC signal and this affects some decoder types and brands more than others. For example the Hornby TTS decoders are particularly sensitive to these incorrect capacitor being present issues.

Your Select 1.6 firmware is actually the latest Select firmware and is identical in features and functions to version 2.0. Version 2.0 firmware is loaded into a slightly later physical hardware version of the Select and just given a new higher version 2.0 revision number to indicate the later hardware. But functionally, 1.6 is identical to version 2.0

Now since you have version 1.6 firmware you have more options open to you. Once you have resolved your power connection capacitor issue. I suggest that you use your version 1.6 Select to factory reset the decoder in any loco that is not responding to DCC control commands. You 'factory reset' a DCC decoder by writing the value 8 to CV number 8. Once the 'reset' is successful, you should be able to then test the loco using DCC address number 3. Once the loco is working on DCC address 3, then you can re-address the loco decoder to the actual address you want to use. Initially testing on address 3 is important as it proves that the decoder is functioning correctly before you make any changes to it.

With regard the shimmying Pecket.

This is more likely a physical issue rather than one of an electrical DCC decoder issue. Compare your coupling rod alignments on both sides of the loco to the image below:

If there is any evidence of rods not being in straight alignment, then a wheel has slipped on the axle and a rocking shimmy can be the result. The audible 'clicking' is also a strong indicator that coupling rod alignment has been compromised. The rods should be the focus area for you to concentrate on. This is where a 'rolling road' can come into its own. It allows you to closely observe the locomotive motion without having to chase the loco round the track.

As this is a long reply and contains images. Please do not try using the 'blue button' to reply. The images will force your reply to be 'held back' for image approval. Just use the 'Reply' text box at the bottom of the page as Fishy described in his reply.

[Stephen-1211844]

Hi, thanks. No changes whatsoever and definitely not dropped.

[96RAF]

In that case I would take the offending loco(s) and your Select along to the model shop and ask them to prove functionality on their test track using your kit.

This would move the focus from your track to theirs eliminating one possible fault cause. If they can make it work on their track then you know the DCC problem is at home.

If they swap your Select with their controller then they can prove-disprove any fault by comparison.

The wombling loco will definitely turn out to be mechanical.

If any of your kit is still under warranty then consider sending/taking it back to where you bought and ask them for a replacement or full refund.

[Chrissaf]

If it is not the coupling rods, then there is a possibility that the motor mount has come loose inside the loco affecting the mesh of the gearing. Gear mesh issues can also cause 'clicking' noises and usually show different observed symptoms when going forward compared to when going in reverse. Sometimes only one direction of travel observes the issue.

[RDS]

I am struggling to understand how something that worked perfectly, suddenly does not work, without any changes having been made.

The answer when it is eventually found, will be fascinating.

Please keep us informed!

[RB51]

Hi Stephen, sorry to hear this tale.

The Select will have no problems with any decoder chips that satisfy NMRA standards, and your Next18 does this. You don’t say who the other one is manufactured by.

The issue is elsewhere. It is possible to mis-code and confuse yourself, the chip etc. As a starting point I would suggest reprogramming the chips back to default.

Do not get disheartened. We will help you sort it out and on no account trash anything.

R-

[ColinB]

It is called electronics and software, it has this amazing ability to do horrible things if it doesn't get the right signal. See how your TV copes with a bad signal, it misses a bit or freezes the screen. Generally once you fix what is the issue, it will ok. If you do have the wrong powerclips, then the DCC signal will be either attenuated by the capacitor in it, or disappear completely. There are some decoders that are worse that others, Hornby are perfectly ok, so are Zimo and surprisingly LaisDCC. Generally when you buy a new loco with a decoder in it, try it "standalone" with its default address of 3 on your layout to check everything is alright before you start changing addresses.

[Fishmanoz]

Colin, I wouldn’t think of the capacitors causing attenuation of the signal, I’d stick with distortion, or degradation, as Chris and I have stated. A quick analysis in the frequency domain will tell you that, for the square wave which the DCC signal is, the higher harmonics will be affected more than the fundamental. The result - a rounding of the square wave and change in the angle of attack. Yes, some attenuation of all frequencies but it’s the differential that counts here almost certainly.

Going from the theory to what we see in practice, with capacitors in circuit we still have a DCC signal, just not a very good one. Some decoders will be able to handle it, others may not. Some will be right on the edge and ok one day, not the next. Stray capacitance and inductance around the layout will contribute. And changes to the layout will affect those, meaning a decoder may be working fine one day then completely unreliable the next after a siding is added to the layout.

[ColinB]

Ok, Fishmanoz I didn't want to get too technical in my reply, that is why I try to use simple answers. The capacitor acts as a low pass filter, the amount it filters depends on the frequency of the signal. So effectively any high frequency pulses will get removed and the lower frequency pulses will get rounded which will mean that the processor in the decoder will not see a clean edge. So that is why you still see a signal, with any digital comms network which DCC is, it only has to lose a pulse to completely upset the data. It does in effect attenuate the signal by rounding off the edges. Either way the DCC signal is not going to work that well. Why do you think the capacitors are there in a DC power connector, to get rid of those high frequency motor spikes that interfere with the TV.

[Stephen-1211844]

Morning Chris, thanks again for your advice and the links: indispensable reading.

Re Power: I'm in the process of removing the link wires, replacing them with point clips (R8232). Then I'll be resetting the decoder as per the addendum leaflet to the v1.6 Select manual on the main track but remembering only to leave the one loco on the track, and then asap setting up a test rack as per the Select's booklet.

Re Shimmy: this occurs in forward and reverse but less pronounced in reverse. If by coupling rods you mean the arms that make the wheels turn, I don't have a rolling road but when at rest the rod on one side is always centred on the wheel and on the other side its positioned at the top of the wheel, is this what you meant? When it's moving it's hard to tell if they are in unison.

Re a to a and b to b: Not that it matters now that I have the clips but inner rail to inner and outer to outer, right?

55 yrs since my last train set. Second loco arrived this morning. Spent weekend making scenics. Great stuff!

[Stephen-1211844]

Hi everyone,

RogerB - Hi, the other decoder was a Guagemaster DCC18. Thanks.

I have now carried out the steps suggested above but the problem persists i.e. the shimmy, slow running and jerky and when at rest the connecting rods are no in identical positions. I have reset the Select controller, removed the offending R8201 link wires and thoroughly cleaned and checked my track, and reset the Hornby decoder in the Peckett. I have done this exactly as described here or in the Hornby literature. Can you suggest what I can try now please. Thanks.

[Chrissaf]

I have now carried out the steps suggested above but the problem persists.

 

 

Just for clarification. The problem that is still persisting is what?

Is it the Pecket shimmy?

If so, as previously suggested by more than one member, this issue is most likely physical / mechanical rather than electrical.