DCC Select - locos constantly crawling

[Hipwell]

Hey guys, I picked up a select today as railmaster was ruining my experience when I just wanted to have a quick 5 minutes to play trains after work it was a pain having to get the laptop out and I didn't find controlling the trains with a mouse very controllable.

So, wired the select in, all good, all locos and points working.

But then I noticed that my locos wouldn't stop to a complete halt when I turn the knob all the way counter clockwise. They would crawl really slowly.

I have got around this by setting the speed steps on on the select to 28 from 128 which seems to have cured the issue, but I feel like it might just be masking an actual issue as the select manual suggests all Hornby decoders (mine are) will support 128 speed steps no problem.

Any ideas?

[Chrissaf]

Yes there is a fix for this known issue. It does not involve making any changes to the loco fitted decoders, thus I advise that you don't tinker with decoder settings as you risk making things worse. The fix is performed only on the Select itself. The things is that I cant remember the detail of the exact fix process that you need to implement.

Once Rob (96RAF) sees this question I'm sure he will give you 'chapter and verse' on what you need to do, as he is the originator of this fix when it has been previously documented on this forum.

[Hipwell]

Thanks Chris, I'll await robs reply.

However, the select manual does specifically say that changing the speed steps on the controller doesn't affect the decoder, it's a controller setting only and isn't programming anything.

Regards

[Chrissaf]

It would be useful for Rob to know what your Select firmware revision level is. The very first number to appear very briefly on the Select screen as it powers up.

[Hipwell]

Its a version 1.5 unfortunately, but I was hoping I won't need to update it as any programming I could do from railmaster.

[Chrissaf]

1.6 / 2.0 firmware is not a pre-requisite for the fix to work.

PS - I have sent Rob an email to bring this thread to his attention.

[Hipwell]

Thanks Chris

[96RAF]

Hi - as Chris says this is a known problem and is associated with loco zero (i.e. analogue loco running).

To run a DC loco (which is not recommended and the manual even cautions against it) a DCC controller sends what is know as a stretched zero bit command. This distorts the DCC signal to provide a pseudo DC voltage to the track.

It may be that at some time your controller has had loco zero selected and throttle left applied, which has left a latent distortion in the DCC signal. It is this distortion which can affect other DCC locos on track.

The fix is simple, either select loco zero and cycle the speed knob to max and back to zero, which will null the stretched zero bit, or reset the Select.

The problem with a simple reset is that at v1.6/2.0 DC running was disabled as default, but previous firmware revisions had DC running enabled as default, however as part of the unit reset procedure you can turn it off. Simply follow the reset procedure in the manual and turn off DC running.

Speed steps - some folk find using 28 steps easier than the finer control offered by 128 steps. You can set this in the controller as well as in each decoder if desired. The only time it really matters is if you have sound decoders installed.

[Hipwell]

Thank you rob. I will double check this after work but I could have sworn that turning off DC running was the first thing I did when I first powered up the select.

Would disabling DC running on my decoders also sort the problem?

[Chrissaf]

I think that Rob's mention of DC running may have confused you. The fix you need to apply does not relate directly to DC running.

To recap and expanding on Rob' posted fix.

1. You select loco address zero.
2. You turn up the speed control knob to max. Ignore any movement the loco might make at this time.
3. You turn the speed control knob all the way back to zero ensuring that the speed knob is hard against the zero speed end stop.
4. You now select the configured DCC address of the loco.
5. You should now find that normal control has resumed.

Rob's reference to DC were comments relating to his Option 2 fix which is to just perform a Select factory reset. He was just mentioning that Select factory resets are handled differently subject to which firmware is installed. And that pseudo DC is what DCC Address zero is all about.

Note : As your firmware is 1.5, you can't do anything with decoder DC operation as version 1.5 can not write to CV 29 on your decoders. Albeit, you could use your RM setup to amend CV29.

[Hipwell]

Thanks Chris. I'll take a look tonight and give it a go. Hopefully this solves it as I'm really enjoying running with a control knob again rather than the mouse!

Just to note, I also own an e link and RailMaster so if any programming was needed as part of this fix I would use that which is why I'm in no rush to update the select to 1.6

[96RAF]

Turning off DC running in each decoder will prevent the decoder reacting to what it sees as loss of the DCC signal, which in turn makes it revert to analogue power and of course it actually gets full DCC voltage of around 15v AC which causes it to shoot off.

Having a decoder set for DC running means it can be run by an analogue controller on a train-set or even by a simple 9v battery for testing.

The controller sending a pseudo DC signal superimposed on the DCC signal is something else, but it can have the effect of confusing a decoder, as can using a DC track connector instead of a DCC one. The loco zero facility is specifically for running a single analogue loco on a DCC layout, which as stated is not recommended as the DCC voltage can burn out a directly connected DC motor.

[Hipwell]

Sorted. Thank you very much.

Can I ask, when programming the acceleration and deceleration on the select, is it actually changing the CV settings on the decoder? Or is it just altering the way in which it makes the loco slow down?

If it's the former, how come version 1.5 cant program any other cvs if it can program acc and dec?

[Chrissaf]

Yes it is changing the CV's in the decoder.

The capability to amend just a limited number of CV's with firmwares 1.5 and below is because those supported CV's were specifically included in the firmware as being absolutely critical requirements for the product to function as a controller. Changing the decoder DCC Address is also a CV based edit requirement, so from a CV's point of view, it is not just ACC and DEC decoder CV's that the Select is amending.

You also have to take into account that at the time the original Select was bought to market, the Hornby decoders also had a very limited range of supported CV's. Thus the products were somewhat tied together in terms of technical capabilities.

The limited range of editable CV's with the Select would originally have been a Marketing decision, probably to provide a justifiable reason to promote the sale of the more capable Elite controller at three times the cost. If you could do more with a Select, then there would be minimal justification to buy an Elite. The expansion of the Select firmware to now edit a wider range of CV's will have also been a Marketing led decision based on current market conditions. What that reason is, I have no insider knowledge to comment on, but possibly a result of Elite and Select sales being lost to competitor products. Thus expanding the capability of the Select could possibly be seen as countering competitor threats. There are still a wide range of things an Elite can do that even the latest Select can not. Thus the Elite has not been completely undermined by the Marketing decision to expand the Select capability. Additionally, during the Select lifetime the Elite capabilities have also been expanded via firmware upgrades thereby maintaining a level of product capability separation between Select and Elites.

There is also a technical limitation on just how far the current Select hardware can be expanded via firmware upgrades. The Select product has a finite amount of code storage capability (usable memory) thus expanding the Select capabilities any further will require a significant hardware revision.

[Hipwell]

Brill thanks for that.

Unfortunately I've just tried again and it's not working!!

I've followed the instructions to the T grr.

Basically, as soon as I plug in the select and it's started up, loco 3 starts crawling, I can speed it up and slow it down, but it won't come to a complete stop.

Tried selecting loco 0 as you said and turning it up and down then going back to loco 3 but its still crawling.

It worked the first time after I reset it, but then after I powered it down and came back, I'm back to square one!

The only thing which seems to fix it is if I change the Speed steps on the select to 28 rather than H5

[96RAF]

I would suggest a reset of that 03 loco by writing value 8 to CV8 in RM via the eLink. You may have to do this a time or two for it to stick. Then change its address to some other higher value, confirm this by reading it and see if the fault is still there.

[Hipwell]

So I've just done a test with an old analogue loco.

And you are correct, DC running is the culprit.

But, each time I power off the select and power it back on it looses the settings and reverts back to DC operation enabled.

Is this normal behaviour? Or will I have to turn it off each time I power up the select?

[Fishmanoz]

The Select reset is reverting it to 128 speeds steps each time, hence your fault re-appears. 28 speed steps is pretty much legacy these days and will stop directional lighting and TTS decoders from operating properly, hence the 128 default.

Turning DC Running on and off is a decoder setting in CV29. Can only be done via your RM/eLink.

I think I’d be returning the Select as faulty.