Tale of 2 Class 20s and an Elite Controller

[Chris1707820829]

I've had one Bachmann Class 20 for ages with no problems powered via my Elite controller. Thought I'd be clever and buy a second for that classic 'nose to nose' running. The second Bachmann class 20 was Hornby TTS-fitted when I got it.

First off, the TTS decoder flatly refused to be reprogrammed as anything other than 0003 despite all indications (clicks and red lights) that the Elite was writing the new running number. I decided to put the TTS decoder to one side for now (anyway, even if this worked, trying to run this with sound on in a consist would have meant the non-sound loco straining to go whilst the TTS version went through its pieces before moving!).

So for consistency I bought them both new bog-standard Lenz decoders of the same type, my thinking being that I just needed to change the direction of one of them, programme them with separate IDs and add them to a consist.

Here again, I went through the motions of following the Elite's flow-chart to reverse the direction and got nowhere. Nothing seems to save, no clicking, no red lights.

I tried the CV29 path and also got nowhere.

Every time I tried to add my locos to a consist, they were still playing tug-of-war.

In the end, I turned round one of the 8-pin decoders, there being no lights to worry about and called them both 0020! Yes it runs but I can't believe this is the only way of doing it since this prevents singling out one of them whilst the other stays put.

Yes, I'm doing this on a separate programming track fed from the programming terminals.

To quote several other posts, what am I doing wrong?

[RB51]

Hi Chris -welcome to the forum. One of our experts will be along shortly with some help.

This new forum set up does not require you to format your post in paragraphs and doing so puts a huge space between them. I have reformatted your post and if you use the three vertical buttons top right - the ellipsis - you can enter EDIT mode and see how your post is set out. Quite different from how it appears when saved. R-

[Fishmanoz]

Hi Chris, welcome to the forums and, without intending to be condescending, can I congratulate you on two aspects of your first post which should be simple but seem to be beyond some. The first aspect is that you’ve posted in the correct sub-forum. The second is that you’ve used an attention-grabbing title that provides information about your problem.

Reading your problem, it firsts seems that there may be a problem with your Elite’s programming output - all efforts at programming IDs and writing a new value to CV29 have failed. However, if I understand you correctly, you’ve then been able to change both IDs to 0020. Am I correct?

The only thing I can see possibly different to your earlier failures is that maybe you put both locos on the programming track and changed both IDs with a single programming operation. Was that the case?

Irrespective, I’m at a loss to understand how you may have done anything different in changing IDs to 0020 versus your earlier attempts and therefore at a loss as to why those earlier attempts failed.

[Chris1707820829]

Just to clarify further, the only ID that just could not be changed at all was that of the TTS chip - on another FB page related to model railways, someone who 'puts chips in for a living' tells me that's not unknown. Anyway, I've more or less written it off especially as it wouldn't function as I wanted it to in a consist. All attempts to programme were carried out on single locos, even when telling the two 20s to be 0020.

After a bit of 'sleeping on it' combined with a bit more 'RTFM', I've now ascertained why I thought the two 20s were refusing to be programmed. It was because I'd already included them, i.e. 2002 and 2003 in a consist called 20. I don't think I'd realised (having never had need for a consist before) that you have to dissolve the consist before they'll answer to their own IDs again. Hence I threw the towel in somewhat prematurely and called them both 20.

I've now set them up with their 'proper' IDs and created a consist to drive them 'nose to nose' which was what I wanted all along.

However, I've still got two matters unresolved.

1. Why the TTS chip steadfastly refuses to be anything but 0003 and
2. Why my efforts to programme the direction of travel of one of the 20s to be reversed relative to the other have come to nothing. OK, I've worked round this by reversing the 8-pin chip but should I ever have to do this again with a loco that has lights, I'd need to know how to do it.

Many thanks for your reply.

[96RAF]

There was a small batch of Class 20 TTS decoders which had programming problems including the master volume which shifted up two CVs.

You can check if your decoder is affected by trying to alter the master volume and see which CV it responds to. From memory CV182 goes to CV184. Set value zero using programming on the main with sound on and see if it mutes. Set it to value four to go back to default volume.

If necessary contact Hornby and arrange a return for reprogramming to fix it.

[Fishmanoz]

To reverse the direction of travel, add one to the current CV29 value and program that back to CV29.

Default values for CV29 are usually either 2 (128 speed steps) or 6 (128 plus plus DC operation). The only other usual change is if you have programmed a long address to the decoder.

Anyway, if CV29 is an even value, add 1 to the value and write back. If it happens to be odd now, then subtract 1.

The other way to change the direction is to swap the orange and grey wires from the decoder to the motor. If these are the correct way around, the loco will go forward when CV29 is even. If they are reversed, the loco goes forward when CV29 is odd. Hope that makes sense?

Of course, none of this swapping should be necessary if there is a socket in the loco and you install a decoder the correct way around in that socket. Although miswiring of sockets is not unknown. But none of it explains not being able to change the default direction.

[Chris1707820829]

I think I may have stumbled on what was wrong. I gave my programme track a jolly good clean and it now seems I'm able to make it do what I want gain. I'd used a junk piece of Hornby steel track for the purpose which is pretty quick to tarnish. Whether that really is the solution is anyone's guess but since cleaning it, I've been able to access CV29 and 'add another digit' to the 0000 that presented itself. The Elite made all the right clicks and flashes leaving me much more confident that it's actually saving what it's been told to.

Then of course I had to re-reverse the chip so it faced the correct direction. I now have my consist of nose-to-nose Class 20s with no silly work-arounds like reversing a chip.

I don't suppose this will help with the old TTS chip I took out, but if I ever do get it working again, I'll probably sell it.

[Chrissaf]

"Whether that really is the solution is anyone's guess......"

Absolutely... track condition is part of the solution.

The programming track has very low power applied to it from the controller. DCC is extremely sensitive to track condition, this is even more critical on the programming track due to the low power level used. I use 'Nickle Silver' track for my 'programming track piece' but I still give it a thorough clean before each and every use of it for programming tasks. Steel rail will be even worse IMO.

[ColinB]

A good idea is before you do any programming get the Elite to read a CV like Manufacturer's Id or the loco address, then you know that the Elite is seeing the loco. Generally I have found Hornby TTS, Dapol, LaisDCC and Zimo decodes work perfectly OK with the Elite, for some reason I cannot get Gaugemaster OMNI and some DCC concept decoders to program with an address.