Loksound V4 & V5 reprogramming problems

[Stuart G]

I have been re-addressing all of my locomotives (Steam, Diesel & D.M.U.) and have done just over 100 with no problems.

The decoders include Hornby R8249, Hornby T.T.S. (various models), E.S.U. (LokPilot, LokSound V4 & V5), Hattons, Uhlenbrock, Zimo (various models) Soundtraxx etc.) - all fine except for three E.S.U. LokSound decoders.

These three are fitted to a Bachmann Class 40 (factory-fitted), a Bachmann (N.R.M.) DP1 (factory-fitted) & a Hornby Class 8F (retro-fitted).

I can read all C.V. values correctly, but when I try to re-address these locomotives, the decoders seem to accept the change but do not respond to the new address, they only respond to the 'old' (previous) address; i.e change from address 78 to address 77, loco. will respond to address 78 despite the decoder appearing to accept the change. 

I have tried to reset the decoders, CV8 = 8, but the address does not revert to address 3.

I have full sound & motion with the 'old' addresses and E.S.U. decoders in other locomotives have not been a problem.

I have even confirmed that CV 19 = 0, plus other suggested checks.

Can anyone shed light on what is going on, please?

[BoD]

Might the decoder be locked.  
 

I’ve not used this myself but this involves CVS 15 and 16 and is explained here.  https://dccwiki.com/Decoder_Programming_Lock
 

 

Edited March 5, 2023 by BoD

[55020]

I think this might be due to the way ESU handle codes less than 128, in that you have to set the decoder to accept "short addresses" only.

 

Activate this by changing CV29.  However it is only the 5th bit of CV29 that needs to be set to 0.  DO NOT set CV29 to zero!

 

If you are using software like LokProgrammer to change your decoder settings, this is easy.  Just tick the box that says "use short address".

 

If you are manually writing CVs, let me know what CV29 currently is for one of the decoders you are struggling with and I will work out the amended value for you.

 

You should then be able to update the decoder address to say 77 by setting CV1 = 77.

[Stuart G]

I've double-checked the settings for the following C.V.s:

CV15 & 16 = 0 on all three decoders

CV29 = 31 on the DP1 & 8F decoders

CV29 = 12 on the Class 40 decoder

 

I have tried writing the new address directly to CV1 but this makes no difference either on all three decoders.

 

I am using a Bachmann Dynamis Controller at the moment; however I have a Hornby Elite/Railmaster as back-up - I will try using this just in case there is an issue with the Dynamis system.

 

My other options are to send the decoders away for 're-blowing' (and hopefully re-addressed to my chosen numbers), or

new decoders (which is definitely a last resort!!).

[55020]

Stuart, the Class 40 value is a little odd as it doesn't have speed steps 28/128 enabled i.e. it is set to only have 14 speed steps.  Anyway your problem isn't the short/long address issue I hinted at, as all three are set up for short addresses.

 

All three are also set up for Railcom, which can sometimes have odd effects.  That would be my first suggested change, just turn it off.

 

Let's take CV29 back to it's default setting of 6

(which gives 28/128 speed steps and DC analogue control, short address, Railcom off))

 

Write the change, read the CV to make sure the change has taken effect, then try changing the loco address.

 

 

Steve

[Nigelcliffe]

12 hours ago, Stuart G said:

I've double-checked the settings for the following C.V.s:

CV15 & 16 = 0 on all three decoders

CV29 = 31 on the DP1 & 8F decoders

CV29 = 12 on the Class 40 decoder

 

 

Do you know which are LokSound V5 and which are LokSound V4 ?   Because the legal values in CV29 are different in each type. 

 

CV29=31.   Would seem plausible for a LokSound v5, and is set to short address mode.   However, it has direction reversed, which strikes me as odd for a RTR loco (unless decoder is 8-pin with the plug rotated).    LokSound v5 cannot have CV29=6 , it is not permitted to set to "3 point speed curve", the 28 point curve is compulsory in the v5 (multiprotocol European version, the US-version "DCC only" does allow 3-point curve).  

 

CV29=12.  That is the factory default value for a LokSound v4.   The decoder will auto-detect speed steps (see ESU manual for a V4).    

 

 

I'd be trying to show that a decoder write actually works, and can be read back.   There are the user-value CV's of CV105 and CV106.  You can write anything to those, and then read it back (power cycle the entire DCC system to be really sure).  That would show read and write are working.   

 

 

- Nigel

[55020]

Nigel, ahh yes I didn't pick up on the reference to V5s in Stuart's original post.  And then we throw in the "full fat V5" or "V5 DCC" just to complicate things a little bit more.

 

In order to keep things simple I do think the first thing we need to eliminate is the Railcom feedback (subtract 8), as this is obviously enabled at the moment.  Keeping the other bits set as is, this would result in the following CV29 values being written:

DP1 & 8F   CV29 = 23

Cl.40           CV29 = 4

 

Stuart, maybe try the Class 40 first (a V4?).  Adjust CV29 as above, write and read the decoder, then power down and back up again as Nigel suggests and read one more time.  Hopefully CV29 still reads as 4.  If it does, go ahead and see if you can adjust the address using the same power down/up cycle.

 

 

Steve

[Stuart G]

So far, things are on the up!

I changed CV29 to 18 on the Class 8F (V5) & Class 40 (V4) and, lo & behold, I was able to re-address BOTH locomotives.

The Class 8F & Class 40 are now fully operational - movement & sound.

Thanks to all for your help so far.

However, I have now encountered an additional problem with the other V5 in that I can neither read nor write to any C.V.s.

I have removed the decoder from the locomotive and checked it on a Decoder Tester (D.M.G. Technical) - this tester has worked before on all the various decoders that I employ in my locos.

When the decoder is attached to the tester with it connected to the Test (Programme) Track source I notice that the F3 & F4 L.E.D.s are on all the time, except when there is communication between the Controller and the Tester - is this just a function of the tester or something else??

[55020]

Stuart, I don't have any knowledge of the the decoder tester you have, so offering any specific advice would be nothing but a guess.  However it doesn't sound good (!!)  regarding the health of your DP1 V5. 

 

ESU are good at providing warranty support and will generally fix or replace defective decoders at no charge.  I used to go directly to ESU in Germany, but I suspect the customs border may now generate issues.  Perhaps others might be able to provide some info about the post-Brexit options?

Your best bet initially would be to contact South West Digital, who are the UK distributors.  I not had much success in getting them to respond to emails in the past, but give it a go.  However you might be better off calling them.

 

Finally, I've just spotted you're in Shetland.  I've spent many many days working out of Sumburgh over the last 25 years, so know the islands well.  I have a ESU decoder testor and Lokprogrammer, so wonder if you might want to get the decoder to me in Aberdeen.  I can easily arrange for it to be hand carried to/from Sumburgh if you'd be happy to consider this.

 

 

Steve

Edited March 9, 2023 by 55020

[Stuart G]

Steve,

I appreciate your offer to have a look at it for me.

I could post it to you if you could pass on your address (e-mail: stujgow@hotmail.com), I have suitable packaging for it to go by post.

[Stuart G]

I contacted Roads & Rails and they suggested that I send the decoder to them to see if they could fix it.

Just received a reply from Richard (@ Roads & Rails) saying that he has managed to reset the decoder and got it working again!!

The sound file needs to be re-loaded and they'll send it back once that's done.

£33 for repair as opposed to £95 for a new decoder - I am very happy.

Thanks to everybody for your assistance - much appreciated.