Reading & Writing CV's - Dapol Loco

[TrainDriverJ]

I have just bought in Hattons sale a Dapol (4D-022-010S) factory fitted DCC Sound Class 68 Avenger loco. I have Hornby eLink & Railmaster operated by a Win 10 laptop.

 

Try as I might I cannot get it to read the CV's of this loco. I do not have this problem with any other locos I've bought. (I have no other Dapol locos in my collection.)

 

Once entered into the database the loco does operate fine and function keys work etc. on the default CV 0003. I obviously, at least, want to change the DCC address and am reluctant to even try as not being able to read CV's back leaves fault finding or any other manipulation difficult to say the least.

 

Does anyone know if this is a known problem with this loco (or Dapol Locos) and /or a Hornby eLink controller? Is there anything I can do? Would it be possible to change CV's without reading them and would this be wise? Any ideas/suggestions?

 

Thanks. John

[David12]

Hi John

I have exactly the same configuration as you describe in your first paragraph, with the exception that the loco is not sound fitted, simply a Dapol Imperium decoder. I have had no problems reading or writing CVs - it is the best loco I own.

[96RAF]

If you know the decoder make/model, then in the reading screen you can double click on CV7 and pick from a drop down list, which will populate the correct CV set for that decoder into the list, then proceed to read them.

 

Also you can write a new address without first reading it. RM will verify the change and report it.

Rob

[Chrissaf]

In addition to Rob's suggestions above.

.

Unfortunately eLink only supports 'Direct Mode' programming where you have to use the 'Prog' track eLink output. The Elite on the other hand supports 'Operate Mode' (POM Programming On the Main) and 'Paged Mode' programming. These alternative programming modes can sometimes work when 'Direct mode' on the 'Prog' track does not.

.

In the absence of a different controller, all I can suggest is covering the basics.

.

Such as disconnecting the main track from the eLink when trying to program the loco on the 'Prog' track. Disconnecting the main track sometimes works because the main track can potentially pick up a lot of electrical noise like an aerial. Make sure that the 'Prog' track is only just longer than the loco, for the same reason.

.

Make sure all the electrical contact areas between the eLink 'Prog' output terminals and the 'Prog' track are spotlessly clean. Strip back fresh clean wires to ensure that any tarnishing is removed. Robustly clean the 'Prog' track rails, loco wheels and eLink 'Prog' output spring terminals. Apply a small amount of weight on the loco as it sits on the 'Prog' track to ensure the best contact between wheels and rails. Re-terminating the 'Prog' wires and cleaning the sprung 'Prog' terminals on my Elite worked for me. My preference is for wires that are directly soldered to the 'Prog' track rails, rather than using a track power clip. If you have to use a power clip, remove, clean all the contact areas and refit. This cleaning regime sometimes works because the 'Prog' output is extremely low power and isn't strong enough to burn its way through the build up of dirt and tarnish that occurs over time. Tarnish isn't always visible.

.

Also, if you use a power track / power clip double check that it is NOT a DC Analogue version. Prise open the cover and look for a capacitor wired across the rail terminals. Remove ANY component you find inside:

.

Photo from Rob.

/media/tinymce_upload/c8c0c976b0ba75d6b5e46cff5869b1dd.jpg

.

Your response might be that my other locos can have their CVs read and written to without issue.

.

That may very well be the case. Some decoders are more prone to these issues than others. I haven't been able to identify which brand of sound decoder Dapol fit at the factory. But the American 'Broadway' and 'SoundTraxx' brand sound decoders are particularly prone to programming reliability issues. So much so that a 'PowerPax Programming Power Booster' is available that connects to the DCC Controller 'Prog' output to boost the programming signal for the Broadway & SoundTraxx brands and all sound decoder fitted locos in general.

.

In the absence of a PowerPax upgrade to your eLink. You could try 'pre charging' the sound decoder on the main track first, before quickly transferring the loco to the 'Prog' track to try programming.

.

PowerPax

.

Why some sound decoders have programming issues and why the PowerPax resolves them is documented here:

https://www.model-railroad-infoguy.com/programming-booster.html

 .

I use a 'PowerPax' and can confirm it does what it says it will do and it is compatible with the Hornby eLink.

.

See the 5th post down on this page for eLink compatibility confirmation.

https://www.hornby.com/uk-en/forum/post/view/topic_id/7148/?p=3

.

I should stress that I cannot guarantee that installing a PowerPax will resolve your issue. For example there may be a fundamental compatibility issue with your fitted brand of sound decoder, but a PowerPax will give you the best chance of programming success.

.

This previous post of mine, documents my experience of using a PowerPax with my Elite:

https://www.hornby.com/uk-en/forum/how-to-get-robustly-reliable-cv-readingwriting/#post-213600

.

PS - It works just as well with non sound decoders too.

 

[96RAF]

On the very vague off chance it may work try turning the loco round to face the other way on track.

There was a problem some time ago with a batch of TTS decoders only working uni-directional with some controllers, so could it be the other way round this time, i.e. another make of decoder not working in both directions with a Hornby controller. 

Rob

[TrainDriverJ]

Thank you all for some excellent ideas and thorough help. This started when I bought this loco the other day. I had no trouble until then. So I naturally thought it must be this loco or a compatibility issue. Anyway today along with trying all your suggestions I thought I would try putting known working Hornby locos with Hornby decoders on the programming track and guess what it now can't read them either nor can it on any others I subsequently tried. Strangely though the locos pre-programmed before the problem started work perfectly fine in operation so you can continue to use the eLink for running only. I wish I had done this before posting as it may have saved you all your hard work. I'm sorry. It looks like it’s nothing to do with Dapol locos at all. Having followed the knowledge base in Railmaster which comes up when there is a problem it looks like the eLink is kaput. Oh well at least Hattons have got them in their sale for 30 quid. Again thanks everybody for trying to help. John

[Chrissaf]

Thank you for your concern, but it is not a waste of time. I have saved my reply to my forum boilerplate document so that I can use it again if someone else asks the same question in the future.

.

Before completely abandoning the existing eLink for a new one. Do try cleaning the eLink 'Prog A&B ' sprung terminals first. The way I did mine (on the Elite) was to take a small piece of fine 'wet n dry' paper and roll it into a thin tube. I then flattened it so that it was about 5mm wide. I then inserted it under the terminal clamp of the 'Prog A&B' output and manipulated the 'wet n dry' paper back and forth whilst clamped under the spring pressure. I then stripped off a bit of wire insulation to create a fresh bared wire end to the wires going to the prog track.

.

At the end of the day, it may be just dirt and tarnish that needs removing.

[TrainDriverJ]

Chris you really are persistent. And thanks for your enthusiasm. I must admit I had given up and was now just going to get a new one. But because of your subsequent reply I am going to have another go, exactly as you suggest.

Furthermore, I would cry if you turned out to be right when I had just shelled out another 30 quid. It'll have to be tomorrow though as I'm sat in front of the tele now watching a film and having a wee dram to console myself. Thanks again. Will of course let you know.

[96RAF]

If your elInk is definitely duff you have nothing to lose by having a look inside. There may be a dry joint you can fix at the Prog Track wire terminals.

 

See here... http://www.halton96th.org.uk/page14.html for info about taking it apart.

 

The likely culprits for loss of programming output are the tiny 8-legged components Q8 and Q9 which can be seen to the left of the eLink board just below the big resistor (R86). This gives you an idea of their size (Elite board shown).

/media/tinymce_upload/efd057bef0bf8e8065ddcb2c78270f44.jpg

 

You can buy these components for pennies, but I have tried and failed to change these out on my Elite and my soldering skills are very good. Not a job for the beginner to soldering.

 

If you have an led and resistor you could try that across the prog track terminals and see if there is any activity when you are programming - with nothing else connected.

 

Rob

 

[TrainDriverJ]

Chris. You nearly got a very large gold star. I cleaned the programming terminals as suggested. I must admit I was surprised but it worked. However, euphoria was short lived. I got up to CV 58 before it started saying unrecognisable CV again. I meticulously cleaned everything again but alas no joy. I can only conclude that there is some sort of intermittent fault within the unit. Thanks RAF96 for the further alternative I know it takes up your valuable time. I may have a look later but I really am getting fed up (time wise) to the stage of cutting my losses and shelling out. My soldering skills will be no were near that which could be accomplished by you - I’m sure of that. By the way I did find out it had a LokSound 4 decoder fitted for my trouble. Thanks everyone. I’ll let you know the final outcome.

[Chrissaf]

It was worth a try....I bet up to CV57 you must have been thinking "Well I'll be a ........... he was right".

[96RAF]

Humour me - let it cool off for an hour say, then read it again but from CV58 onwards. Set the range at the top of screen.

 

If it reads a few CVs then cuts out odds on it is one of the self resetting fuses i.e. F3 seen beside the right hand end of R86 in my Teardown report. Easier to replace than the other bits.

 

My advice, go for the new one anyhow and if you feel up to it try to fix the old one, else you could use it as is as Contr B for your points.

 

Rob

[TrainDriverJ]

Chris. You took the words……. I was doing a jig, promising you the earth and at the same time thinking how dumb am I.

Anyway, I had to leave it for the rest of the day especially after my happiness turned to disappointment. My better half is getting fed up of me as well to the point of offering me the £30!

I’m a glutton for punishment now and going to have another go tomorrow as per Rob’s suggestion. It’s weird how you just can’t cut your losses and move on. Probably a bad personal trait.

Good idea Rob. (It does work perfect from the track output.) I could use it for point control and can still actual play trains whilst all this is going on. I suppose to be fair though I am losing confidence in it a bit even if it does manage to get to the end. It’s just becoming a battle of wills now and I think it’s me that’s going to lose.

 Thanks again. John

[Chrissaf]

John, you started off just wanting to change the address. When trying again why not just go straight into address changing write mode. Then try the read function thereafter to see if the address write function took hold. I also suggest restricting the read range from 1 to 30 as most of the important CVs are in that range.

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

Just one other thing to check out, in case you haven't done this already. Assuming you are using Win7 to Win10. Check your 'railmaster.ini' file. The .INI file editor is the bluish COG icon in the bottom left corner of the RM Help Screen.

.

Ensure that these two lines are the very last bottom lines in the file. If they appear further up the file, delete them, and retype them again at the file end. Save the edit, close and reopen RM and try again. This has been known to occasionally fix weird obscure RM / eLink issues.

.

Alternative comms=1

Check controller=1

.

If the lines are missing, then add them. If either of the values =0 make them =1

But either way, make sure that they are the last two line entries.

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

The fact that you got as far as CV57 would seem to indicate that the issue is not permanent. It really is a pity that there is no 'try before you buy' option on the PowerPax. That would have been a very telling experiment to try. I am certainly not suggesting investing in a PowerPax purchase just on the off-chance.

.

It’s weird how you just can’t cut your losses and move on.

.

I know exactly what you mean. I hate to be beaten by a lump of technology, most times I do end up resolving issues eventually. Particularly software issues on my PC. But there are definitely a few times that tech hardware has been stripped for recyclable parts and the rest gone for scrap.

[TrainDriverJ]

I feel like Scheherazade with the next installment each night.

 

Well after its rest I had another go as suggested. Checked the ini file and deleted the 2 lines and re-entered them at the bottom. Rob, not had the guts to open it up but this may happen yet.

 

Wow. It took it a long time kept stopping occasionally on random CV’s to have to re-read them but it did complete. I tried a second time from the beginning and after a long time and many rereads it managed it. i agree that PowerPax would be a good experiment at this stage.

 

Now do I accept the challenge? Do I risk writing a new loco long number (8008)? Well I decided to go ahead. But before I did I made a note of all the CV entries – just in case. Not that big a gambler. It did it.

 

However it has caused me to have some follow up technical questions that if you don’t mind I would like you to consider and would like your opinions about please.

 

According to my notes CV29 had an entry of 114, CV17 192 and CV18 128. Assuming it had actually read them correctly. After the DCC address change CV29 was 46 CV17 223 and CV18 72.

Whilst CV’s 17 & 18 are as expected for the number 8008 I can’t get my head around the change downward to 46 in CV29. Why?

 

A check on the value 114 reveals an initial set up of 28/128 speed steps, Complex speed curve, Long Loco Address, and (bit 6) special not normally used. The value of 46 reveals 28/128 speed steps, DC Operation, Railcom, Long Loco Address. Both these values seem rather odd to me.

 

I thought I would try writing 114 back to CV 29 as this seemed to include the necessary value for long addressing. However the eLink won’t let me. It comes up with a dialog box that only values up 63 are allowed. Hmmm. Just when I thought I was out of the woods.

 

Do you think the original value could be a fake because of the read problems? But then again it is a correct possible value?

Do you think eLink can read higher values but not write them? Why would this make sense?

Do you think I will suffer any lack of functionality by having the new value or should I not be bothered? Perhaps you could enlighten me? Tried running it and as yet I can’t see what would be different. Seems to respond okay but not checked the many complex operating modes.

Just as a matter of interest what would you need to write a value above 63?

 

Ultimately I think I will have to get another otherwise I get the feeling I am going to end up in deep trouble.

[96RAF]

I have seen some odd values readback by RM. The value should reflect in the long string In the right hand box.

 

Can’t remember if you double click or right click the value or description box of CV 29 in RM to open the tick box of things you want it to do. Then you don't have to fret about working the numbers. Valid numbers are the unique sum of decimal values of bits 0 to 7 of CV 29

 

You obviously know about long addresses and CVs 17 & 18. If not there is a long address calculator on 2mm.org site.

 

You should be able to write any valid value up to 255 in any active CV up to 1024 in RM.

 

It is assumed you are using the latest version of RM downloaded from the RM page top.

 

If push comes to shove you can always write value 8 to CV 8 even though it says you can’t.

 

Rob

[Chrissaf]

Initial observation of your test results would seem to indicate that my suggested eLink terminal cleaning technique was in the main successful.

.

I would not be at all surprised that if you re-ran your tests using a NON sound decoder that you would get much better results. The in-rush current of Sound decoders can and do create CV reading errors with Hornby digital controllers, as my own test results seemed to prove conclusively. (see below).

.

In my page one reply, I wrote the following text, reproduced below to save you scrolling back to the first page:

.

This previous post of mine, documents my experience of using a PowerPax with my Elite:

https://www.hornby.com/uk-en/forum/how-to-get-robustly-reliable-cv-readingwriting/#post-213600

.

Granted my tests related to the Elite and not an eLink but the principles are the same. The results table showed that CV29 was one CV that was prone to having its value misread.

.

According to my notes CV29 had an entry of 114, CV17 192 and CV18 128. Assuming it had actually read them correctly. After the DCC address change CV29 was 46 CV17 223 and CV18 72.

.

In your last post on Page 1 you stated that the Dapol used a Loksound V4 decoder. All my decoders bar one are Loksound V4s......the bar one is a V35. IIRC, Loksounds do support 'Railcom'. The CV29 114 value was most likely a misread, whilst CV29 value 46 may have had 'Railcom' set because RM knows from its decoder database that a Loksound V4 supports it. I have a vague recollection that my last loco purchase with a Loksound V4 had 'Railcom' Bit 5 enabled. I took it out when I set my CV29 to my own preferred value which is 2.

.

I concur that the CV17 & 18 values are correct for a long address of 8008. I don't use 'Long Addresses' myself, but CV17 192 is what I find as being normal for my Loksound V4s when using a short address (less than 128). My value of CV18 varies.

.

If you take the 'Railcom' support out of CV29 so that value 46 becomes value 38, then 38 is the value I would expect to see for CV29 with your particular 'Long Address' configuration. But 46 is equally valid, it would just be that the 'Railcom' part of the decoder would lie idle if it wasn't being used.

.

I agree that PowerPax would be a good experiment at this stage.

.

Everything in this last set of test results you have provided is suggesting to me that a PowerPax has the potential to assist your set-up greatly, particularly with Sound decoders. But I appreciate that is a financial risk that might not pay off. It will at the end of the day be your 'checks and balances' decision whether you want to invest in one or not. On the 'pro' side, it is compatible with virtually all DCC controllers so shouldn't go to waste if you change controller later.

.

However the eLink won’t let me. It comes up with a dialogue box that only values up 63 are allowed.

.

I have heard of this being reported on the forum before, but have not seen this dialogue box myself. But that may be because I have not trying writing a CV29 value higher than 63. Since it is not something that has affected me, it is not something that I have previously looked into further......see EDIT below for update.

.

PS - Follow up to what Rob said. What version / revision of RM are you using? The version number is shown in the little yellow box near the screen clock. If you hover the cursor over the yellow box, the revision number shows in a little pop-up box.

 

Extended CV support is a fairly recent RM update.

.

EDIT: Following extract taken from the online CV29 calculator.

/media/tinymce_upload/57a3f442239587381ba9454cddc974df.jpg

.

What this table shows is that Bit 6 (64) and Bit 7 (128) are not usually used (as commented in the right hand side table column). Of the Bits 1 to 5 that are left. If ALL five Bits are enabled (yellow highlight), then the resultant CV29 value is 63 (also highlighted in yellow). Thus it would appear that in the Hornby controller world, only values up to 63 are supported in CV29. Thus values 1 to 63 cover all the Hornby RM supported CV29 permutations.

.

It therefore may be, that if some non Hornby decoder manufacturer wants to make use of these previously unused bits (Bits 6 & 7) for some unspecified custom decoder feature, then it may be that the Hornby RM controller wont be able to modify them.

[TrainDriverJ]

Thanks Chris & Rob again. I'll digest everything you have said and have a further think and let you know anything of further value. I like your reasoned conclusions and evidence. Looks like the value 114 is actually valid but probably erroneous.

 

I accept that a value above 63 would probably never be needed as bit 6 and 7 would not normally be used and I don't know why you would use them. I'll try and get a screen print of the tiny box that comes up just for the sake of a having a knowledge base for the future of RailMaster limitations.

[TrainDriverJ]

The revision is 1.69 rev 0. It does say each time it starts that it is checking for updates but since 1.69 rev 2 is at the head of the forum I will manually update. Thanks.

[Chrissaf]

The auto-update only detects version changes and not revision changes. Revision 0 had bugs and was quickly followed by revison 1 and then revision 2. From my records: Revision 0 was released 5th June, Revision 1 on 15th June and Revision 2 on 18th June 2018).

.

Revision updates have to be done manually.

.

It is rumoured that version 1.70 will be out soon, based upon my experience of previous years I would expect it this side of Christmas.

.

Just be aware, that the revision pop-up box can't always be relied upon 100%. Just sometimes, HRMS release an updated revision but forget to modify the code that displays the revision number when you hover the cursor.