R8247 accessory decoders problems

[Dai bach]

Help, I have 6 decoders on my layout. Worked fine for a while now only 2 work?

I have read some comments on the forum how to re set these devices using cv8 to get back to factory re set. Once re set I try to allocate a group address as per the handbook ,nothing, readback on re set and after allocating a group address come up with xxx. This is on all decoders . What am I doing wrong? Have even tried re sething Elite but nothing changes.

Are the decoders faulty 

Has my Elite got a glitch

Af Nearly £50 each it could cost me a lot to replace 

Any ides or solution out there before I scrap them and replace with other makes.

Dai

[Chrissaf]

Do you get or have you previously had a relatively frequent number of transient short circuits. The kind of shorts that you get when wheels bridge opposite DCC phase tracks when traversing a cross-over or points. Or when accidentality running a train into a point set the wrong way. Or any other train operating scenario where accidental transient shorts have occurred. Transient as in brief. Did you get a short on your layout just prior to your R8247s failing.

.

The reason that I ask, is that the R8247 is incredibly sensitive to these type of shorts which can cause corruption of the R8247 configuration. Such that re-addressing is required to recover. And in the worst case scenario severe corruption that just cannot be recovered from.

.

Writing 8 to CV8 on the R8247 will perform a 'factory reset' but only to R8247s that have not been totally corrupted by shorts. A 'factory reset' if successful will revert a R8247 to DCC addresses 1 to 4 which equates to Group 1. I would try the 'factory reset' process again, then test using DCC addresses 1 to 4 and confirm working on those addresses before trying to reinstate the addresses you used before. Even if you can use a 'factory reset' to recover a corrupted R8247. I have found in my experience that the R8247 that has been recovered becomes even more sensitive to shorts than before and will lose configuration at the drop of a hat.

.

Remember that the R8247 must be removed from the DCC feed and connected directly as the sole attached device to the Elite 'PROG A&B' output to write 8 to CV8 or to re-address it. They cannot be left 'in situ' on the DCC track feed when performing these tasks.

.

The R8247s can be somewhat protected from short circuit damage by installing a 'snubber' on the DCC feed that connects to the DCC input of the R8247. Only one 'snubber' is needed across a shared DCC feed to the R8247s.

.

A 'Snubber' consists of two simple components. A 100 ohm 2 watt resistor in series with a 0.1uF Ceramic Capacitor [50 volt rated or better], wired across the DCC feed to the R8247s - see drawing below.

.

/media/tinymce_upload/f7b01d626ede7a98194d9b64b1fdf402.jpg

.

Note that fitting a 'snubber' after 'the horse has bolted' will not repair your R8247s if they cannot be recovered, but it can protect whatever you end up replacing them with against future short circuit damage and is worth fitting as an 'insurance policy'.

.

Want to read more on 'snubber' theory, then follow the link below:

https://sites.google.com/site/markgurries/home/dcc-general-best-practices/wiring-planing/snubbers-rc-filter

.

TIP: I write long posts. If you intend to write a reply, it would be appreciated if you didn't use the 'White Arrow in Blue Box' button. This is not a 'Reply to this post button. It is best to write any reply you want to make in the 'Reply Text Box' at the very bottom of the page and click the Green 'Reply' button.

 

Particularly as my reply includes an image. If you use the blue button, any reply you write, may be held back for image approval. Even though it is already a previously published image.

.

See also – further TIPs on how to get the best user experience from this forum.

https://www.hornby.com/uk-en/forum/tips-on-using-the-forum/

 .

[Dai bach]

Chris 

Thanks for the quick response . It is a concern to me that if this shorting is a well known factor then why dont Hornby supply or build in a safety component either as an add on or in the decoder itself? It would appear that my decoders are kaput, unless someone has  a  fix?Anyone recommend another good make of decoders? 

Thanks 

Dai

[Chrissaf]

I believe that there is a rumour of a replacement R8247 in the pipeline.......what timescale I couldn't say.

.

Anyone recommend another good make of decoders?

.

Not easy, as there is a 'forum rule' against active promotion of competing branded products on this forum. Without mentioning any particular brand, I would look at brands that offer 'self learning' accessory decoders. If you read enough of the posts on this forum about 'accessory decoders' including the non Hornby ones. The conclusion that I would make is that they all have their own issues in some way, shape or form, else they wouldn't get mentioned negatively so often in these threads.

.

If you read the article on the link I provided, you will see that all decoders can potentially be affected by 'short circuits', not just Hornby ones. Hence the 'insurance policy' recommendation of installing a 'snubber'.

[Dai bach]

Hi Chrisaff

Further to our conversation above I decided to persevere and tried my selecti  controller to set up the R8247 rather than the Elite decoder. It worked all 4 decoder were re set with new addresses and work with the Elite. The down side is that the numbers start at 61 in groups of 4. So a good outcome, was it me or the controller at fault.

PS, the address on the Elite still comes up xxx 

Any ideas

Dai

[Chrissaf]

A bit too early to give a definitive conclusion. The 'prog' output on the Elite has known to fail, but you can't assume that it has on this occasion until you find that some more DCC products can't be configured......loco decoders for example.

.

If you can configure loco decoder CVs OK, but still can't configure a R8247. Then it may be your configuration methodology that is incorrect. Since you can program with the Select, then that would, on the surface, seem to discount the R8247 itself as being at fault.

[96RAF]

Edit: I was replying to this much earlier but the website crashed, so I went and did something else. Now I have returned and my original posting is still here so I may as well let it stand, noting that Chris has suggested some of my suggestions.

 

Try re-addressing a loco with your Elite as it sounds like the programming output may have died.

 

When you use Select or Elite you can program either a group or port address and the controller will work out which one to use.

 

You could try programming CV1 in Operational mode on the main, but I would advise you to disconnect the decoder from your track and wire it direct to the Elite.

 

The Elite will still operate the points on those Select addresses if you can put up with that range.

 

I presume you are using MENU-ACC-DIRECT-ADDRESS-WRITE. You can address as a loco or CV1 using the port address.

[Dai bach]

Hi both 

Thanks for the responses, as advised I tried to re address one of my locos with  a new number using the Elite it did not work it retained the original numberso as you suggested it looks like the elite programming output has died!

 

Can this be repaired in any way? 

Thanks 

Dai

[Chrissaf]

Contact 'Hornby Customer Services' to discuss with them. Look to the 'Contact Us' link at the bottom of every web page for details. They will give you a 'Returns Number' to quote on the package (Elite) you send them. Many on the forum say that repairs have been turned around within a week, some at no additonal charge to them other than the postage cost of sending.

[96RAF]

This may help with diagnosis, although I haven’t tried it myself yet.

 

The Elite Prog output shares circuitry with the Boost output so you should be able to read an output on the Boost output using your multi-meter set to AC volts scale. Trying to read off the Prog output is not possible as there is only an output there during actual programming bursts, which are unlikely to show on your meter.

 

When programming you can hear a relay clicking in time with the red led. That is the Prog and Boost circuits swapping over.

[Dai bach]

Thanks again both,  will try your suggestions and get back to you 

Dai

[96RAF]

Here is a scope trace of Elite Xpressnet output (green) compared to Boost output (yellow) and Track output (blue). As you can see they are all synchroinsded, but the object of the exercise is to show you the Boost output is continuous (except when Prog interupts it - not shown here) thus allowing you to meter it, which would indicate data is flowing from the main PIC chip past the relaxed relay. Note the ringing (spikes on top and bottom of the square waveform), probably due to not having a DCC load on the Elite just the scope probes.

/media/tinymce_upload/34b44c340ca17e00107fa48ddb43d43a.png