Decoder R8247 problem

[Nidge]

Hi, I recently purchased a R8247 decoder and configured it for Peco Pl-11 side mounted turnout motor. It initally worked but I have notice once I power down and then power up both Elite and E-Link controllers, the configuration settings have not been saved and I have to go through the whole configuration process again.

Interested to know if any of you guys have had the same problem. Is there a fix or have I got a duff encoder?

 

Thanks

[Fishmanoz]

Welcome to the forums Nidge.  Certainly used to be an 8247 problem.  I would take it back. 

 

Just one thing before you do though: you don't have anything causing shorts on the layout, even momentary ones?  This is a known cause for decoders resetting. 

[96RAF]

You could try a reset by writing value 8 to CV8 just like a loco decoder, which will put the accessory decoder addressing back to default i.e. module 1 addresses 1 to 4.

 

When you say you have to set it up all over again are you using an address other than 1 to 4 for convenience of your layout. 

 

What symptoms do you see when you repower after a layoff, the point just doesn't work or the chosen address doesn't respond, if so does it respond to an address within the range 1 to 4.

 

I notice you say both Elite and eLink controllers. I presume you are using RM and have the Elite controlling locos (Controller A) and the eLink on a separate bus controlling points (Controller B). Please confirm.

Rob

[Nidge]

Hi Rob,

At the moment I have one bus and using the Elite and E-Link separately. My plan at some point is to use both as controllers A & B.

I have 3 x R8247 decoders addressed as 61-64, 65-68 & 69-72. Two decoders with addresses 61-64 & 65-68 work fine whether using the Elite or E-Link. Decoder addressed as 69-72 only operates when configured initially. If I power down and then up, decoder with addresses 69-72 does not operate the Peco PL-11 points.

To me it looks as though this decoder doesn't retain the configured addresses when powered off.

Any suggestions would be welcomed.

Nidge

[Chrissaf]

I had a R8247 that would not keep its addresses after a power down. Just keeping on re-addressing it, didn't work. Writing 8 to CV8 to perform a factory reset to addresses 1-4 then re-addressing the ports to the addresses that I wanted improved matters no end. However, I only say improved matters. The errant decoder would still lose its configuration at the slightest provocation. I ended up replacing it with a new one (it was out of warranty). It initially started losing its addressing after a momentary track short circuit. If you do put the eLink as controller B and dedicate it to operating just the points, then those R8247s connected to it will become isolated away from possible transient track short circuits. That will help protect your R8247s from being affected again in the future.

.

As others have said, the R8247 is prone to getting configuration damage if the layout is subjected to momentary short circuits. Once the decoder is damaged by a short (as per my example above) it becomes weakened and prone to any event triggering the fault.

.

RAF96 above opened his reply with a suggestion to reset the decoder. It would appear that you have not yet adopted his advice, which is why I have suggested it again and used my own experience to emphasis it.

[96RAF]

Try writing the GROUP number to CV1 using the Elite, instead of writing the first address..

Group 18 is equivalent to your 69-72 address range.

This method is described in the manual in the bit where it talks about inputting 513 from memory. I dont have the manual in front of me to refer to.

Rob

[mikejj101]

I have an R8247 that keeps reverting to -3,-2,-1,-0. Really annoying.

[96RAF]

Mike

Your problem is the typical points address offset seen when using an Elite in Classic mode, but as your signature states you have eLink then its back to  - 'sorry I don't know'.

Rob

[Chrissaf]

I have an R8247 that keeps reverting to -3,-2,-1,-0. Really annoying.

.

If you look at my reply a few posts up. I didn't mention it in the post, but the reverting to -3,-2,-1,0 is EXACTLY what mine used to do. My only permanent solution was to replace it with a new one. It was damaged by a transient short circuit (R8247's are susceptible to short induced damage). As well as replacing the faulty one with a new one, I placed a home made snubber across my DCC BUS at the location where my three R8247s were connected. In three years, I have not had a single reoccurrence of the fault as a result.

.

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

[96RAF]

Avoiding loco induced shorts is the main reason I have my points decoders operating on a separate bus by way of the eLink Chris and my original four decoders are still working fine to date, including an old R8216 model.

Rob

[96RAF]

It may be that this negative addressing could be a genuine snag and may not be a complete failure.

Anyone seeing the fault is advised to contact HCC to report it in.

Rob

[Philx]

I have a problem with this decoder. I am trying to get it to run four pairs of point motors. I connect first and second pairs works fine. I then connect a third and all three stop working. I disconnect the third and the first two work again. I have tried using different point motors directly wired into third position to eliminate wiring  problems. It is way out of warranty so no going back But I don’t think I have ever used it with three sets of points before. I am using an elite controller.

 

Any advice welcome

[96RAF]

You can reset an accesssory decoder just like a loco decoder by writing value 8 to CV8.

Give it a try then set your points again.

Rob

[Philx]

You can reset an accesssory decoder just like a loco decoder by writing value 8 to CV8.

Give it a try then set your points again.

Rob

Thanks Rob do I need to attach decoder to programming track connections on elite to do this?

[96RAF]

...not quite sure how else you could do it...

as stated there may be an real snag with negative addressing worth contacting HCC about.

[Philx]

...not quite sure how else you could do it...

as stated there may be an real snag with negative addressing worth contacting HCC about.

Thanks again 

[Chrissaf]

My experience of using R8247s is that once they start loosing configurations or having faults where you have to resort to factory resets to recover [write 8 to CV8 on the programming track interface], is that the factory reset fix rarely stick. The most common reason I have found for a perfectly good R8247 to suddenly start playing up, is because the DCC track (or BUS) it has been connected to has had transient short circuits on it (loco shorting on points for example). These transient short circuits create very high spike voltages across the DCC signal, and these spikes damage R8247s for which they are prone to being sensitive to. This invariably results with the errant R8247 finally being replaced with a new one.

.

You can protect your existing and any new purchase R8247s from being damaged by DCC voltage spikes by fitting an RC filter (Snubber) close to where the R8247s are connected. I have not had a single R8247 issue since I fitted my two home made snubbers more than 3 years ago.

.

Another solution is to put all the R8247s on their very own isolated DCC circuit via controller B (eLink), but this alternative controller B solution only applies to when using RailMaster (RM) with the Elite as controller A. Snubbers are a hell of a lot cheaper than buying an eLink, but using the eLink as controller B is a 'cost effective solution' if you happen to have an old eLink in the bottom draw you can use (plus RM of course).

[Philx]

My experience of using R8247s is that once they start loosing configurations or having faults where you have to resort to factory resets to recover [write 8 to CV8 on the programming track interface], is that the factory reset fix rarely stick. The most common reason I have found for a perfectly good R8247 to suddenly start playing up, is because the DCC track (or BUS) it has been connected to has had transient short circuits on it (loco shorting on points for example). These transient short circuits create very high spike voltages across the DCC signal, and these spikes damage R8247s for which they are prone to being sensitive to. This invariably results with the errant R8247 finally being replaced with a new one.

.

You can protect your existing and any new purchase R8247s from being damaged by DCC voltage spikes by fitting an RC filter (Snubber) close to where the R8247s are connected. I have not had a single R8247 issue since I fitted my two home made snubbers more than 3 years ago.

.

Another solution is to put all the R8247s on their very own isolated DCC circuit via controller B (eLink), but this alternative controller B solution only applies to when using RailMaster (RM) with the Elite as controller A. Snubbers are a hell of a lot cheaper than buying an eLink, but using the eLink as controller B is a 'cost effective solution' if you happen to have an old eLink in the bottom draw you can use (plus RM of course).

Good idea where do I get one? Sorry to be dense i am returning to this hobby after some 40 years and am on a steep learning curve 

[Chrissaf]

Where do I get one what? eLink or snubber.

.

Can I humbly suggest that next time you do not use the blue button. This is not a 'Reply to this post' button. It is best if you write your reply in the 'Reply to this post' text box at the bottom of the page and click the green 'Reply' button. Using the blue button has just duplicated my long reply unnecessarily onto which you have added 1.5 lines of new text.

.

Not having a go, using the blue button is something all relative newbies on the forum seem to do for some reason. I blame the forum SW developers for putting it there, just where a 'Reply' button is expected to be.

.

If you mean a 'Snubber', they are very simple to make. The link I included in my last reply gave full constructional details via a further link within it. But here is a simplified drawing showing snubber construction and connectivity appropriate to your particular issue. The 100 ohm resistor will get warm....very warm.....so I recommend using a 2W resistor as the larger surface area will run cooler to the touch. The R C components can go either way round, they are not sensitive to polarity. In your case you only need one single 'Snubber' protecting the R8247s if wired as I've shown it. The 'Snubber' will also provide a certain amount of protection for the loco decoders as well against 'short circuit' damage.

.

/media/tinymce_upload/62b75979ad33bb15a891ac7a47ccaeb0.jpg

.

Maplin Codes:

100 ohm 2W Resistor = Maplin Code D100R @ £0.20p

0.1uF 100VDC Ceramic Capacitor = Maplin Code N43CJ @ £0.09p

.

The 100VDC capacitor voltage rating is higher spec than the 50VDC one shown in the drawing.....so not an issue as an alternative substitution.

.

You won't be able to buy a commercial version of this at the same component voltage / power rating for just a mere 29p anywhere. And Maplin are relatively expensive for their components, I only quote them here for their High Street convenience (saving ordering online P&P charges).

[mikejj101]

I have an R8247 that keeps reverting to -3,-2,-1,-0. Really annoying.

.

If you look at my reply a few posts up. I didn't mention it in the post, but the reverting to -3,-2,-1,0 is EXACTLY what mine used to do. My only permanent solution was to replace it with a new one.

That's what I have now done - thanks to Santa !

[96RAF]

I have an R8247 that keeps reverting to -3,-2,-1,-0. Really annoying.

.

If you look at my reply a few posts up. I didn't mention it in the post, but the reverting to -3,-2,-1,0 is EXACTLY what mine used to do. My only permanent solution was to replace it with a new one.

That's what I have now done - thanks to Santa !

 

And I would still contact Hornby about the original problem with the unit as maybe they will replace the duff one.

Rob

[Fishmanoz]

May I humbly suggest that the only thing that is keeping this product in the market is the Hornby logo printed on it. The reliability and design of what’s inside certainly doesn’t warrant a place in the current accessory decoder market. It’s old, tired and needs to be replaced with a real one.

[Chrissaf]

Ditto that Fishy.....