R8247 Points and Accessory Decoder Stops Working After Short

[Dowse101]

Hi All (Merry Christmas)

Just installed my 5th R8247. CV1 set to 5 (points allocated to 17, 18, 19 and 20). Using JMRI software and DCC++ base station. All worked fine until a loco derails and causes a short. I remove the cause of the original short and turn on power again only to get another short, repeat pressing the power button a couple of times until randomly the short is gone and power stays on (I'm convinced the decoder causes these because I can disconnect it and power stays on immediately). Then everything is working, except points on decoder 5 are unresponsive.

I stuck the decoder on the programming track and it still says 5 on CV1, so it hasn't "forgotten" its address. Just to be sure I reset it by writing 8 to CV8, reprogrammed CV1 to 5 and installed again, it works but the same thing happened after another loco short.

Suggestions?

I got this from that auction site so not easy to return. I'm only buying from authorised retailers in future, another decoder (different seller) isn't working at all. Lesson learned.

[96RAF]

How about this scenario. . .  Does that decoder operate the points on ports 2, 3 and 4 OK after a reset and readdress, but once you try port 1 it goes belly up.

 

If so and given that it was ‘second hand’ it may have been used with a Select controller at v1.1 which had a bug that damaged R8247s - scroll down and read this page http://www.halton96th.org.uk/page16.html for more info.

[Dowse101]

Thanks for the reply RAF96. No, there seems to be two scenarios after getting power back:

 

1) Points on all other decoders work as normal, but all points on the dodgy decoder are unresponsive.

 

Or

 

2) activating any point from any decoder, including the dodgy decoder gives a short. Disconnecting the dodgy decoder from the bus allows everything to work as normal.

 

Seller has said that he purchased 4 new decoders about 18 months ago from his local model shop and they've never been used so I don't think this is the v1.1 select issue.

 

Given that the decoder only fails after a short then I think my solution is going to be, what I should have done in the first place, and take all my accessory decoders off the main DCC bus and give them their own accessory bus with power management circuit breakers to each circuit. Then chances of a short on that circuit will be extremely low.

 

I've actually had the decoder working today quite well. If a short occurs then I just disconnect it from the bus at the choc block, power up and then add it back in at the choc block. That seems to work, it's just a bit of a faff.

 

To be fair the seller has offered me a refund, but seeing as when it works it works and I should be able to stop the cause of the malfunction then I think I'll keep it for the price I paid.

 

I have another question about a potential fake R8247 that I've received, could just be old packaging, but will start another thread in a bit with photos.

[Fishmanoz]

You could just fit snubbers (lots of recent posts on them) which will stop the voltage spikes when you get a short and should stop the problem. 

[Dowse101]

I thought a snubber might help so I fitted a homemade one earlier today, but it didn't fix the problem.

 

It's occurred to me tonight that adding in an extra CDU might be creating an inrush current that is tripping the DCC++ protection. I assumed that the base station limit would be around 2amps (limit for motor shield), but apparently it's set at a conservative 800mA by default. If each CDU is pulling ~150mA when first charging then it's very possible that the 5th decoder plus any locos are momentarily pushing the current draw over the limit. I will be changing the limit first thing tomorrow and see if that makes a difference.

[96RAF]

Good thinking. We have seen that problem with other  acc decoder makes so it is a possibility.

[Dowse101]

Problem solved. 5 R8247s all charging simultaneously was too much inrush current for the DCC++ base station's default setting of 800mA cut off. I doubled the limit and no further problems. Proper snubbers now added too. Circuits to be separated in the new year when the EB-1 circuit breaker arrives.

 

Thanks for the responses guys.

[96RAF]

Thanks for the feedback. Good to hear it is sorted.