Heljan 128 fries three decoders

[Nigelcliffe]

12 hours ago, Izzy said:

If you will forgive me I am bemused that a decoder can be fried on a programme track. I have always understood that the only current that flows is of low voltage for the brief duration that a signal is sent, and that the whole idea is that damage can’t occur to a decoder. 
 

Am I wrong?

 

If not then it would suggest that perhaps there is something wrong with the program track configuration. 
 

Izzy
 

 

Yes that's correct *if* the programming track only ever submits a loco to the short programming pulses.   Some makers programming tracks do this.   

And some don't, because the programming track can also be used for running and "automatically switches over when programming".    So, when not "automatically switched over" its no longer the safe extremely low current and total power environment, but something else which can submit loco and decoder to more power.  

 

Its a mess, and I don't think maker's of systems are helping users with their powered programming tracks.   

 

In this case, I think the owner should send the loco back to the supplier for replacement.   It shouldn't be blowing up decoders.    And, there is, to my mind, a reasonable claim for consequential losses - ie. replacing those blown decoders.  

 

   

 

 

[Izzy]

Thanks for that info Nigel. What a dogs dinner DCC systems and NMRA standards are, the different combinations possible. Never ceases to amaze me. No wonder it can be so difficult for some to cope with it if luck is not on their side. I am just truly grateful I got a Sprog 2 early on into my DCC experience and only use that for setting up and testing all decoder installs, the main Gaugemaster Prodigy (aka MRC) only being used to run layouts and for POM. 
 

Izzy

[spamcan61]

3 hours ago, Deborail said:

I have contacted Heljan and they said that their motors need robust decoders such as ESU. I was

told to use ESU. They said cheaper decoders cannot handle the load of the motor on this locomotive. So far the decoders used were digitrax and a couple of Dapol Imperiums 

I hope they include that info. (which doesn't sound correct anyway, ESU decoders aren't magic) in the user documentation then.

 

Edit: Yes Heljan locos have a reputation for higher current draw than most other makes, but 300mA doesn't sound at all high anyway and it's not like ESU are the only manufacturers of high(er) current handling chips.

Edited October 21, 2020 by spamcan61
clarification

[dhjgreen]

10 minutes ago, spamcan61 said:

I hope they include that info. (which doesn't sound correct anyway, ESU decoders aren't magic) in the user documentation then.

 

Edit: Yes Heljan locos have a reputation for higher current draw than most other makes, but 300mA doesn't sound at all high anyway and it's not like ESU are the only manufacturers of high(er) current handling chips.

I had a problem with a decoder on my 128, I now use Lenz.

[Deborail]

4 hours ago, dhjgreen said:

I had a problem with a decoder on my 128, I now use Lenz.

I will try Lenz. If it blows this too I will let it be a DC loco. 

[letterspider]

When you contacted Heljan, did they advise any way in which to trouble shoot the problem?

If I was in your position I would immediately suspect the PCB and replace it.  See if Gaugemaster have any in stock - surely cheaper than a decoder

Coincidentally I am having problems with a Heljan 47 which runs on DC but refuses to run on DCC, so you are not alone...

[letterspider]

On Gaugemaster website

HN-OO-CL128-105(S) Part 105 PCB 21 Pin w/Connectors5056397828340 £1.00 Currently Unavailable

Well - at £1 it would be worth buying as a spare.

Perhaps they can get one for you on back order?

[John ks]

It would be interesting to see a close up of Digitrax decoder with the blue heat shrink removed

It might be possible to determine which part of the decoder failed & narrow down why it failed

John

[Deborail]

I have decided to return the locomotive to the seller for a refund. I think I lost three decoders, but I have a feeling that I would lose more going forward using this locomotive. I will source out another 128 later on...hopefully that would work well with DCC. 

[WIMorrison]

Why not get the seller to fit a decoder and sell it to you as DCC Fitted?  that way you will either get a working loco or the seller will accept it is goosed.

[Deborail]

18 hours ago, WIMorrison said:

Why not get the seller to fit a decoder and sell it to you as DCC Fitted?  that way you will either get a working loco or the seller will accept it is goosed.

He said that he prefers to refund. 

[WIMorrison]

Hmm, suggests that he perhaps knows that there is an issue?

[Deborail]

On 23/10/2020 at 15:24, WIMorrison said:

Hmm, suggests that he perhaps knows that there is an issue?

I am not sure mate, but anyway, I am posting it back to the Seller today, it is the Scooby Doo Weathered version. 

I have found a Royal Mail version on ebay and purchased it. I also ordered a few ESU decoders.

 

Hope for the best going forward...

[DY444]

"Once is happenstance. Twice is coincidence. The third time it's enemy action."

 

I was a relatively early UK adopter of DCC starting with it in 1999.  In all that time I've only had one decoder destroy itself in a fashion similar to that you have described.  That was caused by a solder whisker causing a short on the circuit board of a Bachmann peak. 

 

My advice for future reference is that you should always assume the locomotive is the cause until you can definitively prove otherwise, and, until you do definitively prove otherwise the only thing achieved by trying more decoders is more blown decoders.

[letterspider]

We've all made this mistake at least once, I am sure. 

[Deborail]

I have I stalled a new ESU decoder in my new Royal Mail class 128 DPU and it works fine-confirming that the earlier locomotive had a fault. Installed it with the CV29 value as ‘2’ address ‘28’...for some reason it wouldn’t move if I set the address to ‘128’. I think that is because in AD2 that would be considered as ‘0’

[Oldddudders]

Address 128 is a 4-digit address in every system I have heard of. Oddly enough, 127 is a 2-digit address in some. 

[WIMorrison]

That would be because 127 is a 7 bit number 0b1111111 whereas 128 is an 8bit number 0b1000000.

 

2 digit addresses are stored as 7 bits hence you can have up to 127 addresses.

 

:)

 

[ColinK]

I had the same problem with calling my 128 address 128, until I switched it to a long address.

[letterspider]

I just received a certain type of class 66 from a certain box shifter.

The chip was new and so I was happy to operate on address 03.

As soon as it touched the track it ran full power and fried the chip.

The shop says they have never heard of this problem before but offered a refund. 

I am sending the item back and asking them to test and chip it...

[Crosland]

8 hours ago, Oldddudders said:

Address 128 is a 4-digit address in every system I have heard of. Oddly enough, 127 is a 2-digit address in some. 

 

6 hours ago, WIMorrison said:

2 digit addresses are stored as 7 bits hence you can have up to 127 addresses.

 

Pedantically, that's 3 digits so some systems limit 2-digit short addresses to 1-99, hence the confusion in the grey area between 100 and 127. If you want your locos to be portable between layouts it's best not to use addresses 100-127.

[AndrewC]

Late to the party and I know you've got a satisfactory outcome but....

I've seen that burn mark in a Digitrax decoder before. Same location as well. I have an Athearn SW1000 that ate decoders. After a lot, and I mean a lot of profanity, 2 more decoders, and finally ripping out the circuit board and hardwiring, I came across the problem when running a meter over the damn thing. Turns out there was a bad component that meant the full track voltage was connected directly to the fwd lighting output. (in 8 pin plug land that meant a straight connection from black to white) Boom. I now triple check or rip out every circuit board before installing a decoder. Likely the Heljan board had a short between one of the track feeds and an output.