DCC Short Circuit using Multimaus

[murray1]

I have been running my layout for several months using a Roco Multimaus. Currently have 5 Hornby TTS fitted locos all performed as expected. Diesels improved by upgrading speakers. It is essentially two circuits with a couple of sidings. It is wired up with bus wire and dropper feeds as standard practice. 

I recently purchased a Hornby R8249 decoder and fitted to a Hornby DCC ready class 101 and when placed on track moved off with no control from the controller.  Thereafter it created a short circuit when placed on track and power applied. Tried same decoder in another DCC ready model - same response.

Sent decoder back and was replaced, which was exactly the same.  Last week I bought a Howes class 108 sound chip to fit in same DCC ready class 101. Straight away it performed the same as the Hornby decoder. I have replaced the original 2 pin power supply which came with controller with a 18v - 3.5a variable laptop power supply thinking this may help - still same performance.

The whole time, the Hornby TTS models all run perfectly so completely confused where the fault lies and how to fix. Any advice or help will as always be much appreciated.    

 

[Nigelcliffe]

When something goes wrong, look for the common item.

 

You say there is one faulty loco (a Class 101), which doesn't work with several different decoder fitted.   So, the reported evidence points to that loco having a fault.  

 

- Nigel

[murray1]

Wish it as that straight forward. Hornby decoder tried in 3 DCC ready chassis locos - all same. Howes sound decoder tried in class 101 only - same issues / fault as basic decoder. 

[Ray H]

Could it be that the Class 101 has a problem and might just be damaging any decoder fitted to it?

 

Can you look at the under side of the socket on the model and see if there are any whiskers of wire incorrectly linking pairs of pins?

 

Do you still possess the ability to run the Class 101 on a straight dc facility - i.e. with the usual (dc only) plug fitted? Can you check it out on that?

 

[murray1]

I am going try the sound chip in another engine. Will try the DCC ready locos with their  blanking plugs on a  DC stretch of track to check locos are still running as expected.  Will also check the pins are not touching chassis and all good connection wise.  

 

Thinking of powering up a separate straight track to check on DCC to eliminate track issues  - though the 5 TTS chipped locos are all fine, which makes me think the track is not the problem.

 

Beyond that I am starting to run out of tests I can carry out. 

[RFS]

Your problem could be a wiring issue with the 8-pin socket whereby the two terminals at one end are electrically joined by a blob of solder underneath. This will short out the decoder and could damage it. However this problem will never show on DC because all the blanking plug does is join these two pins together anyway.

 

If you have a multimeter you can test whether there is continuity between the two positions at each end of the socket. If there is, that's your problem. If you don't have a multimeter you can try inserting an 8-pin plug the wrong way round, ie so as only two pins at one end are connected by it. If the loco then runs on DC, the problem is at the other end.

 

I would also strongly recommend that, having fitted a decoder, you first check it on the programming track and do not place it directly on to fully-powered track. If there is an issue, this will show up without risking damage to the decoder.

 

Edited October 30, 2019 by RFS

[murray1]

tested both two DCC ready locos with blanking plug and both run well. tried the sound chip in the DMU on a short piece of track and it run and it ran also - but with no sound. So know both locos motors are good and the sound chip does power the loco.  

[Phil S]

Firstly - as Ray H and RFS have written - a problem with the socket in the Hornby 101 COULD have destroyed every dcc decoder inserted into it!

It has been known for the wires from different wheels/bogies on Hornby locos to be wired to the adjacent holes  ( track AND motor) which does not show up on aanalogue testing with the blanking plate (ANY blanking plate) inserted.

ONLY checking it with NOTHING plugged in will reveal the problem  because a blanking plate links those pins.

Decoders are not usually protected against Track-to-motor direct shorts as this bypasses much of the decoder's cicuitry and therefore any protection.

If you do not have a multimeter ( and they can be bought for less than a silent decoder) then use a 9V battery and a bulb or LEDandResistor (of 1k or more)  to test for ANY connection from each hole to track and motor...

Engine Sound will start with either F1 or F8 but spot fx may occur anyway with f2, 3 etc...

(Do also do the visual check of the underside of the socket as also suggested)

 

As the default M\ultimaus/Amplifier does not have a separate programming track, which would limit the  current, [ although the multimaus does monitor the current during prgramming and reports an error if out of range)  then a simple test is the 9v battery applied to either the wheels ( if all wheel pickup) or the separate length of track you mentioned ..  but unconnected  to anything else.   By default, dc running is enabled on all decoders.

Edited October 30, 2019 by Phil S
adding alternative 9v 'pre test' track

[Butler Henderson]

Would well recommend using something else for programming;  a SPROG and Decoder Pro for example, that way any issues with the model and/or decoder will show up before full dcc is applied.

[murray1]

Thanks for the information. I removed the Howes sound chip and fitted into a Hornby DCC class 55 chassis. Pleased to say the decoder performed perfectly sounds all good, so no damage to that one. 

The problem as suggested (many thanks) appears to be the socket on the Class 101. Can I ask if these can be replaced / repaired if faulty. I would prefer to have this done professionally as the risk of blowing full sound chips is too high.

The basic Hornby decoder that I believe is blown totally as didn't work in the Class 55 chassis I will just bin as has been suggested wasn't probably the best buy in the first place.  

[Butler Henderson]

First thing is to work out where the fault is. What seems to the socket can sometimes be somewhere completely different. What you need is multimeter with a built in alarm when it finds a short circuit and then test each adjacent set of terminals on the socket. If a short circuit is found unfasten the socket and look at the underside of shorting terminals - it may be a solder blob that needs removing using a soldering iron or a stray wire. If there is no evident fault trace the wiring through the model until you find it. For example on a Bachmann 4 2-6-0 the fault turned out to be touch too much solder on a wire to contact strip at the baseplate that was just shorting on to the adjacent contact strip.

[murray1]

All good advice... thanks 

[murray1]

Below are pictures of the underside of the PCB socket. Is there anything obviously wrong with it. I have not altered in any way since purchase

[Phil S]

Assuning that the pairs of black wires are coming from the bogies/track, and the singles are going to the motor ..... the wiring looks as if it correct  (and has not been wired in mirror image)   Because they use 'blind' holes - there os no danger of a pin touching othe rmetal below. [But I would still put a sliver of insulation beneath it]

 

Pin 1 is a motor connection [ Orange and normally from the Right hand Rail facing forward) , and the diagonally opposite is the other motor connection.

You can verify this - as I have just done - as an aide memoire - by comparing it to a pre-wired 8-pin plug, and orientating it so that orange=Pin1 goes to the marked triangle oe '1'  hole.

 

The other diagonal corners then have (2 connections) to the track   .... in some Hornby models, one of these adjacent wires has been wired to the adjacent pin -  which might be the common +ve from the decoder or a lighting output .... neither of which show up as an error when testing on analogue with the blanking plug .... because these link the obvious motor and track, but also any lighting connection.

 

Testing this would be soo mmucchh eassieerr with a multimeter  - just set to resistance/contiinuity check range, and touch the probes ....

 

I have just fitted a Class 37 TTS deocder to a V-Trains chassis .... needless to say it worked when stationary, but demonstrated its safety cutout actions immediately I tried to move it .... the motor resistance was only 20-22ohms.   As I couldn't locate any of my 10ohm 5Watt resistors, I opted for the simple solution of  replacing the direct-decoder previously fitted, and siimply wired the TTS decoder in parallel... and programmed both the the same address.   Note that this 2-decoder method does not work  with a 'Steam' TTS decoder (because that needs to see the back emf of the motor), but did work because of the compact size of the 'direct' plugin decoder (an old ZTC that is still working) leaving space for the TTS decoder over the loco PCB.   The ViTrains Speaker position is designed for a 20mm round speaker  -and I wanted a quick conversion, so i mounted the speaker below the fan grille.

[Ray H]

I'm not saying that it is a problem but the overly bared wire at bottom right of the back of the decoder doesn't look as though it was last soldered when the other terminals were. Is there any possibility that that excess of uncovered wire is making contact with the chassis or some other part that it shouldn't.

 

I'm a bit loathe to suggest this but if you have a cheap but known to be working decoder, can you install it in the socket with the socket vertically - i.e. where there should be no chance of the back of the socket touching anything else?

 

That way it would suspiciously look like the back of the socket is touching somewhere erroneously if the decoder behaves normally. I'd probably take the added precaution of making a paper "washer" - a piece of paper with a single cut in it to allow the wires to pass through (the cut) - and used to separate the decoder socket electrically from the rest of the loco.

 

Have you been able to take a (very) close look at the back of the decoder and made sure that there is nothing metal - especially a fine whisk of wire that is providing a metalic link between any two of the soldered to pads?

[murray1]

I will look to isolate the bare wire end better. I had with the socket moved away from the chassis, and to the best of knowledge vertical to avoid cross contact, tried a decoder  but short as usual.

I have Contaced Hornby, but with age of model not expecting any help.

Is there anywhere you can take dcc models for this type of repair. Would happily pay for fix or PCB replacement. Even tempted to buy another model to fit sound chip and use this as a non  powered set. Would be a shame as motor runs sweetly on DC.

[Butler Henderson]

Simplest issue with a loco/unit that has an obvious wiring fault is strip out all the wiring bar the wire or wires from the pick-ups and hard wire a decoder in place, checking first of all their is no issue with the pick up wires. Its also a useful approach on many Hornby models that have a unnecessary spiders web of wiring to make for a far neater job while getting rid of the socket can in some models (the M7 being an obvious one if you want to keep all the weights) open up the range of decoders that can be fitted.

[murray1]

Well what can I say - I contacted Hornby by their website, and after a couple of emails the model was sent to them. My boxed and modified / DMU power car was posted, and 1 week later came back brilliantly packed with a note inside. They replaced the decoder socket, replaced the motor and fitted a new Hornby decoder all with no charge. This model is easily 11 or 12 years old, but invaluable to me due to the detailing carried out, and I cannot praise Hornby enough for their excellent service.

 

I wil never know what caused the shorts or fault, but thank you to Hornby and everyone who commented above.