Random start ups of TTS fitted loco's

[mezzoman253]

Hi All,

 

I've got a strange problem with my DCC layout. The problem is random startups and movement of loco's, other than the one under control of the command station.

 

I've recently been fitting Hornby TTS chips to some of my fleet and they are great for my home layout. So far I've done a Bachmann class 37 & 47, and a Hornby class 60.

 

I'm using a Sig-naTrak 2A ACE2 controller with firmware V3.20, which is the latest on their web site. All the loco's are set to DCC only, and Railcomm is off in CV's. I also checked I'd not left any functions on when swapping to another loco'.

 

I thought it may be interference on the DCC bus and fitted snubbers to the ends of the 4 main DCC buses wiring. Prior to doing this, the random startup would happen roughly every 2minutes. After fitting the snubbers the first one happened 10 minutes after switching on. At that point, I switched off and gave up for the time being.  It may be that had I left the power on that it would have occurred again every 2 minutes or so. I can check that.

 

The layout is end to end, and only 12 foot long. The bus is in the form of a letter "K" on its side, the straight part being the longest run. See my topic for the layout plan http://www.rmweb.co.uk/community/index.php?/topic/53821-burnt-mills/.

 

All the sidings, 7 in all, have dual LED stop signals fitted. The station hasn't any fitted as yet. Might this be too much current draw?

 

So, any suggestions as to what may be happening here, please.

 

Thanks, Rob

 

[Dungrange]

I've no idea what may be causing your problem, but the Sig-na-Trak ACE2 displays the current draw to the top right of the main screen, so you should be able to ascertain if that may be your problem.  However, I thought that Sig-na-Trak only offered a 3.5 Amp and a 5.0 Amp power supply.  I have the latter, but I understand that the actual ACE2 is identical regardless of which power pack you went with.  I think that they are only selling the 5.0 Amp supply now.  Personally, I wouldn't have thought that would be your issue.

Edited January 18, 2019 by Dungrange

[34theletterbetweenB&D]

You will probably have to go at this 'bug' problem systematically in my opinion. The one short cut is if you can review back to the point where this problem was first observed, and then recall what change was introduced to the layout immediately before that incident. (Did this only start happening with the most recent TTS fitted loco going on the layout for example?)

 

My plan would start with taking the snubbers off (since they manifestly don't work) take all the other locos off, and see if the TTS locos alone exhibit the problem (I suggest an hour test to make sure you capture an incident if the problem is still present).

 

If no problems at that point, then reintroduce half the 'other locos' and test

 

If no problems at that point, then remove the 'other locos' and reintroduce the half group of 'other locos' and test.

 

That should home in on the interaction causing the problem.

 

Tests with further halvings of the 'other loco' group that created the problem will find the interaction.

[mezzoman253]

I've no idea what may be causing your problem, but the Sig-na-Trak ACE2 displays the current draw to the top right of the main screen, so you should be able to ascertain if that may be your problem.  However, I thought that Sig-na-Trak only offered a 3.5 Amp and a 5.0 Amp power supply.  I have the latter, but I understand that the actual ACE2 is identical regardless of which power pack you went with.  I think that they are only selling the 5.0 Amp supply now.  Personally, I wouldn't have thought that would be your issue.

Hi,

 

You are quite correct, my mistake, it's the 3.5A model. I knew it wasn't the 5A one and somehow got 2A in my head. So not likely to be a current problem.

 

I can see the current in the top RHS of the display. Thanks.

 

Rob

Edited January 18, 2019 by mezzoman253

[mezzoman253]

You will probably have to go at this 'bug' problem systematically in my opinion. The one short cut is if you can review back to the point where this problem was first observed, and then recall what change was introduced to the layout immediately before that incident. (Did this only start happening with the most recent TTS fitted loco going on the layout for example?)

 

My plan would start with taking the snubbers off (since they manifestly don't work) take all the other locos off, and see if the TTS locos alone exhibit the problem (I suggest an hour test to make sure you capture an incident if the problem is still present).

 

If no problems at that point, then reintroduce half the 'other locos' and test

 

If no problems at that point, then remove the 'other locos' and reintroduce the half group of 'other locos' and test.

 

That should home in on the interaction causing the problem.

 

Tests with further halvings of the 'other loco' group that created the problem will find the interaction.

Hi, 

 

Thanks for the suggestions. I'd planned something similar but hoped someone might have had the same and would point me straight to it.

 

The layout only had the 3 TTS chipped loco's on it at the time. I'll take off 2 and see what happens with just one on the track, and report back later.

 

Rob

[RAF96]

I experienced similar odd behaviour in that whilst programming a loco via a Hornby Elite with both the main and programming tracks connected to their respective terminals two TTS locos took off round the track, one fell into the turntable well, one other started its engine and yet another flashed its lights. My bus at the time was a star radial affair and no snubbers were fitted as I was awaiting the parts. Since then that layout has gone to the scrappers and the problem has not been seen since so it may have been that particular bus dynamics. Could you maybe set up a test section and place the suspect locos on it to gauge any reaction.

Rob

[mezzoman253]

Hi, 

 

Thanks for the suggestions. I'd planned something similar but hoped someone might have had the same and would point me straight to it.

 

The layout only had the 3 TTS chipped loco's on it at the time. I'll take off 2 and see what happens with just one on the track, and report back later.

 

Rob

Hi,

 

After my last post, I had put only the class 47 on the track. I switched on the controller and almost immediately it started up but didn't move. The controller had been left on 03 as there's no loco assigned to that address on my system. I changed to address 2(the 47)  and no function was lit, but touching the lights button made it shut down, and it stayed like that until I had to go out. I shut down the controller.

 

Today I've put the class 60 on the layout alone and about 5 minutes after powering up the controller the loco went into start but didn't move. I've again shut it down and it's been dormant for about 25 minutes.

 

Neither loco has taken off on its own, this only occurred when more than one TTS equipped loco was on the layout. 

 

So it's looking like the TTS decoders are doing something odd, especially as others have had this problem. 

 

I can see the current draw on the display and it's sitting at 0.02A(20mA) but does fluctuate on occasion to zero and 0.04A (40mA). I've got 7 twin LED buffer stop signals fed from the track, but they seem to be constant. Not sure if that's significant or not.

 

The snubbers are still in situ, as they need to be unsoldered.

 

I'll update again later if anything else happens to the class 60. Then I'll try the class 37 on its own.

 

Thanks for your comments.

 

Rob

Edited January 19, 2019 by mezzoman253

[mezzoman253]

The class 60 just went into startup after sitting dormant on the track for over 45 mins! I shut it down and the current is now reading 80-130mA fluctuating. Nothing else is on the layout, so is the chip doing this?

 

I changed the loco address on the controller to 003 again and put the class 37 (address 01) on the track .

 

 After 8 minutes the loco started up on its own. Again by touching the lights button on the controller the loco shut down.  I reset the address on the controller to 03 again and the loco has sat for 50 mins without change.

 

Rob

Edited January 19, 2019 by mezzoman253

[mezzoman253]

Been for dinner and then back to the layout and switched on the controller and loco set to 03 again.

 

After 19 minutes, the class 37 started up, no movement. I changed the address on the controller to 01 to shut it down, and back to 03 afterwards. After another 3 minutes the loco started up again. I shut it down and it's sitting on the track after a further 8 minutes good as gold

 

It seems the timing of these startups is random sequence, possibly due to some spike on the bus or other unknown factors that I'm trying to fathom out.

 

I'm loathe to remove all the stop signals as they are quite fiddly to get back once removed. But I may have to eliminate them.

 

I may drop Hornby an email as all the different TTS decoders, 37, 47 & 60 are doing it.

 

Rob

Edited January 19, 2019 by mezzoman253

[cravensdmufan]

Have been following your posts with interest, although I'm afraid I can't offer an explanation.

 

Just to say that when I first started out in DCC I converted my analogue layout by simply leaving all the isolating switches in the "on" position. The lazy option!

 

I used an NCE Powercab starter kit with the basic 3.5 amp output. I usually had 10 or 12 locos stationary on the layout with 2 moving on the main lines plus an 08 shunting around the yard.

 

Getting on to the point of your topic - a good few of my locos randomly behaved strangely. Unxpected start up sounds as you mention, locos "running away" that wouldn't stop, and some suddenly taking off at top speed! Not just TTS ones but Loksounds as well.

 

So I explained my problems to a specialist DCC supplier (Coastal) who suggested I installed proper bus wires with frequent droppers and to break the bus i.e not have it all the way round my roundy roundy layout. Which I did.

 

But the best bit of advice was to upgrade the system to 5amp with a Tam Valley booster. Well I reckon it was the best £40 I ever spent on the layout. Cured all the problems, and every single loco has behaved perfectly ever since. The trains also keep better constant speed, wherS before sometimes they would randomly go slower then speed up again.

 

I'm not saying it would cure your particular situation but I thought may be useful to describe my own experiences.

Edited January 19, 2019 by cravensdmufan

[GoingUnderground]

I'm just wondering if it might be mains voltage spikes or sudden drops caused by appliances turming on or off in your house. These might be affecting your PSU's output voltsge which in turn is affecting your Signatrak controller. Or it mighy be spurious RF interference being induced in the cable connecting your PSU to your Signatrak.

 

I'm not an electronics or radio expert so this is pure speculation on my part. But when you've ruled out the obvious causes it's worth thinking about less plausible reasons. If this was the cause, would a ferrite bead or core fitted at the ckntrollervend of the lead from the PSU fix it? If it did it eould be a cheap fix. Of course, I may well be totally wrong, and I won't take offence if the ecperts say that I'm talking rubbish. But I have read of a case of fitting a ferrite bead or core to the low voltage output lead of a PSU curing odd effects with DCC system.

Edited January 20, 2019 by GoingUnderground

[RAF96]

Another thing to consider (if your controller supports it) is the possibility of loco zero being inadvertantly selected and set to move, as the modified dcc waveform could affect sound decoder much the same as DC runaway.

Rob

[mezzoman253]

Have been following your posts with interest, although I'm afraid I can't offer an explanation.

 

Just to say that when I first started out in DCC I converted my analogue layout by simply leaving all the isolating switches in the "on" position. The lazy option!

 

I used an NCE Powercab starter kit with the basic 3.5 amp output. I usually had 10 or 12 locos stationary on the layout with 2 moving on the main lines plus an 08 shunting around the yard.

 

Getting on to the point of your topic - a good few of my locos randomly behaved strangely. Unxpected start up sounds as you mention, locos "running away" that wouldn't stop, and some suddenly taking off at top speed! Not just TTS ones but Loksounds as well.

 

So I explained my problems to a specialist DCC supplier (Coastal) who suggested I installed proper bus wires with frequent droppers and to break the bus i.e not have it all the way round my roundy roundy layout. Which I did.

 

But the best bit of advice was to upgrade the system to 5amp with a Tam Valley booster. Well I reckon it was the best £40 I ever spent on the layout. Cured all the problems, and every single loco has behaved perfectly ever since. The trains also keep better constant speed, wherS before sometimes they would randomly go slower then speed up again.

 

I'm not saying it would cure your particular situation but I thought may be useful to describe my own experiences.

Hi, Thanks for you input.

 

The bus wiring is not circular it's in the form of a "K" on its side, the layout is end to end, with the feed in the middle. Each leg is approximately 6 foot in length and heavy duty cable, It isn't twisted though.

 

Each piece of rail has a dropper at each end and the points are fed at the toe, as you should. The layout was DCC from day one.

 

The coin test works at the extremities of all lines, As a last resort I fitted snubbers to all 4 bus ends.

 

The controller is 3.5A and should be fine for 3 locos and a few lights.

 

Rob

[mezzoman253]

I'm just wondering if it might be mains voltage spikes or sudden drops caused by appliances turming on or off in your house. These might be affecting your PSU's output voltsge which in turn is affecting your Signatrak controller. Or it mighy be spurious RF interference being induced in the cable connecting your PSU to your Signatrak.

 

I'm not an electronics or radio expert so this is pure speculation on my part. But when you've ruled out the obvious causes it's worth thinking about less plausible reasons. If this was the cause, would a ferrite bead or core fitted at the ckntrollervend of the lead from the PSU fix it? If it did it eould be a cheap fix. Of course, I may well be totally wrong, and I won't take offence if the ecperts say that I'm talking rubbish. But I have read of a case of fitting a ferrite bead or core to the low voltage output lead of a PSU curing odd effects with DCC system.

Hi, Thanks for the suggestions, any help is welcome.

 

I'd considered both your ideas already, but not as yet tried it. The controllers DC lead does have a built in ferrite, as required by EC EMC rules. 

 

I think it's more likely to be mains transients, as there's a long run to the railway room. The room is also my radio room, ham radio, but I'm not transmitting and playing trains at the same time!

 

Rob

[mezzoman253]

I

Another thing to consider (if your controller supports it) is the possibility of loco zero being inadvertantly selected and set to move, as the modified dcc waveform could affect sound decoder much the same as DC runaway.
Rob

Hi,

 

That's something I haven't checked, many thanks. I'll have a look later and report. i'd checked all from 1 to 10 didn't have any functions set, but not 0.

 

Rob 

 

Edit: Just checked address 0 and nothing set.

Edited January 20, 2019 by mezzoman253

[JJGraphics]

I've had the same issue with TTS and my Lenz controller. I had a pair of 20's and every now and then one would randomly start up. Then I got a TTS vent van and that does the same. Glad its not just me.

 

 

I have seen two instances of strange things happen on my layout (ECoS 50200) with two different visiting locos fitted with TTS decoders. One randomly started-up (Class 20) and another (Class 40) made a sudden movement after sitting on the layout ticking over for some time. In neither case do the locos behave abnormally on my Grandson's layout which has a Hornby Select with V1.5 firmware.

 

I have not seen any similar problems on my layout with my own ESU decoder-fitted locos or with visiting locos which have Zimo sound decoders fitted or with non-sound fitted visiting locos.

 

John

[GoingUnderground]

I have seen two instances of strange things happen on my layout (ECoS 50200) with two different visiting locos fitted with TTS decoders. One randomly started-up (Class 20) and another (Class 40) made a sudden movement after sitting on the layout ticking over for some time. In neither case do the locos behave abnormally on my Grandson's layout which has a Hornby Select with V1.5 firmware.

 

I have not seen any similar problems on my layout with my own ESU decoder-fitted locos or with visiting locos which have Zimo sound decoders fitted or with non-sound fitted visiting locos.

 

John

Sounds like another instance of the TTS bug.

[mezzoman253]

I've sent an email to Hornby, copied below, asking for guidance on this. Explaining the effects and that more than a few people are experiencing odd behaviour with TTS chips.

 

"Hi, This is regarding TTS decoders and random startups of locos not selected by the controller.

I recently started to fit some TTS decoders I'd bought online from Hattons, & Rails of Sheffield. They were R8102(class 37) R8103(class47) and R8104(class 60). All were purchased in January 2018, just after the class 37 ones were shipped to the UK. I'd had them on pre-order.

The problem I'm having is random startups of the chipped locos. The loco will be sitting on the layout and will suddenly go into startup. It doesn't move, just starts up a idles. This is with no other loco's on the layout. However if 2 or more are on the layout, one of those not being controlled will startup and move, again the loco "chosen" is random.

I've checked all loco addresses have not got any functions operated. Railcom is off as is any DC control in CV's. Thinking it may be interference I fitted snubbers to the DCC bus  wiring, but that hasn't cured it.

Being on the RMweb forum I've asked about this and several others had had the same or similar using TTS chips, but haven't found a cure.

Any assistance would gratefully be received.

Thanks, Rob Harrison"

 

I'll let you know what they come up with.

 

Rob

Edited January 20, 2019 by mezzoman253

[RAF96]

As there has been a TTS bug around (and fixed) for some time it could be that there is old stock un-reprogrammed for the fix lying around in stock at various retailers.

 

The only way to determine the firmware state of the decoder is to read CV7 and note the number read back. This is not definitive as other CVs can affect the final definition like CVs 156,157, 159.

 

It would probalby help Hornby if those folk with problems could return the values seen in these CVs, then it can better be determined if the problem is an old stock decoder hangover or if it is new build and do they have a new problem.

 

Vague complaints like ‘its another TTS bug’ in post #18 are of no help in the real workd when trying to fault find a real or imagined problem.

 

Rob

Edited January 21, 2019 by RAFHAAA96

[mezzoman253]

Hi Rob,

 

Thanks for the info'. I ordered my TTS decoders from Hatton's in January 2018, but hadn't got the new controller then so couldn't program  CV's. I got the class 47 & 60 chips immediately, but the class 37's didn't arrive until March, being a new release I think. I got another class 37 chip in June from Rails of Sheffield. Looking through the packaging and paperwork I can't see any marking suggesting when they were made. I'll read the CV7 's later.

 

I agree this needs a bit of methodical work to find the real cause(s) rather than putting everything down to "the TTS bug". 

 

 I think I may have solved this, for my system, although it's a bit embarrassing as it looks like pilot error.   Sorry for leading people up the garden path.

 

 

Yesterday, after sending the email to Hornby, I revisited the addresses 0 to 10 to double check I'd not missed one.

 

The Signatrak ACE2 controller has a touch screen for input and the Functions are mainly in dark blue, except the high-low & low-high horn, which are green.

 

When touched they go a slightly lighter shade of dark blue. Not easy to see side by side, well for my eyes. I'd missed one on address 4 function 4, that's brake squeal (momentary action) on TTS. It was set on and I assume permanently?

 

I turned it off and did the one loco on the track tests again. I put the class 37 on and left it for 2.5 hours and nothing. Then the class 47, again 2.5 hours and nothing. So I had a little running session and closed down for the night.

 

This morning I put the class 60 on the track and after 15 minutes in went into startup!! Oh no, my thoughts of a result evaporated. But wait, check the address again you may have left something left on. I had address 4 I'd left the lights on. Was this triggering the class 60?

 

So I turned of lights and tried again. Three hours have gone by and no sign of life in the class 60. Not definitive but encouraging.

 

I'll have to    again to see if I missed anything. It seems the controller is trying to send the set function signal out repeatedly as it's getting no response and that may be triggering the other loco's? Not a DCC expert in that regard. Any thoughts on this?

 

If this is the case, it may explain others similar experiences. Having said all that, does it explain why it was occurring with 3 loco's on track, one of which has address 4.

 

ATVM I'm seeing what 2 loco's, class 37 & 60, do on the track. So not completely out of the woods.

 

My head hurts!

 

Rob

Edited January 22, 2019 by mezzoman253

[GoingUnderground]

Well as it's reported as happening by at least 3 different people with 3 totally different makes of command station, Signatrak ACE2, Lenz, and ESU ECoS, the problem.is more likely, IMHO, to be the decoder and not the command station or user error.

 

The symptoms seem to be different from the previous problem which was the loco had to be pointing on a particular direction vis-a-vis the track polarity for it to work. Turn it round and it didn't work. That problem seemed to affect a particular batch of TTS decoders as it only started cropping up some time after TTS decoders became available.

 

This latest issue seems to be of much more recent origin and unrelated to the previous bug. So old unreprogrammed decoders wouldn't necessarily account for it. In addition it seems to affect both factory fitted TTS decoders and those sold for aftermarket fitment, according to the reports that I've seen. The folks experiencing it all say that the problem is only with their locos fitted with TTS decoders, reinforcing my view that it's the decoders and not the command stations or user error.

 

I agree that Hormby need to know about it and given a chance to fix it. And I sympathise with those that have the problem as it doesn't seem that they can do much about fixing it by themselves.

[RAF96]

The only other thing I can add is that I have had certain sounds play at random intervals when I thought they were off.

Turns out I was using a Select controller which has no indication of functions in use. The offending sounds were always of the play once type, which having played through sat there in the background still switched on. Whether a spike resent the play once command again I don’t know and without some form of data logger monitoring the command stream I don’t suppose I ever will.

 

If folk having problems could read back CV 7 and CVs 156/157/159 if possible then let Hornby know along with the decoder R number if it was a retro fit unit and the make/model of the loco fitted as well as the controller model and firmware state, then at least Hornby can attempt to replicate the problem.

 

If folk would rather post that info on here I am happy to collate the results and send it to Hornby in one lump.

 

Thanks Rob

[mezzoman253]

The only other thing I can add is that I have had certain sounds play at random intervals when I thought they were off.

Turns out I was using a Select controller which has no indication of functions in use. The offending sounds were always of the play once type, which having played through sat there in the background still switched on. Whether a spike resent the play once command again I don’t know and without some form of data logger monitoring the command stream I don’t suppose I ever will.

 

If folk having problems could read back CV 7 and CVs 156/157/159 if possible then let Hornby know along with the decoder R number if it was a retro fit unit and the make/model of the loco fitted as well as the controller model and firmware state, then at least Hornby can attempt to replicate the problem.

 

If folk would rather post that info on here I am happy to collate the results and send it to Hornby in one lump.

 

Thanks Rob

Hi Rob,

 

Here's the CV readings for my TTS loco's:-  ??-??? Bachmann  Transrail  Class 37 406 ( I think this is a respray/renumber)  CV7=132, CV156=147, CV157=6, CV159=147   Hornby code R8102

 

                                                                      32-817 Bachmann   RES Class 47 745                                                                 CV7=132, CV156=149, CV157=6, CV159=149     Hornby code R8103

 

                                                                      R2488 Hornby  EW & S Class 60 026                                                                   CV7=132, CV156=163, CV157=5, CV159=163      Hornby code R8104

 

I'd be happy for you to collate the info' and send to Hornby.

 

Thank you.

 

Rob H. 

 

Edit for loco types and codes.

Edited January 21, 2019 by mezzoman253

[RAF96]

Thanks Rob H.

Could Richard Croft, JJGraphics and Cravensdmufan please add as much detail requested as possible, then I can send a summary off to Hornby for imvestigation.

I appreciate visiting stock may not be available to interogate.

Rob

[mikeg]

Try using a voltage regulating extension lead to supply the command station, this worked for me with my Z21 which would change points at random, also set of locos with 20+ year old chips (lenz 1000's) our mains voltage has spikes which affect the TV and our central heating programmer, so I had an idea what was cousing the problems!

 

regards mikeg