DCC converted train has sound but moves only briefly before needing to be power cycled

[Paul-365510]

Hi. First posting, be gentle, I did search for answers in the forum and online and found nothing relevant.

I returned to railway modelling a few months after 50+ years.

Built a DC layout, then started conversion to DCC with eLink and Railmaster.

Got it working (had to dig out old desktop PC to get railmaster working, just wouldn't install properly on new laptop even following all the guidance) and acquired a second-hand DCC enabled Hornby Caledonian 0-4-0 to master the basics. That all worked fine so I thought I knew what I was doing :(

Next tried to convert a couple of old DC Hornby engines - both around 20 years old: J83 0-6-0 tank engine and Flying Scotsman (following guidance here: Tender Drive Flying Scotsman – Hornby Support.) Used HM7000 with sound. On the J83 I have working sound (it's brilliant!) but no sound on the Scotsman (yet).

Both exhibit the same problem when I try to get them moving - either using Bluetooth and the HMDCC app, or using railmaster. In either case the engine moves briefly and then stops, including sound stopping in the case of the J83.

If I'm using BT I can see the engine disconnect from the app as soon as I throttle up, so the decoder is dying.

To get things going again I have to remove the engine from track power briefly.

No short circuit detected by the eLink.

It appears as if powering the motor causes the decoder to restart, perhaps it can't deliver enough power and keep itself alive at the same time but why would that be? Other people seem to have done these conversions.

The FS is tender driven and on DC the tender runs without the engine, so I haven't even got around to connecting the two together again.

A problem which might, or might not be related. At one point the speaker on the FS started emitting a loud shriek and after a few seconds, smell of burning and the speaker was toast... but the problem with immobility was the same before and after this depressing event.

Am I doing something stupid, or have I just picked the wrong engines to convert?

thanks

Paul

[SteveM6]

In addition to referring you back to the troubleshooting guides at the top of this page, I would also suggest that you check and clean the track, wheels and pickups I have a similar issue with a small 0-4-2 - it’s only a year old but very fussy about cleanliness until the powerbank has charged.

It is also possible that your older locos draw a higher current and cause the decoder to trip a restart its startup procedure.

[ColinB]

It may be you are using the wrong decoders from what you are describing it seems like the motors are taking too much current. Old locos can take up to 0.8 amps which is well above the maximum output of Hornby normal decoders or TTS decoders. The HM7000 according to the specification is supposed to source up to an amp but I have not been brave enough to test it. You don't mention which make of decoders you are using which might be a help. I did a load of tests when I converted all my own old locos from the 1980s and found many of the decoders failed even though the current was within the decoder specifications (I suspect that they were being overly optimistic). In the end I found that the Zimo decoders worked perfectly and I don't have any interest in the company. The ones Jenny Kirk advertise also seem to do exactly what they say on the packet. So in short you can easily convert all locos to DCC it is just getting the right decoder.

[96RAF]

OP says he is using the app so it is presumed it’s a HM7K decoder.

Does it have a power bank fitted or just running without.

[ColinB]

I know 96RAF I wondered that after I had written the reply.

[Fishmanoz]

Rob and Colin, Paul is specific - HM7000, BT and the app. so I don’t see any ambiguity. He also says eLink/RM for DCC. So all that together with his loco info gives us a pretty comprehensive picture of his setup.

Paul, you’ve been told the possibility of the tender drive in particular overloading the decoder and causing it to shut down but there is another overload possibility here. That is overloading the P9100 1 Amp PSU that comes with the eLink.

The 1 Amp PSU is known to have caused a range of issues, all of which were solved by upgrading to the 4 Amp P9300. I suggest it is worth your while upgrading as it will also allow you to run more than 2 locos at once.

But I would also be referring to the Troubleshooting guides.

PS. I’m assuming you are continuing to power the track via the eLink when operating BT, not going to some other PSU?

[Paul-365510]

thanks for all the responses, particularly Fishmanoz. It's good to know so many people out there are willing to help.

I am indeed continuing to power the track via the eLink when using BT control.

I'll upgrade the PSU and report back. If that doesn't fix it then I'll consider a different decoder, although I sort of hoped by sticking to a Hornby DCC in a Hornby train (albeit older ones) I would avoid compatibility issues, and the HM7000 is pretty affordable for DCC with sound.

[Brew Man]

Re-iterating SteveM6's advice, do make sure your track, wheels and pickups are squeaky clean. DCC can be particularly susceptible to dirty track/connections. I find IPA plus track rubbers are the best combination. Wheels and pickups can be be more awkward to get at.

[Paul-365510]

thanks Brew Man.

track is clean, as are wheels. I've tried both of the misbehaving engines on several different sections of the layout as well as single pieces of new track connected to the live, or programming outputs of the eLink, with the same results. And as I mentioned the little 0-4-0 which has an older decoder in it is working fine.

I just upgraded the PSU to a 4A version (it's convenient that some laptop PSUs are also 15V, 4A) and it's not made a noticeable difference. Engines can move 1mm or maybe up to a 1cm before they reboot.

I have a couple of HM7000 powerbanks ordered. I doubt if they can fix the problem since the issue is not an interruption to the supply but maybe they can smooth things a little which might help.

It sounds as if I should give a Zimo decoder a shot at the problem - any recommendations of exactly what I should buy, assuming I want sound?.

The only other thing I can think of swapping out is the eLink itself in case it's not capable of delivering the necessary juice even with a 4A supply plugged in. I don't actually see the eLink in the list of compatible controllers on the trouble shooting page, but maybe I'm reading it wrong.

[Brew Man]

It might be worth your using just the power supply to power the track and leave the eLink out of the equation. That will tell you if the problem lies with the eLink or not.

[Paul-365510]

I don't think that would work: the PSU is 15V DC, and as I understand it, the output from the eLink (or any other command unit) is 30V AC (square, not sign) so at best nothing will happen, and at worst I might do some harm.

The DC trains would certainly enjoy 15V for a while before they burned out!

[Rallymatt]

With only HM DCC equipped locos on the track and a 15v DC supply, you can operate them via the HM DCC app, that would establish if the issue is with an excessive current demand from the older type motors, of if your eLink is causing the problem.

Edit

I should have specified a ‘suitable 15v DC supply’ not all have the appropriate short circuit detection that is required for a model railway application. I have the engineering preference for using the right tool/product for the job 😁

[Paul-365510]

swapped the eLink for another. No difference.

so it's trying a powerbank followed by trying a zimo decoder.

Or I could pop the LaisDcc Pangu decoder from the 0-4-0 into one of the others but that feels like a way to have no working trains at all...

[Paul-365510]

ok, I did not know that.

apologies to Brew Man for not realising his suggestion was sound.

I'll try this next.

[Fishmanoz]

This will only make a difference if your eLink is faulty, a working eLink will deliver over 3 Amps with the 4 Amp supply. And you seem to have eliminated that possibility by swapping eLinks.

Using an eLink to power the track is a recommended way to do it. Any NMRA compliant DCC controller can be used. Don’t worry about the DCC waveform as it is suitable to power HM7000 decoders.

Also, I wouldn’t connect the laptop PSU directly to the track unless you are sure it has adequate overload/short circuit protection.

But all of that still leaves you with your problem unsolved. The only other suggestion I have is to check the Ringfield FS with a DC stall current test to ensure it is not drawing excessive current, usually caused by weak magnets. Google Brian Lambert then go to his DCC pages for how to do it. And you’ll need a multimeter.

[P-Henny]

........Google Brian Lambert then go to his DCC pages for how to do it. And you’ll need a multimeter.

 

 

No need to Google. There is a direct link to Brian's 'stall test' in the 'Useful Links' pinned sticky thread at the top of the 'General Discussion' forum.

[Ighten 1707822464]

I fitted an 8 pin tsx into a ringfield here. But as Fish points out it is important you do a stress test..

One thing though just have a good check of all the wiring on your chip if it's an 8 pin. Of several 8 pins I have had this was the first to have an issue with its soldering - which I had to re do.

One other thing - what does happen when you just connect a suitable psu and use the app. Have you completely confirmed that you have converted to DCC correctly.

[P-Henny]

To add to lghten's reply where he suggests you check if the conversion was done correctly.

There are different types of Ringfield motor, each needing their own Decoder fitting technique. Insulation of track feeds etc. You may not have done the decoder installation correctly for your model even if following the Hornby instructions you linked to in your OP. There are a multitude of Flying Scotsman variations over the years, so you cannot assume that the linked Hornby instructions are the correct one for your particular F.S model.

If you go to the 'Useful Links' pinned sticky I mentioned in my previous posted reply. You will find a link to Brian's 'Ringfield Conversion' page directly below the 'stall test' linked page I documented. This linked page documents the different Ringfield types and how to convert them.

[Paul-365510]

update. Hornby powerbank arrived - put it in the J83. It now works fine. I assume this must mean that there is a marginal power draw issue when starting from a standstill and the powerbank gets it over that hump. Unless someone advises otherwise I'll assume it's ok to run it now and I'm not doing anything any harm.

I'll not try the same trick on the FS until I've followed the good advice on measuring the stall current and taken a look at the different FS conversions that I was pointed to.

btw, the power supply I'm using to replace the original eLink PSU is one of these: Maplin 65W Universal Laptop Charger Power Supply with 9 interchange tips | Maplin | The Electronics Specialist It does have overload protection and short circuit protection, delivers 3.42A 65W and the output is either 15v or 19v depending on the connector you plug in. The connector that works with the eLink shows it supplying just a smidge over 15v. It's less than half the price of the "official" 4A Hornby PSU so I'll stick with it since I can really only run 3 trains on my layout simultaneously.

[96RAF]

@paul365

Assume for now powering the rails direct and not via a DCC controller . . . This will prove the new PSU separate from the eLink.

That PSU seems OK on paper, but for peace of mind I would power a separate piece of track with the PSU and stick a meter across the rails then drop a coin on track to create a short and see what the meter tells you afore the PSU trips. These units tend to be constant current so in a short circuit situation the thing will try to compensate by tweaking the voltage until it gives up on overload, some may pulse in the protective state. That ramping behaviour is what your decoders will see if a short happens, usually from a derail.

Once happy use it to power your eLink and rely upon the controller’s built in protection as your first safety fence.

[Paul-365510]

thanks 96RAF for that advice. A bit nerve-wracking to deliberately short out a PSU that can deliver 4A, but not quite as scary as dropping a spanner on a busbar...

anyway, tried it out and it fails nicely. Voltage drops to zero rapidly, and comes back just as rapidly when the short is removed. no heating of the PSU over 10+ seconds of shorting.

Attempted the stall current test on the FS motor. At full throttle with the wheels slipping, the amperage fluctuated wildly from under 1A to over 3A - not sure if that's to be expected, perhaps the result of the constant resistance change as the wheels spin on the track (strikes me it would be better to take crocodile clips to the motor supply and leave the wheels out of it).

With the wheels stalled for a couple of seconds, the draw dropped to 0.41A. Consistent results. Seems a tad low, I thought it would go up, not down as the back EMF disappeared.

Just for fun I dropped a LaisDCC 8 pin decoder into the FS, sticking with the wiring suggested in the original Hornby support article. It works. Not too happy going over points, but feels like a stay-alive would sort that out. I would like the sound though...

Tempting to revert to the HM7000 but add the powerbank and see how it goes (after checking I'm using the right conversion strategy for this model of FS).

[96RAF]

Good to hear the PSU self protection worked.

Try to find room to wire in a socket which will make replacing decoders a lot easier in future. A simple wired header as sold by Roads and Rails is ideal. https://www.roads-and-rails.co.uk/products/8-pin-dcc-socket-wiring-harness?_pos=8&_sid=3147ff250&_ss=r

[Paul-365510]

that's the plan for the larger engines now that I think I have a clue.

I checked on the conversion method for the FS I have and it's the easier ring field motor to convert so the approach in the original Hornby article was the correct one to follow.

I fitted an Laisdcc decoder and it worked right away, right down to very very slow speed crawl despite not having a stay alive (yet). so the problem only occurs with the HM7000 unfortunately.

The cheapest decoder with sound support other than the HM7000 seems to be the zimo 648 or 645 and both look to be around £150, and I could buy a second-hand FS with the sound conversion already done for that, so I'll leave it here for now.

Many thanks to everyone for the education.

[Paul-365510]

footnote. Might it be possible to add the HM7000 for sound only i.e. not connect it to the motor, just the sugar cube, and run the one train as if it was a consist/double-header?

I'm intrigued enough by the idea to give it a go.

[Brew Man]

I'm sure it's possible but what you would lose is any kind of sync between sound and movement.