Switching from DCC to DC whilst in motion ?

[adb968008]

Just an idea I'm kicking round with.

 

My layout is all DC wired from 3 controllers, and covers an 20m diameter circle (well square) shape on 4 levels and 4 separate loops... 80m circular running + sidings etc.

 

With that size, and the fleet I have, moving to DCC means a massive expense in DCC Chips (Hundreds) and a massive amount of rewire/testing taking my layout down for months, if at all.

 

I could however do it in stages, and set up specific sections to run on DCC.

 

This way chipped locos could run on AC/DCC, but it would mean stopping at a station, and replacing the DCC/AC Feed with a DC Feed to switch modes of operation.

 

Whats the possibility of just having a hard swap in section.... as a train crosses over two isolated fishplates separating the DCC section (say on the left  track side of the fishplates) from the DC (on the track right of the fishplates), if I had the DCC speed set for the train, and the DC Controller set up in the right direction/speed as the train crosses from DCC to DC would their be any risk / impact ?

 

Can the chip handle and detect it or could a sudden flip from AC to DC damage it ?

 

 

As I said it's just an idea I'm kicking around.

Edited January 8, 2018 by adb968008

[smokebox]

As the train passes over plastic fishplates it effectively joins the DC to the DCC and vice versa. Damage to the loco, DCC controller and the dc controller is likely.

 

There are ways to do it safely but require a lot more than the simple idea you suggest.

[eldavo]

The chip might survive but as the loco will bridge the DC and DCC controllers together you would almost certainly release copious quantities of magic white smoke and kill one or both of the controllers. The general safe rule is never allow DC and DCC to come together.

 

IMHO the only safe approach is to have a section that can be switched between the two systems. Bring the train to a halt then switch between systems and move off again. Anything else will end in tears.

 

Cheers

Dave

[newbryford]

Just an idea I'm kicking round with.

 

My layout is all DC wired from 3 controllers, and covers an 20m diameter circle (well square) shape on 4 levels and 4 separate loops... 80m circular running + sidings etc.

 

With that size, and the fleet I have, moving to DCC means a massive expense in DCC Chips (Hundreds) and a massive amount of rewire/testing taking my layout down for months, if at all.

 

I could however do it in stages, and set up specific sections to run on DCC.

 

This way chipped locos could run on AC/DCC, but it would mean stopping at a station, and replacing the DCC/AC Feed with a DC Feed to switch modes of operation.

 

Whats the possibility of just having a hard swap in section.... as a train crosses over two isolated fishplates separating the DCC section (say on the left  track side of the fishplates) from the DC (on the track right of the fishplates), if I had the DCC speed set for the train, and the DC Controller set up in the right direction/speed as the train crosses from DCC to DC would their be any risk / impact ?

 

Can the chip handle and detect it or could a sudden flip from AC to DC damage it ?

 

 

As I said it's just an idea I'm kicking around.

 

A basic principle is never mix DC and DCC.

On the move - the insulated gap will be bridged as a loco with multiple pickups passes over it, will probably kill the - edit: DCC (Thanks Beast!)

 

Cheers,

Mick

Edited January 8, 2018 by newbryford

[beast66606]

A basic principle is never mix DC and DCC.

On the move - the insulated gap will be bridged as a loco with multiple pickups passes over it, will probably kill the

 

Cheers,

Mick

 

In next weeks thrilling installment, Mick will reveal what will be killed.

 

PS - he and the others who posted are right - the magic smoke will probably escape if you mix DC and DCC, and if you accidentally stop across the gap ...

[adb968008]

Thanks for the replies.

 

I was kind of expecting the answer I got.

 

I think DCC remains on the back burner a bit longer.

(If I wait long enough, IOT will render DCC obsolete :-)

[Pete the Elaner]

With that size, and the fleet I have, moving to DCC means a massive expense in DCC Chips (Hundreds) and a massive amount of rewire/testing taking my layout down for months, if at all.

That's what I thought but after a short play on a test track I was hooked.

Howe long is it since you ran your 'longest idle' loco. Years I suspect. I decided to chip locos as I wanted to run them 10 years on & I still have loads I have not run yet. I am surprised at how often I dig out a loco & find that I have already chipped it, but have no memory of doing so. Nobody else can have done them. Maybe there is a chip fairy somewhere?

[Crosland]

As a quick test, remove the DC controllers, set all the section switches to on, connect a DCC controller and put a chipped loco on the layout.

 

It may or may not work, to a varying degree, depending on the wiring. Isolate any reversing loops for now.

 

If you think it's a goer then start upgrading the wiring and converting locos.

 

You could leave one or more loops as DC so long as you are very careful never to cross the joins with a loco or metal wheelset.

[AMJ]

For a while I had AC/DC sections to the same layout. Sorry meant DCC and DC.

 

I kept the locomotive depot and goods yards on DC. This allowed me to run fitted and unfitted locomotives on separate sections of the railway.

 

I'd probably do two of the four controller circuit as DCC gradually changing everything over.

 

Just keep DC trains on analogue and DCC on digital.

 

There are some complex instructions about how each chip will react to seeing a DC power supply as certain signalling systems use it for slowing down a moving train before a signal at danger.

[w124bob]

What would happen if the dead section was longer than the loco and a stay alive was fitted as it ran from dcc to dc?

[Nigelcliffe]

What would happen if the dead section was longer than the loco and a stay alive was fitted as it ran from dcc to dc?

 

If the loco gets to the DC section it will either be almost uncontrollable (because nobody set a "run without DCC signal" timeout value), or will stop (because the "run without DCC signal" value reaches its timeout setting). 

 

If in the "almost uncontrollable" state, then turning off the DC volts will, eventually, lead to the loco stopping as the stay-alive discharges.  Then, if volts are re-applied, things get interesting, the loco might start running at the previously set DCC speed/direction, or it might do nothing.     

 

On most locos, for a stay-alive to work the option of DC running is disabled.   Otherwise the decoder can't tell the difference between DC track volts and DC from a stay-alive capacitor. 

 

 

So, in summary, a DCC to DC "dead section" needs to be long enough to stop a stay-alive fitted loco.  And, in general, stay-alive devices are incompatible with DC running.  

 

 

- Nigel

[w124bob]

On the original MERG system you could control a dc loco in the normal way, from memory setting the handset to address 0 allowed control of the dc only loco.

[WIMorrison]

You can also do that on the Lenz system, but only one at a time and a limited time at that.

[Ron Ron Ron]

On the original MERG system you could control a dc loco in the normal way, from memory setting the handset to address 0 allowed control of the dc only loco.

 

 

You can also do that on the Lenz system, but only one at a time and a limited time at that.

 

 

 

...and not to be recommended, unless you want to run the high risk of burning the motors out in your non-DCC models.

 

 

 

.

[Oldddudders]

...and not to be recommended, unless you want to run the high risk of burning the motors out in your non-DCC models.

 .

The high-pitched whistle that the motor emits, even when still, gives us a clue that things are not good. Not recommended except for the very briefest confirmation that the loco runs forwards and backwards. Say less than 10 seconds in each direction, then take the loco off.

[34theletterbetweenB&D]

Lenz at one time produced an 'LT100 Separation module' to enable this functionality automatically without the magic smoke escaping. Detected a short over the rail break and broke the DC connection until the potential short condition on the DCC system was eliminated. So providing the DC controller was set for correct direction the loco carried right on going.  There were some limitations such as unable to cope with split pick up between loco and tender or two bogies, not more than one motor unit on the train and such like. It looked rather appealing when I was about to go from DC to DCC (and presumably it was a facility designed to this end) but once I had sampled DCC in operation: damn the torpedoes, we only live once and you cannot take it with you, mega decoder purchase time.

[Junctionmad]

(If I wait long enough, IOT will render DCC obsolete :-)

 

requiring you no doubt to spend even greater amounts of money !!!!

[w124bob]

I don't how the early MERG system coped with dc only locos but I ran several for to no I'll effect. I'll ask the owner of the layout, he's been involved for many years now. Pretty sure you couldn't do the opposite with a MERG sourced decoder on analogue.

[Nigelcliffe]

I don't how the early MERG system coped with dc only locos but I ran several for to no I'll effect. I'll ask the owner of the layout, he's been involved for many years now. Pretty sure you couldn't do the opposite with a MERG sourced decoder on analogue.

 

According to the MERG technical bulletin for the old system handsets, the old MERG system used address zero and "stretching" of half of the DCC pulses.  Which is the method used by Lenz, Digitrax and others, and several participants in the thread, me included, have said is a bad idea if you value your motors. 

 

Not overly harmful to heavily built iron cored motors (think old Hornby Double-O),  but a serious risk to lighter motor designs from overheating and possibly brush damage. 

 

- Nigel

[Dagworth]

According to the MERG technical bulletin for the old system handsets, the old MERG system used address zero and "stretching" of half of the DCC pulses.  Which is the method used by Lenz, Digitrax and others, and several participants in the thread, me included, have said is a bad idea if you value your motors. 

 

Not overly harmful to heavily built iron cored motors (think old Hornby Double-O),  but a serious risk to lighter motor designs from overheating and possibly brush damage. 

 

- Nigel

I wrote a fairly long but hopefully simple explanation of how this worked in a topic last year, available here http://www.rmweb.co.uk/community/index.php?/topic/120818-dc-models-burning-out-in-analog-mode/?p=2650142 for anyone who wants to know.

 

To my mind though this is the single most important part of that post

(I still wouldn't recommend running DC motors without decoders on DCC layouts but this is the technical of HOW it can be done)

 

Andi

Edited January 8, 2018 by Dagworth

[Nigelcliffe]

Andi' explanation of zero-stretching in the link given is very clear.  Worth reading.  

 

- Nigel

[adb968008]

As a quick test, remove the DC controllers, set all the section switches to on, connect a DCC controller and put a chipped loco on the layout.

 

It may or may not work, to a varying degree, depending on the wiring. Isolate any reversing loops for now.

 

If you think it's a goer then start upgrading the wiring and converting locos.

 

You could leave one or more loops as DC so long as you are very careful never to cross the joins with a loco or metal wheelset.

 

Its a nice idea, but at any one time I have more than 120 locos on my layout, removing that lot would be a chore !

 

My layout is basically wired as "sidings" (I use electrofrog),Each section feeds from a length of track (longest single section is around 8metres, but can be just a few CMs) from outside the point across the split and is isolated before any points converge into a single track again. This is done to avoid any possibility of tripping out the DC feed at points.

I think I have around 80 independent sections, about 8 -10 sections in each loop, a loop being around c20m circumference/c5-6m diameter, plus lots of extras for Isolation in sidings/shed or platforms and /or 8 Shapes etc etc. I also use the points for Isolating lines on DC.

 

Its primitive but each section is joined using 13amp rated connector blocks (male/female) ultimately wired from four independent AC feeds (2x GM M1s), which are connected to a Gaugemaster Walkabout (I have 4 of them) and each GM HH DC out goes to a splitter which goes into each of the four corners (my 4 signal boxes: but in reality are where I wire up my track sections to be live or dead), which uses pairs of connector blocks to connect manually to each section I wish to be live, so the DC gets to where it needs to be, any line can be wired to any of the 4 DC / HH feeds, and I'm thinking could "jump" the AC in to DC block out removing the HandHeld controllers from these circuit and replacing the M1's with the GM Prodigy unit. Then connect up every section to make each section live...all 80 of them. (I'd have 3 redundant DC feed wires then I assume for removal). The wiring is standard Gaugemaster wire they sell in 100m bundles, not sure it's rating.

 

As it's wired this way to avoid shorting out on DC, is this the same for DCC ? (Though I still have an issue of points closing and isolating a line as all my track feeds originate outside a point).

 

Would this work on DCC ?

 

I was really thinking more of a baby step, I have a dog leg  (about 5 metres long) which leads to a 5 platform station/shed/sidings, before descending an incline and joining one of my running circles .. it was this I was thinking to DCC up first, and running down the incline onto a DC loop, though sods law that DC section is quite a large section hence the "DCC to DC" on the move question)

 

My layout I have to say is a mass of tracks (It s proper old school), no space for scenery, but I like the idea of network sprawl and the logistics of managing it (timetable/diagrams etc) personally, rather intricate detail of scenery..

 

The interest spiked as I recently got a DCC sound 47, and a BLI SD40-2 with paragon sound and running on DC it sounds good (I have a Prodigy Advance and rollers for testing on the bench), I'm just wondering the possibilities of going one step further, but as a full time / traveling employee and family at weekend my layout time is limited so I want as always its quick and dirty so I can continue with the action, but don't want to rip up and start again.

Edited January 9, 2018 by adb968008

[Nigelcliffe]

The layout wiring "probably works" for DCC.  Because, fundamentally, wires are wires and they conduct.   As you say you've used quite large cross-section wires and big choc-block connectors, the current handling should be fine.

 

What may cause a minor annoyance is dropping the power in sidings as turnouts are changed.  If power is dropped, then a loco stops.  So sounds stop. And when power is restored as the turnout is moved, a loco may then go through is start-up power sequence.   No actual harm in this, but may be annoying. 

 

 

The issue over combining DC and DCC is one of safety of equipment.  Firstly your connections and switching should ensure that power to a section comes from one system only (and you can't "common return" things either, a lot of DC wiring articles in the past have recommended common return).    
Then the more difficult bit: you need to be totally confident that your operating processes and track switching arrangements will ensure that a loco or rolling stock with metal wheels (just the wheel, doesn't need pickups) NEVER bridges a DC to DCC section.  If confident that won't happen, then you're fine.   But, should that happen even momentarily, there is a significant chance of damage to either the DCC or DC control system.  ( And chances are its the expensive DCC system which goes "pfff" ). 

 

 

There are a few articles around on low-tech means to control current flow such as car indicator bulbs. Those may be worth reading and thinking about. 

[Sir TophamHatt]

1) Save up the hundreds and convert a whole bunch of locos in one go.

 

2) If you have a "ring" that isn't joined (IE, think about the basic Hornby oval), convert one track to DCC, the other keep on DC.  Never run trains onto each others sections.  When enough trains are chipped, dump DC all together and connect the other track.

 

3) Convert what trains you have to DCC and just use those for now.

[Gordon H]

According to the MERG technical bulletin for the old system handsets, the old MERG system used address zero and "stretching" of half of the DCC pulses. 

 

Yes, I remember the discussions with Mike at the time regarding which zeros should be stretched, as their number varies with packet content which could produce variability in the DC offset produced. I believe my suggestion of only stretching the Start bit in each packet byte was the method eventually adopted.

The zero stretching principle is defined in NMRA S9.1

FWIW, the MERG Decod10 and Decod11 decoders were designed to operate with DC traction systems too.