Nigelcliffe

Both types mentioned in the thread above work well.   
I have the CK version of the Toolstation (top of thread) device, used it for years.    The pistol style can be quicker to use, but can't easily do "pull insulation back in middle of wire".   

The CK/Toolstation types can vary in how well they cope with fine wire, there is some limited adjustment in the design which can help, but some of it will be down to manufacturing differences. 

I also have a set of the traditional "plier/scissor style" cutters with a diamond shaped cut-out and a screw adjuster for depth of cut;  I use those for very fine work.  I also look after them carefully: stored closed, not abused on the wrong material, etc... 

Of course, if you're worried about data loss, you must put special green ink on the wire to stop the bits falling out of the sides of the cut wire.  

- Nigel

If you can't work out the drivers (which may be the case, Win11 is tighter on security), then the options are (a) new device, but Lenz' own ethernet product is reassuringly expensive,  or (b) running the Lenz interface on a different hardware device, such as a Raspberry PI.  

Whether (b) works for you depends on what software you're using on the PC, and whether it can remote-connect to the Pi.  

The "old way" to tell about phasing was to use a short burst of DC, and observe the cone move.   If both cones moved the same way (in, or, out), then things are in-phase,  if they moved opposite ways, they are out-of-phase.    But tiny modern speakers may not be suitable for such approaches.  

- Nigel

5 minutes ago, chrisb74 said:

Just picked up an M1 for £19 inc postage on eBay 😊

Well done.  As they are a fairly heavy small metal brick, that's very little after posting costs.    

Add some decent on-off-on switches and you're in business.   

DCC-Ex (current production version) works fine.   I have a cut-track on the motor shield, cut following the guidance on the DCC-Ex website.    Official Arduino Motor shield, and an Elegoo Arduino Mega. 

That requires two different voltages supplies so the Arduino voltage regulator isn't overloaded, which I get by
a)   16v DC, to the motor shield. 
b)    same 16v DC,  via a DC-DC converter to approx 7 volts, which feeds the volts-in barrel connector of the Arduino.   The power pins on the stack of shields is therefore powered from the Arduino and its on-board regulator.  

This works fine with JMRI/DecoderPro,  including programming track,  via USB connection.     

I've also added a Hans Tanner "IoTT stick" to the stack (own home-made shield, with a M5-stick in a header), which gives me WiFi control via LocoNet-over-TCP/IP from a computer running JMRI.   That mostly works, with one very specific bug in the programming track reads, which I've reported to Hans (I think its in his LocoNet code, as the IoTT stick programmes correctly via an Engine Driver App connection,  so the correct data is getting from programming track, through the DCC-EX and into the IoTT stick,  just getting lost when the data exits as LocoNet-over-TCP/IP  ).  

- Nigel

4 hours ago, Chuffer Davies said:

Does it depend on whether the chip has the Stayalive control built in or whether a separate interface has been used?  I use small Zimo decoders which need a separate interface wired directly to the common (supply) rails on the chip. I don’t see how the chip could then differentiate from power supplied via the track and power being taken from the capacitor.

They can, and they do differentiate between track-power and stay-alive power.   Read the Zimo manuals :-).    There are settings for "run time without DCC signal", which would be sensible to set quite low in most situations.     I'd expect any decent decoder design to include run-time settings for stay-alives.

The benefits of a large stay-alive are two-fold.  Firstly, the voltage won't decline by a significant amount in normal use, and secondly (on Zimo only), should the loco come to a stop and as it stops it looses track power, then the decoder senses the loss of DCC and deliberately nudges the loco a bit further until DCC is restored or it is out of power.  Moving a motor from stop takes a lot of current. 

Any reasonably decoder should respond to an E-Stop signal (which is the emergency stop on many systems), and stop instantly.  

- Nigel

probably not.   There are many reasons why some decoders are cheap.  Reduced features is one "benefit". 

This is the Common Return version of wiring:  

It is VERY IMPORTANT to have a consistent rail-side for the return (black),  so as a loco moves around the layout, the same side of the loco is always on the return (black) rail.     If not sure, start at one power connector, and select one rail as "return" (black).  Follow that rail around until you get to another power feed, and mark the same side as "return/black".   Keep doing this tracing for every power feed.   (For a circular layout this results in the "return" being "inside rail" or "outside rail" - doesn't matter which, so long as you're consistent).  

Isolated rail breaks only needed in the switched rail (green),  but most would put isolators on the return (black) as well.   

And, yes, this does all work,  you can have a loco in Sidings-1 on Control-A  going in the opposite direction to a loco on Inner Loop on Control-B.  

Kevin's suggestion of using the outer control knobs on the QS seems very sensible;  much more likely to grab the "correct" control knob.   

DCB suggested adding additional on/off switches for sub-sections.    
This shows two additional on/off switches for the "Sidings 1" area.   The first feed (A)  has power all the time from the main on-off-on switches,   the other parts (B,  C)  are individually switched with an additional on/off switch.  That would allow a locomotive to be isolated in either B or C,  whilst another locomotive worked the other parts.  
Similar on/off switches could be used for other sections as required.  

18 minutes ago, SRman said:

..........

My posts at present are aimed solely at ESU using the LokProgrammer, but much of this work is also possible using JMRI Decoder Pro software, which also allows modification to other brands of decoder. I find Zimo sound decoders a little obscure though, so don't tend to make as many drastic changes to the order on those. ..........

You probably need "Zimo input mapping".    That allows you to swap anything to anything.   eg. Swap "project writer's F5" to "throttle F2" and also swap "project writer's F2" to "throttle F17",  etc..   

As you indicate,  for the ESU decoders,  DecoderPro will do exactly the same changes, with everything laid out in a very similar table to that for the LokProgrammer.  

8 minutes ago, chrisb74 said:

I appreciate this is probably like talking to a child, but bear with me 😁. Rather than switching out track sections, would a system work whereby feed 'A' from the transformer ran to an on-off-on switch, the outputs from which ran to the two controllers which subsequently provide power to the inner and outer loops? Feed 'B' could follow a similar path, eventually terminating in the two sidings sections. In my mind, I could then run a loco from the outer loop onto the inner loop via an insulfrog-point crossover without fear of short circuits, as this system would not allow both loops to be fed power simultaneously. Or am I talking out of my hat?

May work, or may be a disaster, I don't fully understand what you're proposing.    And doesn't save anything.  

The earlier proposal is four switches,  not dozens of them.   You must already have the wires as you're proposing to connect one controller to each section.    So, wiring is a matter of connecting up four switches to the existing wiring, and those switches to two of the controllers.   

Each switch is:   Switch "up" for Control A,  "down" for Control B, or "centre" for Off.   
Four switches, either controller can control anything or anywhere.  

( There's an similar switch arrangement for four controllers, but it goes back to needing two M1 transformers...   And that does seem to be overkill for the available running spaces  ). 

- Nigel

12 hours ago, kevo said:

Hi nigel  just to let you know i followed them steps and big relife the decoder is responed ing  ☺️motor spins each way. i will give it a break this eve and try a reassemble the motor  .tomorrow  eve after work . just to re cap  i doint need to use tape etc  as it was dcc ready  is that right  thanks  kevin 

That's good progress.     I'd have a bit of tape around the orange/grey wires at the motor brush connections to be certain they can't touch anything when things are assembled.  (I'd have used heat-shrink on the wires for that, but that's more complication).    

If you have a multi-meter, you should be able to show no connection between the motor brushes and the motor body.  

Then move to the next assembly stages.   You should be able to have assembled chassis, with the decoder red/black not wired to chassis, and test that with the "lash up" contacts.   And then finally, connect the red/black to the chassis block - could even do that checking one wire at a time, with the other wire going to the lash-up connections to prove things, provided you keep your "red/black" to track arrangements very clear in your head.  

Chrisb74's various running suggestions are between "won't work" and "could damage stuff".  

As buying two M1's is considered too expensive, and only two trains running at any one time,  I come back to the "only use half of it". 

With the four sections of track, put the wiring feeds for each to a Double Throw Centre-off switch.   This selects between the two controllers in use.   Thus, you can drive any section of track with any of the two controllers in use.   If going between two sections, you just set the two switches to the same controller.   If wanting to isolate a section, set the switch to the centre-off.   

This could be done with either:

- Single Pole switches, and "common return" wiring (name one rails, eg. "outside" as "common", and all wires to that rail come together, and go to one side of both controller's output,  the other rail goes via the switch).  

- Double Pole switches, and swap both rail feeds between the two controller outputs.   

Suitable switches will be two or three pounds each.  Look for " on - off - on " to indicate the centre off position, and the words "on" must NOT have brackets around it, ie. NOT " (on) - off - (on) " as that indicates a switch which has to be held in position.  

Some people have difficulty getting their heads around "common return", even though its been standard practise for 60+ years on model railways, and its recommended in the Gaugemaster instructions for their controllers.  

- Nigel

1 hour ago, jpendle said:

Gaugemaster sell 2 suitable transformers.

 

https://www.gaugemasterretail.com/gaugemaster-gmc-uq.html

 

John P

Hmm...  Gaugemaster's information is a bit suspect, for their own products..

The UQ four-track controller needs four independent power sources.   So, that's either four of the small WM1 power-bricks, or two of the cased-in-metal M1 transformers (the M1 has two independent outputs). 

The instruction sheet says two of the M1 cased transformers - that's what they use in the fully-assembled version with a mains plug.   

If only wanting two-tracks, then only power half of it, and ignore the other two knobs.   
Which means one M1 cased transformer, or two WM1 bricks.   The M1 would be fine for running point motors, the WM1's might be marginal.   

Unless budget is really tight, get the M1.  

- Nigel

I'm unsure on addressing. 

It might use "blocks of four" addresses.  Thus, CV1=1, means the outputs are 1,2,3,4,  but CV1=2 would be outputs 5,6,7,8,  and thus CV1=20 would result in addresses 77,78,79,80.  

And regardless of which addressing scheme, there's "how does your DCC system maker interpret accessory addresses",  so a value of 1 in CV1, might result in outputs being accessory addresses 1,2,3,4,  or they might be 5,6,7,8.    If applicable, the shift will always be +4.     (There's an ambiguity in the DCC standards). 

The use of CV9 is non-standard, and with no other information to go on, its more guesswork.  
 

1 minute ago, kevo said:

Hi thanks for the replys  no it was only a few seconds  on my test track  and no power at all was aplied  . so it was quickly  taken off . 

If the track had DCC power, that's a toasted decoder.   Its not a matter of "turning up the throttle" - that's just a data signal of 1's and 0's.   The voltage and current is available at all times.  

Motor wires don't matter.  Other than motor may go the wrong way, ie. "forwards" when you want "backwards",  but that's easily dealt with. 

Basic testing and assembly.  

1)  does the motor work on DC (no decoder in sight).   Not sure ?  Use a 9v battery connected to the motor terminals.  If it spins, motor OK. 

2)  connect orange and grey Decoder wires to motor.   Connect temporarily the red+black decoder wires to *programming track*, and read back some CV's (eg. CV1, or CV8).  You should get a value which is not zero and not 255.   That indicates decoder responding to commands.    No sensible reply, decoder is toast, or wiring wrong this far.  

3)  If you pass step 2,  move the red+black temporary wires to the live DCC main track, and try running the decoder on its address (probably 3, you'd have read it in CV1).  Motor should spin up and down.  

4)  Now, connect the decoder red+black to the pickup screw terminals, and re-assemble the loco, being absolutely sure that no stray metal parts of the orange or grey can touch the metal of the chassis - ie. if you weren't tidy at the motor terminals, you could cause a problem.   

33 minutes ago, smokebox said:

When you isolate the motor from the chassis/frames you don't (normaly) need to isolate the body of the motor, just the electrical connection terminals.  Neither of the electrical terminals are normally in contact with the body of the motor.  One exception is ringfield motors, as were quite common in older Hornby models, which often have the left hand motor terminal connected to the metal body of the motor.  This must be remedied before a decoder is wired in.

Agreed,  and I doubt anything was needed to be done with this model.    Originally it was supplied with a decoder socket,  a previous owner took that out and connected the motor wires direct to the pickups, so works on DC.   Simply disconnecting those wires and inserting a decoder would be all that was needed.     (As I explained earlier).   

But, I expect the decoder is toast due to wiring things up the wrong way round.     

Unfortunately, connecting the decoder outputs (orange and grey) to the track power can destroy a decoder.   When you did that, did you put the loco onto a full power track ?   If so, suspect "destroyed decoder".   

If you only put it on a "programming track" which is normally off, except for programming bursts, then it should have survived.

4 hours ago, dj_efk said:

The black Z21 relies on a smart phone, which to me is not ideal (I do prefer a physical control knob).

As others have said, you can have a conventional handset.   From the Roco Z21 product pages, the Z21 supports: 

Roco Maus (numerous types, including the old basic LokMaus devices), 
Lenz LH30/LH100 (and I'd expect LH101), 
Digitrax DT402, DT300, UT4 (and I expect DT500, DT600, UT6), 
Uhlenbrock Daisy II and Fred (and the Piko badged DaisyII version),     
And a few others...   

You should be able to find an acceptable handset in that lot. 

The "other sockets" on the PowerCab layout socket device area all the same, so secondary handsets (ProCab), or an extension to another set of sockets can be added to any of the spare sockets; front or back.     
What is critical is the primary socket, and the correct NCE cable for the PowerCab (as its the Command Station).    That's got the entire layout power going through that cable, up to the PowerCab, and back down again.    

This document from NCE shows one arrangement for extensions, which I think matches Colin's description above: 
https://ncedcc.zendesk.com/hc/en-us/article_attachments/200259415

- Nigel

I doubt it matters, because you're bringing the second cab up "on power" and changing it before it does a cab-network connection to the command station (the PowerCab). 

What does matter is the socket and cable used;  First handset is "PowerCab" and uses the correct socket and cable for operating, and will work on its own.   Second (optional) handset is now a "ProCab", and uses the other socket and other type of cable.  

( You can also extend the socket for the second ProCab to other places around the layout, and the ProCab will be OK if plugged in, start a train, unplug, replug somewhere else.     Whereas the PowerCab MUST stay connected to its socket at all times.   You might want, in time, to change the access and ease of unplugging the PowerCab to avoid mistakes of unplugging the wrong one.  ).   

try this

https://bearwoodamrg.wordpress.com/2016/08/22/fitting-dcc-to-a-tenshodo-or-hanzono-motor-bogie/

2 minutes ago, Chrisr40 said:

Hello Nigel,thanks for your help. Before I realised that this might be the issue I changed cv1 to a new address to see if that would help. Do I need to do anything with that as reading it now just returns what I set it to ? 

Thanks

Chris

Have you read your system manual on how to set addresses ?    Most systems have a command for setting addresses.  

If you're going to set things manually, then you need to set CV29 correctly as well as setting CV1 or CV17/CV18.     The text around the calculator attempts to explain how they all work.   

10 minutes ago, Southernman46 said:

Finally ditched my (very 60 y.o.) old H&M Powermaster a couple of months ago on the basis that it may try to kill me at some point in the future (kept the box though - yes I still had the box 🙄). Surviving on a Duette now - I don't understand DCC nor can I afford the number of decoders needed 😌

Arguably from a safety point of view you may have ditched the wrong one....   

The Morley Vector twin, at under £100, might make a very good replacement for the Duette (Morley has centre-off detents, just like the Duette, but also chucks in some remote handsets for free).   Or there's a Gaugemaster Twin with separate direction switches at closer to £140.     

CV29=38 is a long address (see calculator also referenced below). 

1)  Check its not in a consist.   If CV19 is anything other than zero, then CV19 is setting the consist address.  Change CV19 to 0. 

2)  Assuming your system has a programming track,  the consult the system manual on how to read the loco address.   Most have mechanisms to do this for you.  

3)  If your system cannot read on a programming track,  then use "service mode programming" to Write the address you'd like, following the instructions in the system's manual.   This will overwrite any address currently in the loco. 

4)   If you want to work out a long address manually, the you need to decypher CV17 and CV18.   See calculator I wrote over a dozen years ago.    But you shouldn't need this.  

You don't need the common positive blue.   

The positive of the lights should connect to one of the chassis pickups.  That gives "half-wave" power into the lights, and they then connect to the decoder function outputs (white, yellow, etc..).  

( It will be in the manual for the Lenz decoder, or any other properly documented decoder.  ).