Nigelcliffe

Yes, some people resort to disabling it (bit of paper inside blocking the button making a connection on the PCB), or building a little wall around it (couple of layers of thick plasticard) so the button can't be used or not used by mistake. The Momentum button reprograms CV3 and CV4 to new values. There isn't a "go back to what they were" capability, except for a decoder reset (so you can get back, but its a bit drastic, back to the settings when you got it, including address and any other changes). Sound decoders can have CV3 and CV4 set to values by the sound creator to allow sounds to play out correctly. So, changing them can cause sounds to not play correctly.

What's the minimum duty cycle of the PWM controller ? Some of those on sale on Ebay/Amazon/etc have a lowest cycle which averages out at about 1.5volts, others will go all the way down to zero. You can reduce the 1.5v ones with a couple of ordinary rectifier diodes in series with one of the output wires (before the direction switch), the diodes will drop about 0.7volts each.

You need to use the Recall Stack in the PowerCab. That allows you to resume control of an existing locomotive. If you select a loco manually, then the PowerCab assumes it should be controlled with all functions initially to off. Whether this behaviour is sensible or not is another question, but its what they do.

Duette's are fine for Common Return wiring. You can select a different terminal for each controller as the common. And thus get your required control behaviour. There are a lot of reasons to consider not using antique electrical appliances such as Duette's, but that's a different matter.

I think its possible - in the advanced function mapping. I could only find a German language manual for the decoder, and I'm not proposing to work out all the CV's necessary to do it. The advanced function mapping appears to have the notion of "loco moving" and "loco stationary" (could set both) to control any function output. Therefore I expect it is possible.

Are the back-to-back dimensions of the loco correct ? This is crucial !!

I think you are misunderstanding the manual. CV8 isn't "set" as such. CV8 is a mechanism for changing some internal decoder settings (pseudo-programming as Zimo call it). So, an instruction of CV8=4 (or CV8=3) changes the output of FO3 and FO4. In dependently an instruction of CV8=6 (or CV8=5) changes FO5 and FO6. You can set both pairs of outputs as "normal" or as "logic" or whatever combination you require. (And if you attempt to read CV8, it will always return a value of 145, which is Zimo's manufacturer ID). Zimo manual page 27 for more details.

The things I've made have all been "keep it really simple". So, there isn't an on-off switch, instead the battery connector is easily accessible, and I just disconnect the battery. With battery disconnected its possible to attach the charger to the battery. I could add an on-off switch, but don't find it needed. I doubt that 4mm wagons need more than a single axle drive. Gearmotors with a final bevel drive are more power-efficient than worm drives, so everything else being equal, the battery lasts longer.

You need to buy some stuff and start experimenting. Yes, motorising the wagons seems sensible. Unless you fit something to indicate run time of each wagon, you notice its going to need charging when its not as quick moving as normal - ie. battery starting to die. Typically you'll recharge a single cell battery in under an hour, so you won't need lots of spare wagons, or you arrange the batteries to be easy to swap out of a wagon. I'd expect a single Lipo cell would give you around 30mins to 60mins running of wagons (actually moving), assuming your drive is reasonably efficient and low friction. That should give you a day's exhibition running from the wagons in practise, as most of the time they'll be stationary. Micron Radio Control are one retailer(*) of stuff you need for the radio side (electronic receivers, transmitters, model batteries and suitable charging circuits). If you explain what you're trying to do, I expect they'd give you a recommended set of starting devices to work from. Yes, you can run 5 (or more) wagons independently. Motors will be "how long is your string"; you need things which are geared down to the speed you want the wagons to move - many Radio Controlled toy cars are going to be too fast unless you change the gearing. There are suppliers of motors/gearboxes for radio controlled HO road vehicles which run at sensible speeds, but many of those are in places like Germany, and you've got Brexit hassles to sort out imports. (I've used such things for slow speed controlled model vehicles in 4mm and 2mm, they can work well). (* I've no connection, other than buying a few things from them over the years ).

Just use the same wires with your DCC controller. That should show up any issues. ( Most decoders should support DC running, there are a handful that don't. But its an end-user setting in CV29, so may have been turned off. And they require a clean DC signal, some DC controllers are extremely noisy. ).

How long is your bit of string ? Out of the box, the N gauge Electrofrog switches the frog polarity using the blades. There is no wire to attach, no links to add/remove. This is different to the OO version. Some people report that the contact from blade to fixed rail isn't reliable. (There is a correlation with how point motors are installed, painting track, ballast, and ballast glue, but not always the case). So, they choose to add a wire to the frog, which could be at the frog, or one of the rails away from the frog (ie. next bit of track). Such a wire needs switching polarity, and the switch needs to happen at the same time as the blades moving (otherwise the blades might move, but the switch follows later and that's a short-circuit). Devices like Frog-Juicers can achieve this switching at the same time as the blades move. A smaller group of people go further, and modify the turnout so it has the gaps like the OO version, and electrically the same as the OO version. This is moderately drastic. The Unifrog is a different arrangement. That has a wire attached to the frog, and the wire needs switching (Frog Juicer will do this, or a switch associated with turnout motor). Without using the wire, the Unifrog becomes an "insulfrog". - Nigel

JMRI support is on the group "jmriusers" on groups.io Report the issue there, and follow advice given on checking things. There are many possible reasons for the observed behaviour. "bug" is only one of them. Reports here don't end up at JMRI development. - Nigel

Depends on the quality of the D-connector. Some are really low rating. Some are 4 or 5 Amp. Look for solid or turned pins, rather than pressed pins (proper electronics suppliers like RS, Farnell/CPC, etc.. have specification sheets so this can be checked). In practise, I think you'll be fine. Burntisland-1883 uses D-connectors to distribute power around the layout. I don't know how many are involved, but longest run from command station might be ten of them. - Nigel

Two devices on the network with same address, the Z21's static and a DHCP address assigned by your router to something else. What happens then ? Stuff stops working, or works intermittently - intermittent is the most likely, and you'll be wondering why it appears to work, then doesn't, then works, then doesn't, or does something "odd" (because its received half of what you thought it was told to do). "reservation" is a mechanism to tell the router which is allocating DHCP addresses to treat the Z21 and its address in a specific manner. But, this is well inside "if you know what you're doing" territory. I suspect its simpler to pick up Iain's suggestion of controlling the DHCP address range in your router so the static address on the Z21 falls outside the DHCP range, but even that is well above basic-level of network device setup. Which is back to why Roco ship a router with their Z21...

Body on and then fails suggests something is displacing electrical contact when the body is added.

The Z21 out of the box has a static IP address. If you don't know the dynamic address range of your router and add a device with a static address, then you risk two things with the same address. Your system works by good luck. That luck could run out.

Trouble with all the above.... The reason Roco ship the Z21 with a configured router is the scope for folks to make a dogs-breakfast of their home networking. Half a bit of knowledge about networking is usually the way to get into a mess (been there, in 1986, got the tee-shirts). It should be possible to configure the second router so it takes its DHCP allocations from the "primary" router, and thus everything is on the appropriate number range. Or, if everything in the house (other than the Z21) was on a dynamic address, then just change the house router so it doesn't clash with the supplied Roco unit.

Works but usually unnecessary. Simply connect the LED+resistor negative to the decoder output (white or yellow, or whatever) and the LED=resistor positive to one of the track pickups.

I explained CV29 over a dozen years ago... https://www.2mm.org.uk/articles/cv29 calculator.htm Correctly fitted (you have put them in the right way round ?) the decoders should run the loco without lots of faffing. DCC system, unless you've done something odd to it to put locos into Consists, should be fine. Possibly consider a system stack reset (think that's Op-Switch 36, consult the Digitrax manual). Thereafter, ask the supplier to explain why the decoders don't work. You've bought a Bachmann Loco, with the recommended Bachmann badged decoder to go into it. (The decoder should be decent, its a Zimo with Bachmann packaging). And, just to avoid confusion: Digitrax systems don't do RailCom. However, a Zimo decoder will likely be shipped with RailCom enabled, as its a European decoder, and European systems tend to support RailCom. And the RailCom setting shouldn't matter. (Digitrax systems can do channel-1 of RailCom with additional hardware, but you need to be a DCC-nerd to know what hardware is needed ) - Nigel

The MERG steady-state decoder can be assembled to do the "switch to ground" for the GF signals directly. No additional parts required. One MERG unit would control eight GF signal devices. The combination of the assembly instructions AND the technical notes for the steady-state decoder will cover the build option required - ie. don't just plough through populating everything following only the build instructions; some parts in the kit not required and will need "wire links" in place of them. ( I thought the requirement was to work from existing DS64's). - Nigel

45 turnouts I'd be looking at touch operated computer screen (eg. an Android Tablet, or an old iPad), rather than creating complex switch panels. Which then reduces the work to accessory decoders to power each of the turnout motors, and connecting those to the chosen system to send instructions out. Exactly how it is done depends on choice of system to send instructions to the turnouts. The options using the Prodigy system are very limited, so you may find it cheaper/quicker/simpler to use a different system to send out the turnout instructions (ie. a second DCC system purely to send instructions to the turnouts, which is independent of the track DCC system). That second system can be very cheap (well below £70, possible below £40, including power supplies).

Which brings us back full-circle.... Option a) - costs a little, but definitely works... Relay (or opto-isolator as switch which is smaller and solid state) on output of the DS64. This creates a contact which opens/shuts for the GF board. DS64 configured to operate a Tortoise style motor, so output is continuous. Opto-isolator would be very simple DIY electronics with VeroBoard, an Opto-isolator and a resistor. And probably nothing else needed. Option b) - might work, might not.... Keep trying various combinations of outputs/connections on the DS64. I'd be looking at whether the 1L/1R goes to any of the common connections on the DS64 in one state. If so, that may be part-way to a no-devices solution. Additional diode may be useful.

If you've not tried 1R to Common, then you should. When you said "didn't work" I assumed you'd tried more combinations. Another option if its "half working" would be a series diode in the output, so the accessory decoder is only outputting in one direction. I think its more likely the GF board is looking for a circuit which connects the input to ground (which closing a simple on/off switch over the terminals achieves).

DS64's can be put into "constant output" for devices like tortoise motors. So a simple relay will do the job. I agree that DCC throttle handsets are usually awful for controlling signals/turnouts. However, there are alternative methods including computers (JMRI and other software), or switch panel input devices for various systems. So they don't sell something suitable (or didn't understand the question). I'm sure there will be many accessory decoders which will do the job, as they come, out of the packet. The requirement is electrically very simple; I think the GF controller will be "switch to ground", so all the output of the device will require is "output to ground" in one state. It might even be possible on the DS64 without any additional parts - I don't know how it works, so would need one to measure things. I'd be looking what happens on the outputs (1R, 1L, etc.) in relation to the decoder's ground (0volts). If those outputs go to "ground" when inactive, then that may be the solution for the GF inputs. The use of a relay just ensures that things are correct and isolated.

"Blue" is positive volts. You can also get the positive from the track, though (unless you fit a rectifying circuit) this will be half-wave power. At its simplest, just connect the pickup on one side of the trailer car to the positive side of the LEDs . ( This is how all 6-pin decoders work, and its documented in many of the more thorough decoder maker's manuals ).