Junctionmad
-
Posts
2,486 -
Joined
-
Last visited
Content Type
Profiles
Forums
Blogs
Gallery
Events
Exhibition Layout Details
Store
Posts posted by Junctionmad
-
-
While everyone supports “, but I need, and will increasingly need, a simple and easily maintained system.” , the primary function that makes a system maintainable is to build one that doesn’t need much maintenance . This means quality wiring , good robust connectors ( no nasty chocolate blocks etc ) and professional wiring installs
for example I’m the designer of our big club O gauge , lots of complexity ( 50 MERG CBUS ( can bus) modules ) , DCC etc etc , it’s done 6 exhibitions and the total electrical maintenance over two years was one broken wire ( our diagnostic system identified the culprit in a few seconds )
like u propose we install all the electronic modules , electronic frog switching modules , signal servo & point modules in a “ conduit “ that is an integral part of the baseboard . 120mm high , running all the way round the inside of the layout ( a big roundly roundy ) ( which comprises 16 baseboards )
only 6 wires run between each board carried on a 3 pin XLR and a 4 pin XLR. We have around 1000 metres of wire ( were it not for the electronics we’d have over 5km ! )
Hence whether you build a mimic panel , or a distributed system or whatever , if it’s well built , it’s maintainable
largely because it won’t need much maintenance .
I recently spent considerable time “ debugging “ a large loft OO layout, huge unreliability due to poor cheap chocolate blocks , spaghetti wiring and confused thinking . It self generated all its own maintenance issues
-
Ah I see, I didn’t think you wanted to move the controllers around , as this can’t be done while trains are moving , unlike walk around control in dcc
a radio controlled dc handset would save a lot of wires and connectors
something like this http://www.micronradiocontrol.co.uk/rc_model_rail.html
where the receiver unit wasn’t actually mounted in the loco but into the baseboard wiring would remove considerable wiring and so forth
-
20 hours ago, Bonkers said:
Sorry re read and digested your reply,,, nothing will alter the heat thing, thank you for your quick answer,
As Scotty used to say “ ye canna change the laws of physics “ !
-
6 minutes ago, cctransuk said:
Surely, by having the DC supply for each track section switched between the two controller circuits via a single pole, two way, centre off switch; with common return to the transformer / rectifier, it is not possible for a track section to be connected to more than one of the two controller busses?
..... or am I missing something basic?
John Isherwood.
There was a mention of 7 busses !
I agree for two controllers after that it gets a little more complex
I note you mentioned 2 controllers only but layer you said
“
I think that the way forward is to use 'chocolate block' connectors to support the seven bare copper buswires,
“
Which I took to mean 6 potential controllers and one common return
-
While less common, 2mm dibond is available
-
17 hours ago, Oldddudders said:
And you are proud of anti-social behaviour? I wonder if you've been invited back.
Ah it’s was good fun , we toned it down then , crowded exhibitions aren’t the quietest places
-
there is an excellent description of common return here https://rail.felgall.com/crw.htm
it should be pointed out that as was incorrectly mentioned , is that common return only ever related to the DC side of the controls ie from the outputs of various DC power sources ( which could be transformers etc . It never applies to anything AC ( normally , though common return to isolated power supplies can of course return AC currents )
it should be pointed out that today many DC power sources are switched mode power supplies rather then transformer rectifier systems and many SMPS systems have DC negative bonded to mains earth , so that common return can’t be achieved
the types that have no earth pin clearly are isolated and can be used
With modern dc controllers there is no reason why a given rail can’t be a common rail for a given set of section switches and all dc controllers fed from a common DC supply
the issue arises at controller boundaries where a dissimilar polarity is selected
however good dc can control eliminates that issue because the train should never need to cross a controller boundary to complete its journey
I recently planned out a conversion to cab control for a large dc loft layout. The main issue is to ensure that any given track section couldn’t be switched to two controllers at once , this required interlocking logic on the mimic panel to achieve this. (Mechanical “ radio “ buttons could also be used )
In Johns case there seems to no protection against switching a track section onto multiple controller busses AFAIK
- 1
-
I think there may be some confusion of terms.
Cab control in DC involves a system whereby track sections , typically forming a route from the train start position to its destination , are switched to ONE dc controller. There maybe many controllers and many track sections but the principle remains the same.
this is the opposite to the more simplistic solution of fixed dc controllers switched to fixed sections of track whereby there are “ bridging “ points where the train has to straddle two control zones.
neither of these solutions out of the box need separate Isolated AC Power supplies or common return wiring , ( there are some advantages of isolated power supplies in this sceanario. ) they can all be run on a common DC power feed. Proper cab control can be wired using a common track feed but that’s not necessarily the classic MR common return
Classic common Return means that wiring can be minimised and completely separate return currents. Run back to their respective power sources , hence with isolated supplies even currents of opposite polarity can return to their respective power supplies , ie one section /controller can be opposite polarity to the other but the common return still works
this needs isolated power supplies otherwise the current may flow via a wrong return route ( amongst other things , like double voltage risks )
- 1
-
16 hours ago, fezza said:
Sound is a very personal thing. Personally I turn up the thrash until I feel the need to open the bedroom window and flail...
But maybe that's just me...
we seriously pi$$ed off our fellow exhibitors at one event !!!
-
Nothing better then a clatter of 2-stroke EMD diesels at full bore , lovely lovely , on our big O gauge we had 20 diesels all with “ big” sound running ( the booster registers 9.6A )
what what did you say , can’t hear a thing over this clatter sweet
Railways are big greasy noisy places
-
How will you fold dibond !
-
5 hours ago, Crosland said:
Another problem is the linkage to the point/signal, etc., and the servo position setting. I always use something like an omega loop to ensure that the servo can get all the way to the commanded position without continuously driving against the mechanism.
For points I find I have room for Z loops ( I have a Z forming pliers )
for signals even in O gauge I find it difficult to find linkage space for Z loops so I arrange the link to pop out of the servo arm of over driven
-
I’ve completely switched to the “ opto at servo “ approach and this has eliminated any noise based twitching ,I see no utility in an opto at the pulse sender.
the technical advantage is improved Common mode rejection , and primarily a dramatic decrease in the input resistance of the servo control input. This is the primary cause of interference pickup as the long control wire coupled with >10Mohm CMOS input of the servo control acts like an antenna
remaining sources of twitching are largely from two areas , startup conditions and or internal servo issues.
startup twitches can be eliminated by pull up or down resistors , and or controlled power up sequences , in my case these have been eliminated by software
this leaves internal servo issues , I have seen servos make what’s seems like calibration movements ( digital servos ) from time to time. This is usually a movement to a fixed point and then back to the new commanded position.
this was particular to a brand of micro digital servos and was not present in other brands
removing the pulse sequence, after movement has completed, has some utility , but only works for analog servos. Digital servos will continue to “ hunt” for the last commanded position even if the pulse is removed
of course all this assumes the basic generation of the pulse timing is consistent , I have seen ( and measured ) some inconsistent timing pulses where for example the pulse timing was interrupt based and the priority of the interrupt was too low or where timing was “‘polled “ based and the loop timing consistency was contaminated by other frequent interrupts
these issues are software ones and need fixing anyway
good servos will have a defined “ dead band “ so once the signal jitter is well within the minimum dead band , the servo “ should” stay motionless
dave
-
On 22/08/2020 at 17:52, farren said:
Thank you all for your answers it was mainly just a idea, However I didn’t release Peco did EM track, and I have been thinking a small Diesel MPD layout. this would land itself to a new comer to EM no valve gear to worry about.
EMGS commissioned PECO to make the track, so peco technically don’t sell EM track they are just a contract manufacturer
- 1
-
I would suggest at this stage that ply sleepers ( I cut my own on a proxxon saw from a4 ply sheets ) and functional plastic chairs represent the zenith of model track making , this can be backed up by the occasional rivet based sleeper to add strength or where functional chairs arbt suitable ( or brass shims etc )
It’s relatively cheap ( certainly cheaper then copper clad these days ) works well and looks excellent . They can be a bit flimsy until installed especially around the slide chairs.
the plastic and ply is a better approach imho , then plastic /plastic constructions as it’s possible to reposition the chairs on ply if you need to , whereas it can be a damnation with plastic timbering usually requiring removal of the whole timber
I tend to use brass shims on more complex track work , and cut up functional chairs to make cosmetic covers
I now create jigs in PTFE , on my CNC Miller for common crossing setups !!
in 00 modelling , 00-SF is by far the best compromise and I use it exclusively in point work married to RTR bullhead flexi ( all types except DCC concepts )
in O I use O-MF
- 1
-
The “ old “ typical operating currents for red , green , yellow , blue, leds are largely out of date , as modern small leds are significantly more efficient
for mimic panel and general indicator use I use 2mA as a typical led forward current and even at that, certain blue leds are too intense
- 1
-
If you get the maintenance app, you can configure the Z21 directly, you can use any WiFi router running in simple bridge mode to connect your z21 to your home network
-
I started ( supplied by Millholme Models) building copper clad , then I moved to ply / rivets and then onto ply and plastic chairs and also plastic sleepers and plastic chairs.
i firstly believe, that you need to want to “ model “ track , because quite frankly you can take RTR track and with a few cosmetic mods , good coloring /weathering and ballasting and make it extremely “ believable “ . ( well as believable as 00 can be)
ive built to several 00 standards over the years and learned the pitfalls of several , I certainly regard 00-SF pointwork, flared back to RTR 16.5 plain track as a very good compromise , hand building “ miles” of plain track gets old fast.
I recently evaluated , c&L , Markway , dcc concepts and the new PECO bulkhead. Realistically when finished , ballasted and painted , all will make excellent representations of bullhead track. It’s largely down to personal preference which manufacturers system you want to adopt , PECO is a good balance between looks and cost , C&L is arguably the most prototypical , DCC uses stainless which is a mega pain .
in the end , you can make all this work fine ,
- 1
- 2
-
On 19/08/2020 at 20:09, Theakerr said:
As I understand it the Pololu DRV8825. is specifically for a stepper motor. The L298 on the other hand will drive a servo or, my special interest, a small DC motor. So is their a better unit than the dated L298?
Thanks
What do you mean by “ servo “ ,
-
remember LEDs are designed to be supplied by a constant current , not a constant voltage , hence any power source thats not current limited runs the risk of destroying the leds . Hence with a voltage source like a traffo , resistors are always needed
- 1
-
15 minutes ago, ikcdab said:
It's great when things run of at tangent and get complicated.
Nice controller, but prob over the top for what you need.
Anyway, you have it now. Just make sure that each controller section is totally isolated from the others. Nothing else is required, except for direction and speed matching when a loco goes from one section to another.
As mentioned , this “ may “ work , but often doesn’t
-
The L298 is a very dated product . It does not have any instrinic current control . It has over temp protection but only on a short term basis , it provides a sense current oupu5 , but that in itself isn’t current control
there are far better driver chips these days
-
On 08/08/2020 at 22:49, Ouroborus said:
Good luck with that.
The rest of the world is 649USD. The UK is 847USD. I think the amount and the quoted currency is far from a mistake. He must be a very bitter little man. Still, he/Germany might need all the money they can get soon.
I don’t think he’s the only “ bitter “ person around , perhaps we could move on from this aspect.
- 2
-
Depending on the design of the controller , it is possible to “ speed match “ both dc controllers , in theory , allowing a loco to cross from one circuit to another , this assumes that the polarity of both sections is the same , ie a consistent wiring strategy is followed.
However it is always less then satisfactory and there is always a point where the loco is potential paralleling both controllers together and some controllers cant handle this
this results in speed changes , jumps , and stops etc
switching the track sections to the controller, known as DC cab control, is about the only practical way to handle multi controller implementations , the idea being that only one controller is activated for all the track route/sections need for the train to complete its journey . The situation can become complex , and can be further complicated should you want to incorporate interlocking to prevent users from assigning a track section to more then one controller at a time.
- 1
Boosters for Roco Z21 & G Scale
in DCC Help & Questions
Posted · Edited by Junctionmad
Err not so
we use a Z21 on our O gauge and we quickly exceeded the 3A limit ( it resets and stops by the way ). As we have a single DCC bus of considerable capacity , it’s not possible to split the DCC system.
We used the 10Amp MERG DCC booster kit ( 30 odd quid) and simply fed it from the Z21 track output , essentially not using the 3A Z21 capacity anymore. Works perfect . The booster has its own mains driven DC 15A capable switching power supply
dcc doesnt care whether you have one big booster or multiple smaller boosters , clearly your DCC bus wiring must be up to the job however ( our DCC bus is good to 25A) , each baseboard has local short circuit / overload protection also