eldavo

I don't think I'd go as far as to say it's elitist! SMD is clearly great for saving space and actually is not that difficult to use once you give up on the idea of using a soldering iron. The trick is to use solder paste and a hot air gun. I use an old Black and Decker paint stripping gun and a complete circuit board can be soldered in a matter of seconds. Positioning the components is a bit tricky but the solder paste holds them in place and you solder everything in one go. Just don't drop the components on the carpet! Cheers Dave

I'm sure Gordon said he was never going to build a layout with gradients and definitely never going to build a full train traverser again. Who are you and what have you done with Gordon? Mad. Cheers Dave

My experience is that 5 to 10 seconds is far from uncommon. My network is about 17+ Mbs download with a ping time of 30ish ms. Seeing these sorts of delays on a lot of threads/pages. Cheers Dave

I seem to recall a couple of years back we had photos of exactly that though I think he only got one side done! Cheers Dave

A few snaps of Redbridge Wharf at the Fareham exhibition...

Test

I certainly do. All my devices exhibit this problem to some extent. My household broadband delivers around 20mbits/sec download and right now I am getting 11mbits/sec across my WiFi. The worst performance is with my Google Nexus 10 tablet running Android using the Chrome browser where the performance is frankly dire and puts me off visiting the site (of course this is the device I use most!). The best performance is with a Windows laptop directly connected via ethernet using the Firefox browser where the page loading speed is acceptable but the flickering is still visible although as the load is quick enough is less of a problem. In my view it is a fundamental problem with the way some of the pages of RMweb are delivered. They are simply too complex with a ton of stuff that 90% of users will never use, umpteen bits of JavaScript which are probably unnecessary and to top it all advert streams that deliver varying sizes of content. The latter is an absolute no no when it comes to designing usable interfaces. Many studies have been done on this topic and screen images that flicker and reorganise in view of the user have always been the number 1 annoyance to users. I don't know what you can do about it but as you can see from other posts in this thread I'm not the only one seeing the issues. I'm afraid your software provider has sold you a poorly designed piece of software that is suffering from serious code bloat. I'm sure if we all went out and bought the latest and greatest hardware the problem would go away but that ain't gonna happen. Hopefully this comes across as positive rather than a winge but for someone who was in the business of building complex software systems for 40 years this sort of stuff really gets my goat. Cheers Dave (dinosaur software designer who knows how this stuff works rather than some of the unskilled code jockeys who are churning stuff out these days)

The slowness of page loading is really hacking me off. Apart from the layout of the page changing as the adverts load as things resize the avatars on each post, even in compressed mode, take for ever to load. I've lost count of the number of times I touched a post only to find the stupid system takes me to some other post entirely as the layout has changed. Anybody know of a way of changing settings to get rid of the avatar on every list entry? Cheers Dave (aka fed up of Winchester)

I doubt very much it is down to WiFi channel interference. If this were the case we would have problems with every web site and page loaded even if only intermittently. To my mind the crux of the problem is the complexity of the page layout and the sheer number of different things that have to be loaded for each page. Even if these are cached locally much of the time it causes significant load in the client browser. From my basic experiments this seems to impact Chrome more than other browsers and the older the hardware it is running on the worse it will be. As a grumpy old software developer I would put this down to stuff being developed by folks who have little understanding of the underlying hardware and just add stuff because they can without thinking about the overall performance of the system. Software bloat. I do find using RMweb increasingly frustrating. Cheers Dave

Waton has been abandoned crated up in a corner of the workshop for another year but it's due out to Brockenhurst exhibition this weekend. Stop by and have a chat if you are in the area. Cheers Dave

As everyone else has said this is truly awful news. There is no way the time and effort can be repaid or the emotional attachment healed. As a practical matter I note some have suggested that stock should be removed from layouts and taken to overnight accommodation. If you intend to do this please, please check your insurance very carefully. My understanding of the usual show insurance cover is that your items are covered on the journey from home to the venue, while at the venue and the journey back home. If you remove items from the venue overnight you will likely need your own insurance policy to cover this. Cheers Dave

OK done some more testing. On a fairly aged Windows 7 laptop running Firefox performance is not bad but certainly not snappy and there are still twitches as bits of the page resize. Not such an issue as the complete page loads in a reasonable time. On my Android tablet with Chrome the performance is poor. The layout twitches around as various bits load and things resize. Several seconds to completely load a page. Running Chrome on the Windows 7 PC again the performance is poor and the pages dance around as various bits load. Tried it on a much more recent Windows 10 laptop, again with Chrome, and while better it's still not quick enough to avoid seeing all the bits of the page dancing around as the pages load. Not a definitive set of results but to me it would indicate that there is something in the page layout that is causing Chrome to struggle. Not ideal given the number of devices and users that use Chrome. Taking a look at the page source it would only take a miniscule difference in the download and caching behaviour between browsers to cause mayhem as there are an awful lot of downloaded scripts and other stuff! Seems somewhat OTT to my dinosaur eyes. Cheers Dave

I really don't want to be one of those that looks a gift horse in the mouth but... We've had this new forum software and layout for a while now but there still seem to be a fundamental problems. Maybe it's just me but just about every page that loads from RMweb is slow compared to other sites. I'm not using the latest super turbo whizzy devices and I'm on mere 20Mbs broadband link but there are fundamental issues with the way pages are being served and rendered. You can see the layout twitching about as different elements of the page are being loaded as the sizes of images and other things are being accommodated in the browser rather than being computed ahead of time. This means that you go to touch or click on an item and if there is a delay in an element downloading it can move and you end up clicking on something else. When I was designing user interfaces this was always seen as a big no no as it frustrates the heck out of the users. Much of the general slowness seems to be with loading adverts or sidebar content and the user avatars. I never observed this with the previous software so something is being done differently. I'm not sure what can be done about this but I wouldn't mind betting the developers never test their software on anything other than in-house ethernet or fibre broadband so never see how poor this all is out in the real world with a complex site like RMweb. I have to admit I do less on the site these days as it just too frustrating to use. Cheers Dave

All perfectly doable and not really that complicated. For large numbers of inputs and outputs look at examples of using 74HC595 and 74HC165 chips. Lots of examples on Arduino playground and other places. Also cheap as chips! You can use I2C devices but these basic chips can be strung together very simply. I've never used an external flasher circuit as I've always found the Arduino is fast enough to do the work as long as your code isn't dreadfully inefficient. Timers are easy. Do it all in the loop() function. Grab the current time using the millis() function then simply compare against the time you last made a state change. Multiple ifs don't cost much. Just note down when you last changed a flashing LED from on to off or changed a switch etc. and do a compare. Only use millis() once at the top of your loop() function as it's much more efficient. Happy to share examples of all this if you need it. The Arduino is a great tool for the job. Cheers Dave

Welcome to the homemade point club. I can't imagine you will have a loco that is heavy enough to cause a spread of your switch rails. In have used code 40 flat bottom and have never had a problem. As has been mentioned, clean the rail and sleepers with something like a fibreglass scratchy pen and use decent electrical cored flux solder and you will manage with much less solder. Get the iron in and allow plenty of heat to transfer then feed in a little solder. Keep the iron in place until you see the solder flow nicely. There is no hurry as the rail and sleepers can stand quite a lot of heat. Have fun. Cheers Dave

Haven't spoken to Phil for some while, he's into military dioramas at the moment. Bumped into Geoff at Southampton show a couple of weeks ago. Certainly up for a thrash around the course. Seem to be stuck on handicap of 17 myself. Not over enthusiastic in all this wet and cold. I think you missed the peak of the market for Klear. They've recrntly released a new formulation which some say is even better than the old one though you can't buy it in the UK! Now stop surfing the net and get on with those platforms. Cheers Dave

Great progress big G. Personally I wouldn't use Klear to seal weathering powders. It's sold as a high gloss acrylic but in practice gives something between gloss and satin on models. Keep it for treating your glazing where it really changes the optical characteristics making things look much more like glass. You will need a Matt varnish which as we know is the spawn of the devil and must be used very carefully. Cheers Dave

Getting perilously close to Southampton show so lots of activity around the club. No more soldering on control panels but I can't seem to let the soldering iron go cold. Roped in to produce some railings for the signal box... Sometimes you get a bit sidetracked by small details. Somebody mentioned fire buckets so I had to see if my 3D printer could solve the problem. Looks like it did... These are small, really small. The buckets are a smidgeon under 4mm tall and the handles are about 0.3 of a millimetre across. Lots to do. Cheers Dave

Worth checking you are using the correct Nano definition in the IDE. There have been changes to the bootloader so now there is an option in the later IDE versions to select "... with old bootloader" for used with older Arduino boards. Cheers Dave

A comms line length of 2.5m will probably be OK with I2C. You can slow down the transmission rate to around 30kHz if necessary. We have bus extender chips on all our boards but we have some cables of 10 or more metres. The Arduino implementation of I2C is not the finest of efforts but it works pretty well most of the time. If it didn't we would be in big trouble in a couple of weeks as we are taking our layout to Southampton exhibition. It relies on I2C for 90% of its operation! Cheers Dave

Interesting stuff. We, the Winchester railway modellers, have done something similar on our Redbridge Wharf layout. No mega but a lot of nanos, 4 alone in one of the control panels. Beware long cable lengths with the I2C bus, we had to use bus extender chips. Cheers Dafe

More progress albeit slow. We now have a fiddleyard/main control panel. Here it is in test mode doing an impression of a Christmas tree. Sorry about the reflections from the workshop lights the top panel is somewhat shiny! The idea is that it is fairly simple for the operator. To set a route you simply push the button representing the starting point, then the button for the end point followed by the "set" button. Some routes require the pressing of a "via" button to disambiguate, for instance when traversing the Romsey curve. When a button is pressed a red LED lights next to it to confirm the system saw you press it. These red LEDs flash until you press enough buttons to define a valid route at which point they come on continuously. When the set button has been pushed and the system has set the route appropriate LEDs are lit to show the route set and the red LEDs go out. To unset a route simply press the button corresponding to the destination of the route then press "clear". Up routes are indicated by yellow LEDs, down by blue and shunt moves within the fiddleyard by green LEDs. Simples! Underneath things are not quite so simple... There are few wires under here. Good job it's DCC! There is an LCD information panel (not yet fitted), 3 Arduino microcontrollers, 24 assorted shift register ICs, 41 push to make switches, 144 resistors, about 250 LEDs, a handful of other components and a lot of wire and solder. Scarily it mostly works! More to do. Cheers Dave

The scenery team have been busy with a tin of yacht varnish adding a gloss coat to the water representation... How many times have you seen a log pond modelled on a UK based layout? Cheers Dave

The less pretty progress is typified by this chaotic mess... This is the underside of the main fiddleyard control panel. The external view is (will be) clean and simple allowing any of 300 odd possible routes around the layout to be set up using 2 or 3 presses of the 38 route setting pushbuttons. The panel is about 50% wired up. Every route setting button has a red LED alongside it which provides immediate feedback to the operator, these are all in place. The pushbuttons will be added last. These pushbuttons are monitored and the red LEDs are controlled by the Arduino processor at the bottom right which is connected to a bunch of shift register chips. The majority of clutter on the panel is the wiring of groups of LEDs that provide feedback of the routes currently set. These are colour coded to show up, down and yard shunt routes. Of course it's at this point that I discover I have underestimated the number of LEDs required, I only ordered 200! Lots to do. Cheers Dave