Saturday 24 March 2018

Garden railway, adding cheap lights to a g scale carriage

There are various commercial solutions available to add lights to a model train carriage. These are perfectly fine and work either from track power of battery power. 
My g scale railways is all battery lowered RC controlled, no track power. The rails haven't been cleaned for over a year.
So track lowered light isn't convenient. And I don't want to keep on buying 9V batteries. So what to do?

Then I saw these us b battery chargers when searching for 18650 batteries for the RC control i was building in one of my locs.
Also contained a 18650 and the full control for it and booster to 5V.

And all for the same price as a 18650 battery. So battery sorted, now the lights.
LED, of course. Use 5V of course. That was not to difficult. I saw 5V usb led string that can be cut after every 3 leds. Perfect. Use connector, even better!

So the battery is protected by the electronic s of the us b charger it's in and will feed the 5V led chain by the boosted voltage the charger gives out.

Now a switch and a way of charging. Can use a micro us b extender cable but it's a bit big and may get clogged if left outside.
I had some small 3.1 mm power sockets and a usb to 3.1mm cable. That sorted it.


Now what about turning the led s on and off?
I played with some switches and these all required holes of taking bits out. Or perhaps a patching ball effect transistor switch? How to turn it off again? Too complicated, better stick to the old kiss: keep it simple silly.

Then I noticed that the carriage had pop in chimneys that were not glued in but had barb shaped books holding them. And the rotated and where kind of rectangular. Lever switch!, I thought. By rotating the chimney the lever switch is operated. Quick rummage through the component bins in the shed and he's i had some.

Advice: make sure you got well stocked odds and sods cupboard full of thingymebobs, ignore the better half complaining about the mess, it'll pay of.

Anyway, got all the parts, just solder and thermal glue. 
I am using One of these cheap but quite good looking Newqida Train carriages. 

These must be about the best value for money in g scale. But peop!d are catching on and China is getting slowly more expensive, so better get them altogether later if you want some. The first i bought for £27 now they are £33 plus £4 shipping, or $99 in the US for 4 plus $27 shipping, it's one of the other.

Then one ish to micro US b cable is needed for power to and from the charger.

One small bit of hole enlarging for the charging connector. Cut it half and use both end.
This is the layout of the components before soldering and glueing.

Some connecting.

And the switch.

Finished roof.


And the switch is a chimney, no ugly exterior additions.

Cost:
Usb to micro USB cable £ 0.99 and up
3.1 mm   socket: £ 0.99. 
LED strip 3 , from a £5.00 2metre strip about 50 on there so guess about £0.30.
Lever switch, £1.50
Usb 2600mA battery pack/charger cheap from ebay £4.95, UK seller so quick
Solder 0.6 meter wire few pence.
Total £9 and some time.

Hope you like the idea. Post a reply if you want to know more.
 


Friday 26 January 2018

How I converted my g scale locomotives to Radio Control, Newqida 99-6001 , Bachmann 91199, old porter 0 4 0 (91041) and LGB locs

Preamble.

For a few years I have a small 45mm track in my garden. I am running G scale or IIm scale trains. It is a small single track loop with a couple additional branches. It run at ground level for some added realism. On a bed of slablayer mix. I find Wickes' slablayer great for tr a k foundation. You dig a 4" 10cm deep thrench will It with slasher and possibly add a tiny bit of cement , water It, trowel it flat it the water leaves marks and leave it. Brill and so far in the 4 years of held up well. It is a single track and not too long, about 25m. It has a main station a halt a branch station when it get to it and had a small goods yard which I will rebuild this summer and add a small branch loop. 

As with any track powered train cleaning the rails for power pick up is less pleasant, even if on g scale it takes a lot less time than on my previous in door N scale. But because it is at ground level you get the occasional grass messing up the power flow.

Therefore a method less reliant on clean rails had to be found. I had seen some videos from the USA all having radio control several years ago but I was hardly available here then. But revisiting this subject YouTube has loads more information it used to have.
AS I do have DCC before I thought about stay alive DCC. I had experience with it. After some procrastination on it I decided to look into RC and on board batteries.

The first G scale train I bought myself was a Newqida 99.6001-4, one of these German type narrow gauge locs. Looks  very much like the LGB one, but if you look better they may have been inspired very much by the same locomotive but there is enough difference to think it can't be a copy of the LGB one. 


Anyway....
The Newqida came with rc and battery pack. Great... I thought.....
Well the loc looks very nice. But the RC is from a simple RC car. Control is a bit abrupt let's say.

Missed opportunity if you ask me. For a few £/$ more for proper RC it could have been a killer, but alas not so. Still, the loc looks good. And is quite robust.

So out with the existing RC and replace it with something else. DCC stay alive perhaps? That would not be easy as there is no power pick up of the track.
Then perhaps DCC by air. Yes.... Well that is twice as expensive as what I paid for the loc, so not quite viable from a financial perspective.

And this is the story how I went about getting Radio Control in my G scale locomotives.

Investigating Radio Control on various sites and youtube I saw Ge Rik's blog on his Peckforton layout. He had some excellent posts on RC and he used Deltang which used some other protocol. And the receiver settings can be changed to accommodate the motor fitted in the locomotive. And you could later expand with sound.

So I thought I give Deltang a go. In first relatively inexpensive Newwida loc it is just about affordable. The most suitable Deltang receiver for the G scale loc was the RX65(B) and I chose the TX22 transmitter kit which can control 12 locs. And I bought an additional transmitter to make a tx21 for programming.

This is the RX65, it's very small, like a Christmas postage stamp.

And this is The Deltang TX22 kit (images from Deltang website)



The kit came with everything you need and the instructions how to build it. If you have a bit of soldering experience it's not hard, really. I'd go for the kit unless you want to source all the components your self, can be done but when you add postage it may work out more expensive. And no I have no affiliation with Deltang or it's resellers. The Rx65 is £40-£45 and the TX22 kit about the same.

Well then what to do with the batteries? 

I don't particularly like the NiCd or NiMH batteries. Last a year and then lots of chemical waste. So I diverted to what many do. It had to be Lithium. I went for the standard 18650 batteries, LG so reliable and hopefully not faked as much like the Panasonic or Samsung ones. The Lithium batteries also have much more energy per volume. A C cell NiMH has about 3Whr where as a 20% larger 18650 Li batt has 10Whr, almost 3x more energy per volume and a low internal resiatance. And the 18650 batteries have become very widespread with e-cigarrettes/vaping and e-cars becoming popular. Guess one thing to thank the tobacco industry for 😀


Lithium batteries can be dangerous and I know people who's kitchen burned down by batteries of a drone. So I need to be dead careful. So serious reading up what to do. I am lucky that I have an very sturdy electronics background. To cut a long story short, I found all work had been done for me. No baking my own pcb necessary. In fact the battery management system that is available ready made, is cheaper than 20% of the components I need to buy to make one myself. So rather then component level design, it's building with "Lego blocks".
There are tonnes of battery management system pcbs. But first I needed to know how many 18650 I could fit in the locomotive. I could cram in 4 with difficulty and 3 normally and 2 with great ease. I need about 7 to 8 volt for full speed drive so to be sure I went for 3 batteries in series getting 10-11V. with hind sight 2 would have been enough but the higher voltage make the motor starts at a slower speed as the pulse width modulated signal has a higher kick to it. So beneficial to have a higher battery voltage than you strictly need.

So add a BMS for 3 lithium batteries from eBay for a couple of pounds.


The BMS does not limit the charging current so I need some current limiting via resistors or some simple electronics. I can make a current control (CC) with a transistor a small resistor and a diodeled but I have used  step down PSU before that had CC. The ready step down or buck PSU module is smaller than a transistor and heatsink and runs 87% efficient from 30V down to 2V about the battery voltage. A no brainer if you also know that the ready PSU pcb is smaller and about the same price of not cheaper and does not get as hot. So another "brick" added. Brill and so much less hassle. Source eBay or Amazon or other online shops, price about £2.00.

Advantage of the CC PSU brick is that I can run a large range of wall warts with it, anything that delivers 0.5A and 14V or higher, even AC as I added a bridge rectifier on the input. I

So I connected the step down PSU to the BMS via a diode that via switch to the RX65. The BMS in turn connects to the batteries. The RX65 has that switch on the input to prevent battery drain when not in use. The diode block current flowing back to the step down PSU when not charging. And of course I put a polyfuse/polyswitch in the power lines to the BMS, just in case, 1.5A 16V. I tested short circuiting and I works perfect, nothing gets hot, nothing blows and I get a very short 20A current spike that dies down rapidly in 2 ways, the polyfuse and the BMS current limit. If the BMS locks then I have to give it a short charge. If the polyfuse fired after 2 shorts then I only need to let it cool down.

Insert drawing.

I found the loc was running too fast. The good thing about the RX65 is that you can program the maximum speed. The programming looks a bit difficult at first but if you print and follow the instructions On the Deltang website it can be done but you do need a TX21 transmitter or reconfigure your TX22 to become a TX21. It is all on the website. Ge Rik's website has a great guide as well and a video on it. I recommend to visit his blog.

Looking at the train running with load I estimated I needed a slower speed of about 60% of what is was. When it did the programming it was after reading a few times the instructions and watching the video so I know what to expect. And it turned out it was quite easy and done in a minute or so.


This Christmas (2017) Santa for me a Bachmann 91199 locomotive.



On my Bachman 91199 Lynn but not Lynn, I could stash 2 18650 batteries in the tanks max without losing the weights and with that lose traction. This gives about 7Volt. But the motor needs about 12V for realistic max speed. No worriesI though, I simply up the voltage using a step up PSU brick. Motor needs about 0.5A so chose a 3A boost PSU brick. These bricks are rated for maximum input current, at 7V in and 13V out 0.5A out at 85% efficiency would need 1.2A input. Always overrate the components with 1.5 to 2 times to have some spare capacity and for longevity so I went for the 3A version.


The BMS was rated for 2A to 2.5A and has over everything protection.


The 2s BMS pcb is really tiny.

The circuitry is slightly different due to the addition of a step up converter PSU.

Insert diagram

I also had a cheap 2nd hand old Bachmann Porter 0 4 0, I think Bachmann 91041. This was a bit in a state. The drive gear wheel had split from using the wrong grease by the previous owner and some other minor bits wrong.


This has quite a small motor that runs at realistic maximum speeds at about 6.5 to 7.5v. I decided to go for 2 lithium 18650 battery and no step up booster.

I added a 2s BMS board and a current control step down psu for charging.


Psu from ebay, 2amp 2s BMS from Amazon.

The 0 4 0 has plenty of space inside. I put the psu and RX65 in each side tank and the BMS next the the two 18650 batteries.


The front had space for on/off push switch and charge socket.


I used melty glue as I nothing else held the type of plastic, not even superglue. I removed all the electrical contacts for the rail power. Replaced the front bulb with an LED.


Then my LGB locomotive. The motor is very robust. And needs some power. I build the same circuitry as used in the Bachmann 91199 with step up booster ps u between batteries and RX65 receiver.. But the BMS current limiting kicks in if you accelerate too fast. I had underestimated the low internal resistance of the LGB motor and maximum current it'll pull when starting to move due to inertia. 

I had to replace the 2A BMS with a 3.5A version. It is only a few mm larger and fits easily. It is a thing to watch if you build a similar circuit.  



Thursday 2 March 2017

Steampunk Speaker Project

Welcome to my blog on my steampunk speaker project.
The speaker is a mix of craft art old electronics and a some new for convenience clad in metal and hardwood. And I could not leave a plasma ball out as fictional energy source.



This is the end results.
I did not want to overdo it on the gears so left it to a minimum.
Control is via keys salvaged from an very old and rusty typewriter; press o and n to switch it on and o  and f to turn it of(f). Volume via a old style valve knob.

Having used the speaker for many months, I find the sounds remarkably good. Better than many a speaker and of course better that these small Bluetooth speakers. So perhaps this is the best large steampunk Bluetooth speaker. I am biased of course ☺

Below you will find the build process, all in chronological order. Hope you enjoy the story.
And I am in the process of building another streamline speaker in the next 12 months. Perhaps with some 3D printer parts and surely some wood and metal.

The build:

01/03/17 

After years of thinking about it I have decided I will finally make my steampunk speaker.
The most important thing about steampunk is of course the look.
There is loads of stuff out there that looks fabulous, but there is not much stuff that actually also works.
The speaker I am building will properly work and will be properly designed to function like a modern speaker.
I will design a "one of" amplifier. Valve amplifier of course!
I made sketches and decided the choice of materials, well to a degree.
For the amplifier I will use either 6P3S or 6550 valves and 6SN7 and 6SL7 valve equivalent of a Russian origin because these are excellent and look great. And for transformer I will use Hammond valve transformers as they look the part and are not ridiculously expensive, you need a 2nd mortgage for some others. I will be using these are nice signal lamps

more expensive that I hoped for
And I have ordered the wooden planks for it.

And here are the valves I shall be using.



























(update: I am using 6 of 6P3S and not 4 of 6550)

08/03/17
I have been looking into stylish fixtures and so on.
First the screws. It's steampunk so it needs to fit in and compliment the other parts. So steel screws are out. Perhaps blackened blacksmith type heavy bolts will do, but for the amount I intent to use these are prohibitively expensive. Copper, cheaper but still too expensive.
That made me think. Lets look for reasonable looking one of any material and then do something about it. So I have bought a copper plating kit and will get coach screw with flange from Screwfix.


The kit and screws are less than half the price and they look better than the best copper ones you can get. Surely I will use some minor PPE kit with that sulphuric acid stuff.


11/03/17
As promised an update on the plating.
Well mixed results from the plating. Some disappointment, some learned.
The first 4 screws went quite well, slow but well. But the zinc plate on the screws seriously hinders the copper plating when I want to speed it up and with the next screws. It seems he sulphuric acid eats the zinc even when using plating voltage. I tried with stainless steel and the copper plate is quite good. Well the lesson I take away from it is that you need the take the zinc of, then Nickel flash it to form a barrier as iron and copper will cause a kind of battery, nickel forms a kind of barrier. But I have no nickel plating kit so I have decided to go for large brass screws. These are more expensive than I had in mind but after potentially contaminating my copper plating kit with the zinc from the screws and needing a nickel flash plating kit I fund the ease of using ready screws more comforting.

I have also tried paint, gold paint as that looks closest to brass, the others looks terrible. And I have asked for quotes from professional platers, I'll see what they charge.
For oldish looking Copper using shine copper spray paint and a hint of mat black and a bit of rubbing works well, follow up with green copper oxide like paint in small faint quantities and it can look convincing.

Anywho, enough about plating.
Let me give you an update on the wooden speaker box itself.
I have cut the wood.

And planed the edges of the planks to make them fit a circle and mounted the first bit.
Then mounted them on the cut wood to form the half barrel like shape.

And this is how it looks.
It was quite some work, but the planks fit very snugly.
You can see the shape I had in mind. A half barrel on the bottom and an arrow like shape as a whole. The inner edges on the top is where the valves will be. I won't give away the rest, watch the blog updates.

The size of the box is quite large due to the speaker I have chosen. I agree, it will be a bit large.
This is worrying my wife and she's already asking where in heavens name I am going to put that thing ;-) bless.

Warning !!! Geek talk in the next paragraph or so, look away now if you have allergic reactions to it.
I have build many speaker, all calculated from the Thiele/Small fundamental parameters from the speakers.
This time I wanted a single full range driver/speaker and not a coaxial type. I want to avoid the phase difference/distortion you get if you use multiple speakers and crossovers. So single full range speaker it is. Other criteria were a reasonable efficiency and acceptable low range. I can't stand speakers that look good on paper but are so inefficient that you need 10kW to hear a faint sound from it. Laziness from the designers if you ask me. Well that is my gripe on speaker designers over with.

Quite some time ago I had a look and I decided on a speaker from Fane. They do a 12" full range speaker that is quite efficient, nicely powerful and has an acceptable low end range.
It's the only one that I could find that fitted all criteria. Amazing that with so many speakers available there are so few proper full range speakers. Well you do have one or two others that seem to be made. But pricing of these is beyond my reach, perhaps they are made from Dodo feathers and Marsupial wolf skin, looking at the prices. Sorry I will now really stop my rant on speaker designers.

So the speaker is a Fane speaker, SOVEREIGN 12-250TC to be more specific.
The main cone is aided by 2 additional smaller cones (sometimes called whizzers, ugly name) that take over when the 12" cone stops emitting sound due to inertia.                              
That's all.


20/03/2017
Finally an update, I was busy with other things like Karate grading and preparations for it and the weather was particularly unpleasant the last several days.
So my work on the speaker has been sporadic and mostly indoors.
Well the wooden box is ready. I'll show in a later update. I need to put some filler on the inside to stop airleaks and rattling of the wood when the speaker is put at higher volumes.

I have aslo designed the valve/tube amplifier and decided to go for a turret board build rather than printed circuit board, more in keeping even though it will be out of sight for safety reasons: just under 400V DC at 1A is not quite healthy.
Anyway, below a pictures of board number 1 with components but no valve yet.



Wiring connections still to be made.
Board 2 will be made next weekend, family permitting.

And I received a Bluetooth USB TF card receiver sound thingy for £2.80 from China via Ebay.
Was skeptical at first but it works great, Really plug and play, brill. it must be the same module or chip set as in my Bluetooth speaker as it uses the same cute Chinese voice to announce it is paired.
I also received some gears as well, ordered various other brass/copper bits. I am playing with the idea of putting a Stirling engine on to use as "power on" method or a flame licker/eater vacuum engine but these are a bit expensive. I still have o find the right model, and it's really too much work to make any of these myself if for £20-£40 you get a small Stirling engine. I am going to procrastinate a bit more on this idea.

23/05/2017 update
I need some control on the speaker and amplifier, for turning on and off and input select.
I wanted to be original and not have toggle switches. Knife switches (Frankenstein) are one option, but when thinking about it, I came across a nice old slightly rusty typewriter. It was on Ebay. It's really nicely in the style. And I won it, brill. I collected it yesterday from this nice chap Pete.

I will wire it up so you need to press O and then N to switch it on, and O and F to turn it off, yes off has 2 fs but that gets more complicated than it is worth. And for the Input selection I do not yet know, perhaps I for input and then a number. and Plus and Minus or < and > for volume, I'll see.
I also got thin copper sheets in and dome head pins that are like small rivets, and some more stuff.
I am so ready for the weekend!

25/03/2017
I have copper clad the front of the box. In pieces held together with rivets.That took a lot longer than anticipated. The whole copper and rivet stuff is just for decoration of course. Its all done with 10cm wide strips of 0.2mm copper in irregular pieces to look like made from whatever was available, and it now looks like this.

Not bad if I say so myself.
Now some meters, the valve amp, indicators, switches and some decorative copper tubing. I am still waiting for one major part for the operation.
I hope I'll have some time for wiring the valve amp soon.


Update 03/04/2017
Lots of stuff done. The 2 part valve amp is almost wired. The bare black Bluetooth control unit pcb and front has been clad and made presentable. Its only small and looks like this:
With some cliche cogs.
I also spray painted some wooden cogs with copper spray paint, but I am not yet convinced they stand closer scrutiny. I may use them or not.
And after deciding on the means of support I ordered and got the wheels in. A great way of alternate use of a utility item. See if you recognize what they are when it's ready.

10/04/2017
After a longish weekend with nearly all of the time spend with the family not much time was left for the speaker.
The power supply is ready and installed.

Control of it is half ready.
And I have the support; I am going to use these wheels.


18/04/2017
The valve amplifier is finished and I have been testing it.
Found that above 205V AC mains (variac) there is some motorboating, where the very low frequency total feedback combined with the phase angle leads to positive feedback rather than negative feedback.
I applied a simple solution of reducing the overall amplification a fraction by adding a moderate resistor in series with the decoupling capacitors on the power valves. That instantly sorted it to regular mains Voltage +15%.

And above the speaker with the PSU spilling out of it's guts. I know, not quite safe yet that 380V DC laying about.
The SD card/radio/USB/bluetooth unit works great, but when in BT mode there is an audible whine. I think I need to separate the digital and analogue ground a bit better and am working on finding the best star point ground location. Easier said than done I have the mains ground, switch mode 12V PSU, this USB-SDcard BT unit, shielding and split in half amplifier to deal with.

29+30 April 2017.
Signal paths are sorted.
And the whole box is ready to be closed and speaker attached. I have manged to finish it! It's all working.
Got over some more minor oscillation issues due to high impedance signal feedback that arose by building everything in, something you hardly see in transistor or chip based electronics.
Tested if the plasma ball would cause major noise issues or "statics" on the electronic, but seems perfectly fine.
Now all is ready and I am happy with the results.
Pictures of it finished I have but at the start.
I may put more on later, like carrying bars on either side, only for decoration.

Hope you enjoyed reading it.

If you want to do something like this yourself here are the cost of the major parts:
Speaker £85
transformers £200
HV PSU £30
valves for amplifier 6x £12, 2* £8
turrets and pertinax board £25
hard wood for box £100
brass screws £40
various decorations £50
time: many hours build, if I were to guess it's over 80 hrs, and many more on preparation, design and sourcing material, don't underestimate the time it takes to get the right material.