Sunday, 11 June 2017

More stay-alive, back with the Class 02

Some time ago, I reported on fitting ceramic stay alive capacitors to my little class 02 shunter.   Subsequently I found out about the performance drop-off in ceramic capacitors near their rated voltage, and that means their performance for stay alive isn't as good as one might hope. 

So, mindful of that,  I've replaced the stay alive unit in the 02.  Out with the ceramics, and replaced with four 220uF, 16v rated Tantalums.   Those are fitted in both cab-sides, two capacitors to each side, so the cab-light in the loco had to go (but it didn't look particularly good, so its main function had been to test whether the stay-alive was working !). 

Net result, about 1/4 of a wheel turn without power at moderate speeds, far more than what's needed to get over a spec of dirt.   And, that's made me think about the DY1 shunter I built over a dozen years ago, and about finding space in the cab for capacitors in that loco. 

Saturday, 30 July 2016

More stay alive experiments

Prompted by a posting on the MERG forums about capacitor performance, I returned to the test bench to see what happens in the real world.

The MERG posting pointed to the performance of Ceramic capacitors as the applied voltage is varied.  At very low voltages (1 or 2v) they achieve their rated capacitance.  But, as the voltage rises towards the rated voltage, the capacitance declines quite dramatically.    In contrast, Tantalum capacitors have a nearly flat graph, with full capacitance at their rated voltage. 

My use of capacitors in stay-alive devices uses capacitors at or near their rated voltages, so the ceramics used in earlier postings were 16v rated and are used at voltages of around 13v. 

A typical Tantalum capacitor is about twice the volume compared to a Ceramic for the same uF and same 16v rating.   So, with a given volume in a loco, one can fit twice the uF of Ceramics to a loco.

Which brings me to the test bench.   My 4mm Scale Y6 Tram loco, fitted with a Mashima motor, High Level gearbox, and underneath, an 0-4-0 with coupling rods linking the wheels.   It was built with a CT DCX75 decoder, and originally fitted with 22 of 100uF Ceramic capacitors. 

An equivalent volume is achieved with 6 of 220uF Tantalum capacitors, or 1320uF in total. 

This was tested on the bench, and appeared to show a small but significant increase in how far a wheel revolves when power is removed.  
So, the first conclusion is that for equivalent volume, the Tantalum capacitor performs a little better than the Ceramic.

The next test was to increase the Tantalums to 2200uF by using 10 capacitors.  On the track, this gives about 1/4 of a wheel turn without power at modest running speeds.  This is substantially more than the about 1/10th of a wheel turn which the Ceramics would show.    So, clearly for the same uF rating, the Tantalums work much better than the Ceramics, which is consistent with the maker's specification sheets of declining performance mentioned above.

Overall conclusion for model building - for the same unit volume, Tantalums work better than Ceramics.  For the same distance travelled, about 0.3 of the uF in Tantalum will give the same movement as 1 unit of uF in Ceramics.  Tantalums are cheaper and available from more suppliers than Ceramics.  So, I'll be using Tantalums first and only if stuck on awkward space going over to Ceramics. 

Sunday, 21 July 2013

Stay Alive Capacitors in 2mm Scale

The last post on this blog was about Stay Alive capacitors for DCC, and this is the same topic, only much smaller. 

At the 2mm Scale Association summer Expo in June 2013, I had a long conversation with Jens Emmermann.  He showed me some locos fitted with tiny ceramic capacitors, and reminded me to read Carsten Berger's Digital1001 website.

Carsten's site has diagrams of CT decoders, including the locations of the positive and ground solder pads.  It also retails ceramic capacitors and other small parts,  the capacitors are about half the price of buying through Farnell in the UK.

The ceramic capacitors are 16v rated, 100uF.  They are rectangular and measure approx. 2.6 x 2.6 x 3.4 mm.   Being rectangular, they can be assembled into larger rectangular blocks.

So, to my test loco.   A class 02 shunter, which is 45mm long over the buffers.  Inside the cab is a space of 8 x 8 x 4mm below one cab window.  This space had carried a lump of lead, but as the loco has a very heavy solid brass chassis block, the lead only made a small difference to total weight. 

A rectangle of nine capacitors, totalling 900uF, was assembled onto a piece of 0.3mm thick PCB.  The upper surface of the capacitors were linked with a strand of wire which helped solder flow between the capacitor poles.   A second small PCB was assembled with the charge/discharge diode/resistor components.  Following advice on Carsten's site, a 16v Zener diode was also fitted across the capacitors as an over-voltage protection device.

Having made the parts, they were wrapped in very thin paper insulation,  I use model making fine masking tape for this, slicing pieces to the required shapes.

The final part of the assembly is the difficult bit, soldering two wires to the back of a CT DCX75 decoder.  The insulated film on the rear of the decoder has to be cut away over the pads, and a very small soldering iron tip is required to solder to the pads.  I filed an old soldering iron tip to the required shape, and I use a temperature controlled iron.  I practised a few times on some scrap PCB. 

This image shows the location of the solder pads:
The positive pad is accessible, being at the edge of the decoder.   The negative is much harder, its small and in the middle of the board. 

I found making the connections onto the decoder to be very difficult, soldering under lots of magnification to see the parts clearly.   It needs very good light, careful fixing of wires so they don't move and a very steady hand.   In future I might pick a different decoder, the solder pads are in better positions on some of the newer and smaller DCX76 models.

The end result seems superb.   I can't actually measure the stay-alive time, but it feels like around 0.2 to 0.3 seconds of running.  The observed behaviour is that the loco doesn't stall on the first tiny spec of dirt on the track. 

And, if its possible to get effective stay-alive in such a small loco, there can't be many locos where it is "impossible".  It's just difficult due to the tiny size of the components.

Monday, 6 August 2012

Decoder Stay Alive, mixing Zimo and TCS

Over the last couple of years, I've fitted a number of locos with "stay alive" capacitors, as detailed in the Zimo decoder manuals.   At the weekend, I bought a TCS KA2 stay alive unit, and have tried it attached to a Zimo decoder.

Bottom line; it works well, keeping the loco running for 20 seconds or so, covering a yard of track with power disconnected !

The KA2 has two wires, blue (to decoder blue) and black/white striped, which goes to the decoder ground ("mass" on some Zimo diagrams).   Compared to the DIY circuit in the Zimo manuals, the KA2 appears to be lacking the inductor (for programming sound and changing firmware) and a discharge resistor.  The discharge resistor is probably a good idea, as the KA2 can keep the Zimo decoder memory running for more than five minutes - put a loco back on the track with its memory still alive, and it will run at the last setting it remembers !!

Internally, the KA2 contains a couple of diodes and a charging resistor (to limit in-rush current), and another component (probably inductor, not sure yet), plus six 1F capacitors rated at 2.7v in series.  This gives a total of 0.16F rated at 16v, which corresponds to the maximum track voltage which TCS recommend for the KA2.   If the shape is wrong, then the capacitors could be removed from the circuit and strung around in a different configuration.  
A DIY version would be possible if equivalently compact capacitors could be found from electronics suppliers - I've found US sources of these, but not yet identified an EU/UK source with modest postal charges. 

Overall, recommended in any loco which might have pickup problems assuming there is enough space for the stay alive unit.   Unfortunately, its usually the smallest locos which have pickup problems, and those often lack any space.

Friday, 24 June 2011

Bench for RSPB Flatford

I am a volunteer for the RSPB in south Suffolk and north Essex. Usually this involves site maintenance in large woods or on fields which are breeding areas for estuarial birds. But, this year has had an unusual project; making a wildlife friendly garden at Flatford.

The Flatford garden sits alongside the National Trust property at Flatford which is world-famous as the scene of the Haywain and several other Constable paintings.

One thing a garden which will get lots of visitors needs are benches. So, wanting something a little different, I produced this bench in the RSPB workshop using off-cuts of chestnut from other parts of the garden project (the chestnut is local, felled as coppicing to maintain a wood, so about as sustainable as it gets). The shape of the bench does not dictate how you sit on it, you can choose any direction of view. Some of the ideas came from a bench made by Trannon for a museum, though with many modifications to suit outdoor use and the locally available timber.

If the bench proves successful in the garden, I'll make a few more. The 2nd version will have some changes to the seat shape, though will keep the octagonal theme.

Thursday, 23 June 2011

YADCV ( Yet another DCC Coupler Video )

I've done another DCC coupler design, in a Farish class 14 diesel. This one is a commission installation for another modeller (like certain famous motor cars, if you have to ask the price, you can't afford it!)

This time the coils are within the N scale NEM coupler pockets, and the DG couplers are on the loco body. This makes maintenance very much simpler as the fragile parts are well protected and the mechanical parts are slightly modified standard aftermarket couplers. The couplers are "handed" to match the stock on the layout in question where DG loops are only fitted at one end of the stock. So, the loco has a normal loop at one end which can be lifted, and a "lifter" at the other which lifts the loop from a wagon off the loco coupler.

I'll post some drawings at some point, but a bit of video to illustrate how it all works. The loco is a Farish Class 14 diesel, with the wheels turned down and re-gauged to run on 2mm Finescale track. Chip is a CT DCX75.

Blogger's video feature seems to be broken, so I've had to put it on YouTube

Sunday, 14 November 2010

The train on line 1 is running itself....

A long time since the last posting, but here is a train one...

For the Warley show this year, I will have a demonstration on DCC and 2mm scale topics. This will include some basic issues, such as how to fit decoders in limited space, and some more advanced topics, such as train detection, interlocking, and automation.

The advanced topics are illustrated with a simple track arrangement which can be run automatically or manually. A fairly rough video is shown below (bring back the old version of Windows Movie Editor, it was easier than the current one !).

The automated shunting sequence includes careful buffing up to the train, and if the loco running has lights, these are switched on/off at appropriate times.

The train detection is through MERG DTC-8 detectors which register current in a track section. Those feed their output to a HDL LocoIO board, which places the information on a LocoNet network. From the network, information is picked up by other LocoIO boards (displays status on LEDs, etc), and via a computer interface, to JMRI. The LocoIO boards also provide input from the manual lever frame, and output to LED signals at the trackside and servo motor drivers to move turnouts.

JMRI provides the interlocking for the signals rules, an on-screen duplicate of the lever frame, and the automation scripts which send instructions to run trains, change signals and turnouts, alter functions on trains, etc.. The scripts respond to the state of the track detectors. JMRI can track a named train around the system, so it is known which train is in which section.

Train movement comes from a Digitrax Zephyr command station, which relays instructions from LocoNet to the track. It also offers the manual running controller.

A second manual handset, a Uhlenbrock Fred, provides a hand-held controller.

And, if any visitors turn up with an iPhone with the WiThrottle Application loaded, they are welcome to connect it into the system and drive the trains on their iPhone.