This arose from some comments from a fellow DOGA member a couple of years ago. DOGA had their stand at Watford show that year, I was helping on the Saturday, and I took along the boxfile as a display item , and also something to provide intermittent movement (We had a Hornby Sentinel on it for a while and it looked the part - I really must built my Judith Edge Vanguard Steelman..)
However this also displayed the boxfile's glaring practical weaknesses
When I built it , some years ago, my knowledge and experience on the electrical side was very limited, and the boxfile represented a major step forward - for the first time I was fitting live frog points and point motors to drive them. This led to some mistakes.
Even longer ago I bought several clearance packs of electrical "goodies" from a company called Greenweld. They were job lots of connectors , cables and such like which in a fit of enthusiasm I thought might be useful. Few have been. A rummage in this stash produced an audio cable with a 5 pin DIN plug on each end, and another with a 7 pin DIN at one end and fine wires hanging out of the other end. These, I thought, would make pukka connectors for the boxfile. DIN sockets were duly sourced and we were in business.
Here they are:
The first problem is that those very fine wires are the devil to secure in the screw connections at the back of the Gaugemaster. I tried making them solid with a bit of solder - that just made the job desperately fiddly instead of completely impossible
The second problem is that the points don't throw particularly well. The siding into the wagon hoist is completely reliable, the nearby headshunt can overheat after extended use and the point into the coal siding is a real problem. It worked until I stuck the weighbridge building on top of the motor. Then it would only throw in one direction....
Of course I didn't think to build in a CDU. And since the wiring is hidden inside sealed buildings I'd have to destroy parts of the layout to retro fit one.
My friend recommended an external CDU , which would then allow thicker wires to connect to the Gaugemaster , and make setting up dead easy.
So far so good - and an All Components CDU was duly sourced and has been sitting in the study ever since waiting for me to acquire a round tuit, or more accurately a suitable enclosure.
It was only when I read the instructions that the real problem leapt out and hit me. They recommend using 6A wire, or as a minimum 3A to carry the current to the motor. I don't know what the current rating of the wire in those audio connections is, but it looks well under 1A "layout wire" (7/0.3 wire I believe). Any internal wiring within the file was carried out in blithe ignorance and 1A layout wire.
A little measuring suggests there's 2.4m of extremely thin wire between the 16V AC outputs and the point motor into the coal siding. Plus a couple of foot of 1A wire and various connections. No wonder that point struggles to throw.
A crude hasty test on the remains of a 10m hank of 24/0.3mm wire (say 5-6m) and the 1.2m interboard connector using the multimeter suggested resistance through the audio cable is about 30-40% higher than through a run of 5A wire at least 4 times as long. I'm aware that resistance becomes more serious the more current you push through. Oh heck.....
By this time I'd also come up with the scheme of resurrecting Tramlink by fitting DIN sockets and audio connector as interboard connections to replace the extremely crude arrangement currently in use , whose wiring has come loose on one side, leaving one board dead. Tramlink serves as my DC test track when I dig it out from under the magazines, so something needs to be done. I probably need to replace and relay one point , and the idea of retrofitting point motors to the two points was and is rather appealing. So I could face the same issues there.
Anyway I pressed on, hoping the thing would deliver some improvement. The external connector (the one with a DIN plug at one end) was shortened to about 18". This should remove about 20-25% of the distance from the power source to the furthest point motor (and half the run of wiring to the two points on the first file). Logically therefore , it reduces the total resistance by 20-25% to the worst affected point.
Now if half the power leaving the power supply is lost due to resistance in the wiring, (and the very poor throw of that motor suggests the loss is substantial) a 20-25% cut in resistance should equate to a 20-25% boost to available power at the solenoid. If the loss due to resistance is less than half, the gain in power is less. But if the loss due to resistance is more than 50%, then a 25% cut in resistance would translate into a boost to available power of more than 25% - perhaps significantly more
This is before you add the benefit of upping the current and voltage by using a CDU. If the path from the power supply to the motor is too long in too thin wire, the cumulative resistance can strangle the output from a CDU and you may see very limited benefit
Here are some of the basic components before starting - the Maplins PSU enclosure , the audio cables, and some Maplin grommits (Wallace is not in the shot)
A piece of 5mm balsa wood was glued to the base of the plastic box using aradite, with a strip of doublesided sticky tape under a recessed area and some UHU along the top of the sides. I don't want this base breaking loose if the box gets knocked about . The CDU unit is screwed down to this - the balsa allows for seating of the underside where there are projections caused by soldering components to the circuit board . I took this approach with the MERG decoder I built for Blacklade and it seems to work fine, though there the mounting screws do pass through the balsa onto the ply board top.
A second strip of balsa was wedged/araldited across one end to take a small connector from one of the Greenweld bags. It had tags with loops on one side and a larger tag , presumably for some kind of spade connector , on the other side. These fouled the CDU board , so after 24/0.7 wire ("5A") had been soldered in place the prongs were bent over . These form the connection between the two wires in the audio cable which connect to the track and heavy duty black and red wires which run to the controller .The fine wires from the audio cable were soldered to the tags on the opposite side. I was a little nervous about whether the joints might be dry, as the metal is not terribly good for soldering to - a hasty test with the multimeter gave readings of 0.07- 0.05 on the lowest resistance scale through the entire set up fron DIN plug to the end of the 5A wire - a little under half the value originally measured through a 1.2m audio cable. So the joints are presumably good
The two wires in the audio cable carrying the current to the point motors were then extended with short lengths of "1A" 7/0.7 wire , the joints protected with heatshrink, and the extended ends connected to the output terminals of the CDU
A general view (heavily zoomed and not in perfect focus) of the contraption is shown below
I then set it up, managed to get the wires connected to the correct terminals on the CDU output and we were in business.
Despite my fears , the improvement is dramatic. Instead of throwing with a loud buzzing , the points flick over instantly with a click. Even the point on the second boxfile works perfectly
And connection to the screw terminals of the Gaugemaster is now simple reliable and a matter of a couple of seconds , instead of the previous fiddle trying to get tiny wires caught by the screws
There's a further, unexpected benefit. When I tested the traction current with my lumbering black 05 , running was much surer, smoother and more reliable. Since I'd done nothing dramatic to the traction circuit and I was testing on the second file , with a further audio connector in the path to the motor, this was a real surprise
I can only conclude that the connections at the terminals of the Gaugemaster may have been a significant part of the problem. It looks as if the fine wires were not only fiddly to trap in the connectors, they were making a poor connection even when trapped.
In short , a big improvement all round, and I think I will probably chance my arm and rewire Tramlink using the same set up
- Read more...
- 0 comments
- 1,089 views