T Gauge Experiments
I have been intrigued by the possibilities of T Gauge (1:450) ever since it first appeared a few years ago. While the original Eishindo product range was quite limited and had some operational and reliability issues, the selection and quality has improved. It is now produced by Railway Shop in Hong Kong. There is also a fair range of UK model body shells from various individuals made using 3D printing. Some modellers have done very impressive things with this new scale - for example, a superb model of Orbost, Australia.
I decided to look at making a small(ish) T Gauge display layout, called Sarum Bridge, basically just a tail-chaser with some simple automation to run 3 or 4 trains. I bought some basic T Gauge stuff to play with - an HST starter set, a few one-way points and other oddments of track. Surprisingly, there is very little useful information and tips on the Web - apparently there used to be an extensive forum, but that has since closed down. There is a new forum - TalkingTGauge - and traffic there is gradually increasing.
After playing with my new train set for a while, I noted the following..
- the HST set is a very nice model. The basic set is a 2+2 configuration (2 power cars and 2 Mk III first class coaches. It runs well, down to reasonably low speeds. The size makes it fiddly, and the hardest part of setting it up is getting the tiny couplers to engage. The directional lighting is a nice touch in this scale, especially since the PWM controller makes it almost constant brightness. I will definitely be getting the extension set (4 Mk III second coaches and a powered buffet car, making up a full 2+7 rake). Update - the full 2+7 rake looks good and runs well.
- the PWM controller is a major improvement over the basic controller supplied with the original sets. However, speed control is not as good as advertised - you can see the train jerking along at low speeds. I hooked up the 'scope and found out why - the pulse frequency is only 100 Hz. This is generally a bit on the low side even for larger scales. A few experiments show that using a higher frequency (200-250Hz) helps, but not as much as expected. The rough running at low speeds is more of a power pickup issue due to the small size. It seems that the designers may have picked 100Hz to get a good deep diesel growl. Technical specifications: plain vanilla square wave PWM, 4V output (when running on 4.5V batteries), 100Hz. The HST runs at scale speed (200kph) at about 50% power (both knob setting and actual duty cycle). The motors stay cool even after a few minutes of running, though there is a noticeable growl. The lowest speed setting still gives about a 1% duty cycle output - just enough to dimly light the head and tail lights and hum quietly. The back EMF from the HST running at scale speed averages about 1V. It is interesting to see the lack of momentum of the tiny motors here - the back EMF declines from 1.25V to 0.75V over the 5ms between pulses, so the train is literally accelerating and decelerating a hundred times a second. Each motor unit draws 20-25mA at full power at 4V, plus another 5mA for the directional LEDs. The controller comes with two optical detectors to reverse the train with an optional 5 second delay. These work well (although the detectors are comparatively huge), so you can easily set up a simple automatic point-to-point layout or various other designs. I have my own T Gauge PWM controller design which offers comparable performance plus full overload protection.
- The motor polarity is reversed compared to larger scales: in N or HO/OO the RH rail is positive when looking in the direction of motion - in T gauge the LH rail is positive.
- the one-way points were a disappointment. These behave like spring-loaded points, except that due to the small size they are implemented differently. The entire point blade mechanism has been removed, and a section of the steel curved rail has been replaced by plastic. Operation relies on the wheels being magnetic - the wheels stay attached to the remaining steel rail and keep the train straight and level. It works beautifully with the HST power cars (all wheels magnetic), but the coaches tend to derail. I don't even want to think about how 4-wheel wagons (plastic wheels at that) would behave. I would expect that the Japanese-outline EMUs would work perfectly since even their unpowered wheels are magnetic (actually, no they aren't). I haven't tried them, but the manual points would probably work OK too - comparable to N gauge insulfrog types.
- gradients and curves are no problem. The HST reliably gets around 132.5mm curves (mid-sized) at full speed (400 kph). The manufacturer suggests a maximum gradient of 6% (1 in 17). I elevated one end of the test oval for 1 in 19 ( about 5%) and the HST coped with only minor speed variations. The 2+2 set is overpowered with each loco hauling one coach, while the full set has 3 locos hauling 6, so I removed one power car and retested. The speed variations were more noticeable, but still acceptable, even on curves. My planned steepest gradient of 1 in 30 should be fine, even with conventional loco-hauled trains.
- keeping track and wheels clean is essential. I am still experimenting with the best methods here. I have already learned to check the wheels and bogies of the power cars under a powerful magnifier with a fine pair of tweezers at hand to remove fluff and dirt.
- railing the stock is much easier than expected - I find the rerailer is not necessary. On the other hand, coupling up is much more difficult. The couplers are quite tight and if not meshed perfectly then one vehicle will lift slightly and cause derailments. With the 3d printed 4-wheel wagons, I found it best to omit the coupler springs and then use a pair of tweezers to squeeze each pair of couplers together. It was also easier to couple up a short rake of 4 or 5 wagons lying on their sides, then put them on the track.
Anyway, I am going ahead with the display layout. I had two basic designs in mind - a wrapped single track oval with one or two passing loops, or a wrapped dogbone or double track oval with no points. The problems with the one-way points means that it will be a dogbone. The size will be 4' 8.5" x 1' 8", and configured to run up to three trains.