Typo

Dear Joe,

I believe there is a typo on "Page 4" of this article in the Premium Edition,

That .05" should likely be a 0.5", (somehow gained an extra decimal zero)
based on the fact that it's the largest Capacitor in the associated picture,
and the next smallest is 0.32" in diameter...

Happy Modelling,
Aim to Improve,
Prof Klyzlr

The diode in figure 3 ...

is shown as a 1N4001, as I have used and mentioned in many of my articles and columns.

That works fine in the larger scales.

Rick is working to wring every little bit out of the space in a N-scale locomotive.

A standard diode, such as the 1N4001, will drop about 0.7 volts when the capacitors are running the locomotive.

Thus, if you have, say 12 volts on the track and the capacitors are fully charged, there will only be 11.3 volts or so available, as the diodes in the decoder will drop some voltage. The capacitors won't start supplying current to the decoder until the (blue wire) voltage drops to about 10.6 volts.

We don't have any easy way to help the 0.7 volt loss within the decoder. Outside, we can use Schottky diodes, that only drop about 0.2 volts, picking up about a half volt (5% or so) in "run power" when utilizing the stay-alive capacitors.

One choice is the 1N5819 (1 amp 40 volt) Schottky diode. It is currently available from AllElectronics.com at 3 for $1.

Good work, Rick. Thanks for sharing.

Thanks Bruce.  I will be

Thanks Bruce.  I will be adding the Schottky diode to my "use these parts" list.  Getting everything to fit, especially in diesels, in N-scale is a true jigsaw puzzle.

Klyzir - good catch! (not sure what happened, proofs were fine).

For those who want to work with SMD

Here is a 1 A 40 V SMD Schottky diode. 4 for $1.

https://www.allelectronics.com/item/stps140a/1-amp-40-volt-schottky-rectifier-smd/1.html

Test using Rick's idea.

Hi Rick,

I do not use sound in my N locos but do want slow speed reliable running for switching, I took an Atlas GP15 fitted with a TCS decoder. Set up a bit of a test track with an isolating fish plate, running at mid speed on the Digitrax throttle, which is nearly full scale speed after the last axle passed over the isolated fish plate the loco run a couple of mm carried on by the flywheel, at slow switching speed it stopped instantly, I wired up 2 x 220 UF capacitors then carried out the same test, at half speed the loco run on for 10mm, at slow speed it run on for just 1mm, I then stripped the trucks down, cleaned and freed everything off [surprising how tight they were], repeated the test, at half speed I get about 12mm and low speed 2mm so worth doing, I have drawn up some replacement gears [3D printed] for the trucks to try for more RPM at slow speed so the flywheel should also help, that might take a bit more trial and error though.

Alexander

Interesting Results

Alexander, thanks for posting your results. Clearly, the lower capacitance circuits are not primarily for motoring over long stretches of dead track, but for maintaining sound through a short interruption.  Your results demonstrate the effectiveness of even a small amount of capacitance. 

More interested in slow speed

More interested in slow speed of insulated frogs in GP38's and RS-11.  sound is not on priority or novelty I plan to add.

Flywheels vs Stay Alives (or Batteries)

The problem with attempting to get more "stay alive" capacity out of a flywheel at slow speed is - well - slow speeds. Yes the energy stored in a specific flywheel is proportional to the square of the angular velocity ()¹, with model railroad drive system we are already dealing with very very low flywheel energies. So if you calculate the potential increase in energy derived by changing the gear ratio of the drive you will find very little improvement. But if you compare the energy density of the flywheel (at low RPMs) to capacitors, the capacitors (or even small batteries) win by substantial margin. Given that flywheels provide so little "momentum" at low speeds (even for those still running DC) and modern decoder designs with adequate Stay Alive capacitors can outperform flywheels, the question is why are we not deleting the flywheels and using the space for additional capacitors (or batteries).

I pulled the flywheels off of a Kato N GE P42 and in the process of machining the frames to allow the installation of four SMD caps (Stay Alive). A test setup with four 47 uF tantalum capacitors and Digitrax SDN144KOA decoder kept the engine running for about 1/8 to 1/4 second after the power was cut. This is usually more than enough time for the engine to clear a dead frog or dirty spot on a wheel which the engine could not do with the flywheels. I hope to be able to fit 68 uF caps which will further improve the performance. Note: The several engines with decoders run just fine without flywheels including Bachmann's N GE 44 ton and HO Shays (if the gears don's break).

1. Flywheel Energy Storage,  https://en.wikipedia.org/wiki/Flywheel_energy_storage

Motor speed

Phew ... thank goodness I model railways and not operate NASA equipment.

What removing the flywheels and having all the capacitors in the world will not [for want of a better description] stop the motor "clicking" round the poles as just enough power is applied, at this point the flywheel is useless, as the motor hits a marginally tight spot it stalls, re-gearing will allow the motor to run faster for the same speed,eliminating the potential stall and with Rick's capacitor solution coupled with the minimal flywheel effect should give better slow speed running, after all when tested using JMRI spedometer, my GP15 runs at 140mph on full throttle, re-gearing should allow better use of the throttle range.

Forgot to add on the previous post - using a RRampmeter when the loco is placed on the track the 2 x 220 uF capacitors draw 0.5amp for just under 2 seconds, I suppose there is a complicated formula out there to calculate it accurately but that is good enough for me.

Alexander

... complicated formula - Nope!

Fairly simple formula(s):

Someone is enjoying themself.

Someone is enjoying themself.