Model Railway Forum

Why would that loco need stay alive, is it for the sound.

Walkingthedog wrote: ↑Mon Jun 26, 2023 6:41 pm Why would that loco need stay alive, is it for the sound.

It’s never been a great runner Brian and very susceptible to the slightest spec of dust. Unfortunately with sound decoders they reboot and go through their whole startup routine each time.
I’ve been fitting them to each of the new HM7K by default as they simply plug in and are cheap - completely transformed the way my layout runs.
But you should know with me that one thing always leads to another.

I find that odd, I had one and it ran beautifully very slowly over my not often cleaned track, and it weighed half a hundredweight.

Walkingthedog wrote: ↑Mon Jun 26, 2023 7:43 pm I find that odd, I had one and it ran beautifully very slowly over my not often cleaned track, and it weighed half a hundredweight.

It certainly is a hefty model Brian! But I know that yours would never have let anywhere near a DCC chip let alone sound They tend to show up any weaknesses in the set up. It’s also spent much of its life on display rather than running which may be a factor.Still, it runs like a dream now and it’s slow creep across the complex pointwork on the layout has to be seen to be believed.

Good point Steve.

Well, it didn’t go bang. Made the first diy stay alive this morning and although I do need to optimise the layout, it seem a to be working when hooked up to a TTS decoder and gives up to 5 seconds running time when the power is interrupted.

For the technically minded amongst us, it has three 3v 1F capacitors and to limit the voltage supplied to these to prevent overcharging, the circuit incorporates a 1w 8.2v zener diode. The circuit also has a 1w 100 ohm resistor in parallel with a bog standard IN-4001 diode. I know that excess voltage/current (?) is dissipated as heat through the resistor but although it runs as expected it seems a little warm to the touch.

2023-06-28_11-57-03 by Steve Mumford, on Flickr

I would experiment with using a higher value resistor than the 100 ohm you are using. Maybe 200 ohms. It will slow down the capacitor charge/recharge rate but shouldn't be too bad and the heating effect should be much reduced.

Experimentation may well be the way forward (theory is no use as I don’t know any )
I had thought to start with a higher wattage so that the resistor is less likely to be damaged by the heat but changing the other variable ie ohm rating as suggested, may help.
I noticed on the second one I built (tidied up the layout of components) that the Zener diode also gets a bit toasty - I know it’s running within its tolerances but locked inside a loco body may be asking for trouble.

Now to dabble with a bit of a theory - DCC puts out say, 15v but the capacitors have a total capacity of 9v. The diodes and resistors are in the circuit to limit voltage and direct the current around the circuit. The difference between the two voltages means that they have to dissipate 6v by turning it into heat.
I suppose I could increase the number of capacitors to reduce that differential but that then makes the whole unit too chunky.

I thought the stay alives should go to the DC side of the decoder, not the DCC input? In which case it would be 12v. Happy to be corrected.

They normally connect to the output side of the bridge rectifier, which is at the front end of the decoder and can thus provide power to keep the microprocessor and other electronics working to provide current to the motor.