Hornby Smoke Units [DCC operation]

[Chrissaf]

What decoder?

[Chrissaf]

Does anyone know about the workings of the Hornby smoke unit fitted to their 1980's range of locos like Stowe. I fitted one of these units to a tender driven Duchess of Sutherland loco. I have recently converted it to DCC. The smoke unit is controlled on a function. The funny thing is when I turn on the smoke unit I get quite a lot of smoke, it then tails off to a whisper, but when I turn it off I also get quite a lot of smoke for a second. If I then turn in on again I get a lot of smoke. It seems as if when the smoke unit is running normally the element gets too hot and the smoke basically vaporises, but turning on and off the element gets to cool so you get more smoke.

Has anyone got any ideas why I get more smoke when I switch it off or on. I am using proper Seuthe Smoke oil.  The resistance of the coil in the smoke unit is about 100 ohms.

[Ruffnut Thorston]

It is possible that the function output is being overloaded by the smoke unit.

 

Depending on the voltage supplied by the function outputs, the smoke unit could be over voltages.

 

This is not good with the Hornby plastic smoke units, as they can and do melt, and can take the locomotive body with them!

 

Take care...

[ColinB]

No, that is not the issue the decoder is working perfectly ok, it is not overloading, I checked all that before I started. The decoder is Zimo, but that makes no difference as its limit is 500 mA, I am drawing about 100 mAmps. No it seems to be a function of smoke oil, I am wondering if the element in the smoke unit gets too hot, so rather than getting smoke the oil just vaporises. The actual smoke unit is screwed to the Duchess's metal chassis, hence why I took it out of the Stowe, so it shouldn't melt the body as it is at least 20 mm away from the body. It is DCC decoder so it will be fed by constant voltage, I could put a resistor in the lead feeding the unit, which in effect would lower the voltage. I just wondered if someone had a clever idea.

[Chrissaf]

A resistor passing about 100mA is going to get hot ... very hot ... maybe even hotter than the smoke unit itself.

.

1N4000 series Silicon Diodes drop 0.6 volts each and run virtually cold. If you were to string a number of diodes in series, say 6 that would drop 3.6 volts from the function feed. A reduced voltage would also reduce the current.

[ColinB]

When you put the resistor in, it won't be passing a 100 mAmps anymore, but point taken. If I use 50 Ohm that would be half a watt at 100 mA., which is not that hot, but it will be lower as it is now passing about 70 mA. In fact seeing as I have a whole chassis I could even use one of those wirewound ones in a case, that you screw down.  Just to test it out I could try the diode trick, if I remember back to my analogue electronics days I think you can do something with a transistor and zener diode. I was really more concerned if anyone had had a previous issue.

[ColinB]

Well after what Chrissaf said, I thought I would try it with no DCC decoder but with straight DC. Would you believe you get most smoke at about 8 volts, above this it smokes in the reservoir but not out the unit.  

[ColinB]

3 diodes in series seems to work, so it obviously works better at about 10 volts. I used 1n4001s to generate the drop, I will go for 1N4148s eventually as they are smaller.