Super Capacitor for 'Stay Alive'

[kebang]

Whilst perusing U tube I saw that a guy from the USA had fitted a 12v supercap (22mf) directly to the deoder on his HO loco. This raised a couple of questions for me.

1. surely it would be safer to use a 25v - 35v capacitor? (To prevent overcharge & subsequent explosion)
2. Can a supercap be connected directly to the decoder? (possibly directly to the blue wire & the brown wire fitted to some decoders for stay alive)

In the video the loco ran for at least 6 seconds - more than enough time to negotiate insulated set track points. (I would be happy with 2 seconds. The DCC Concepts stay alive are a fraction of a second = not enough to cross set track points at a crawl)

If the answer to '2' is 'no' I would be grateful if someone could show me a simple circuit. I'm not going to assemble it myself, my eyes & soldering skills are not up to the task - I will pass it to my local electronics firm. (I retired to a small Philippine Island & the young Filipino technicians are intrigued by my locos!)

[GWR_fan]

For starters maybe begin here -

http://www.rmweb.co.uk/forum/viewtopic. ... 0&p=604189

[kebang]

Thanks for that info GWR.... I've read it, very interesting. Hattons have confirmed that their 'own brand' decoders are fitted with the stay alive brown and blue wires. At 8 quid each (Like you - no VAT ) I can afford to risk blowing a few up to see what works!

I will report back here once they arrive & I start to experiment, but it takes 3 weeks for my parcels to arrive here from the UK. They probably get to Manila within a week but by the time they have been processed there & forwarded on.......

Meanwhile if anyone has further comments/advice they are most welcome!

[james2k]

I put stay alives in all my locos (were they fit).

I follow the below method which is out of an ESU loksound manual.


U+ --> Function (Blue) (Pin 16)

GND --> Ground (Brown) (Pin 20)

I tend to solder this directly to the 21 pin board rather than the chip. As it is a bigger pad than the chip itself.

This does work on other models or chips to. But please be cautious!

[Peterm]

Although I make my own stay alive type circuits I'm no expert, but 12v at 22uf supercap doesn't sound right. Besides which, you need to be really careful with them. The circuit that James has put up doesn't have the extra resistor to discharge the cap/cap's when you've finished using the loco' and this can cause problems because of the stored energy. I use electrolytic and super caps with Zimo decoders and a Lenz 100 system. I set the track voltage at 14.5 and use 16v electrolytics or three x supercaps at 5.5v and 1f each if I can find somewhere to put them. I also limit the amount of time the loco can run on cap' power to save embarrassment. I don't know if other decoders can do this.

I'm not knocking James here, but if you go to pp 60 & 61 you'll find a circuit that prevents programming and also drains the cap/s while the loco is out of use.

http://www.zimo.at/web2010/documents/MX ... oder_E.pdf

[kebang]

James & Peter - Thank You very much! This is just the sort of info I was looking for! It would appear (reading page 59 of the link provided by Peter) that where the decoder is pre fitted with stay alive wires then a direct connection to the cap is ok? I will find out when the decoders arrive!

[Suzie]

Most supercapacitors are typically 2.7V or 5.5V and either have to be attached to a power converter to boost the voltage up and charge or are wired in series with voltage sharing circuitry in a module - you are unlikely to find a single capacitor that would be suitable for DCC track voltage that could be connected directly to the blue and black/white wires of a decoder.

There are a few ready made modules around. Lenz make the best one which has to be connected to a Lenz Gold decoder, but most of the others are interchangeable and it should be possible to connect them to any decoder. You have to be a bit careful with track voltage though with some of them - they are not all suitable for full track voltage of up to 24V.

[kebang]

so 3 x 5.5 volt supercaps wired in series (If they are small enough) should be OK, as long as I can be certain that the track voltage does not exceed 15 volts. If these prove to be too large I might have to look at the range of 25 volt caps to see if I can find one small enough with a big uf number.
Thanks

[Gordon H]

so 3 x 5.5 volt supercaps wired in series (If they are small enough) should be OK, as long as I can be certain that the track voltage does not exceed 15 volts.

So long as you also take steps to ensure the voltage is evenly distributed so that no single capacitor sees a voltage outside its individual rating.
Also remember that the total capacitance would only be 1/3 of the individual capacitance as they are in series.

[kebang]

So 3 x 0.022F 5.5 volt capacitors wired in series still provide a capacitance of 0.022F, but the voltage is tripled to 15.5, correct? (fu=22000 in this example, or does it drop to 7333?)

[Peterm]

Wiring in series reduces the capacitance. This is what I do when using supercaps. Wiring in parallel increases capacitance. This is what I use with electrolytics.

[kebang]

Thanks for clearing that up for me - I was using the same rules that apply to batteries....

[Gordon H]

So 3 x 0.022F 5.5 volt capacitors wired in series still provide a capacitance of 0.022F, but the voltage is tripled to 15.5, correct? (fu=22000 in this example, or does it drop to 7333?)

For capacitors of equal value in series the total capacitance reduces by a factor relating to the number of capacitors. For two, it halves, for three it is a third, and so on.
To calculate it for any combination of capacitor values in series you add the reciprocals of each value, then take the reciprocal of the sum.

[kebang]

The hatton decoders finally arrived today! 3 weeks is about par for parcels arriving from England.
As they already come with a tiny capacitor it's pretty obvious who manufactures them.
It's impossible to buy supercaps here - even reputable electronic shops haven't got a clue what I'm talking about - so I will be using standard 16v or 25v caps together with the diode & resistor as shown in a previous post by James. As the 16v are smaller for a given capacity I would like to use those. I'm still waiting for my NEC Power Cab to arrive so until it does anything I make will be tested with an old Hornby Select.
In a recent post by Harry one of the answers to his problem stated that the Select can spike at up to 60v!
(My track voltage appears stable at 15v, but that is measured on the AC scale of a domestic digital tester...).

Anyone have any guesses as to what will happen with a 16v cap, or for that matter a 25v cap if its used with the select?

NEC are now producing their own 'Keep Alive' module. On their website the say do not allow the track voltage to exceed 16v when using this module. I take that to mean that
a) their module comprises of 6 supercaps in series (16.5v), and
b) the power cab will not exceed 16v output.

So do I risk testing with the Select or wait impatiently for the arrival of the Power Cab?

[Peterm]

I know nothing of the select except that it' doesn't conform to dcc standards. I would also think it's a fair bet to say that you can't adjust the track voltage. If the track voltage exceeds to rated cap' voltage, you can get some damage to your loco' or more importantly, yourself.

I would wait and you'll hopefully be able to adjust the output voltage with your powercab.