Art Dent Posted January 7, 2019 Share Posted January 7, 2019 (edited) About a month or so ago, I learned that Hornby were no longer going to manufacture the Thomas the Tank Engine range So, as my 3-y-o grandson is train mad ( ), I have bought Thomas, Percy and Thomas' two coaches. As my layout is DCC - the two locos needed decoders fitting, and as Percy is only an 0-4-0 I wanted a cheap decoder that would accept a cheap stay-alive. I finally obtained a LaisDCC decoder (860012) which was cheap-enough at under £10, and tried ou their two stay-alives - the 'small' (860007) and 'large' (860009). The 'small' had virtually NO effect and the 'large' certainly had an effect (running the motor for around 20-25 seconds) but ran the loco in reverse! So, I started looking at RMWeb and other model railway forums as well as trawling the web, to find a suitable circuit to make a small-ish stay-alive that would run the loco for around a half to a second or two. Displayed here is the result of my musings. First a circuit diagam ... ... and secondly a 3D drawing of how to fabricate the unit. I would welcome any comments / suggestions. I am currently fabricating the unit using a 16V 2200uF electrolytic capacitor and I will photo-document its fabrication and testing. Hope this helps, Art Edited January 8, 2019 by Art Dent Link to post Share on other sites More sharing options...
Art Dent Posted January 7, 2019 Author Share Posted January 7, 2019 (edited) Okay - so first test. On the positves ... Nothing blew up. Negatives ... NCE Cab reported 'Short Detected' so I could get no further. So ... either the 100-ohm charging resistor is too low or the Zener voltage of 15V isn't low enough to regulate the voltage across the capacitor properly As I understand it (and my electronics theory is VERY shaky) The Zener will keep the voltage across it at 15V - providing the voltage across the network (Vin) is greater than this. The voltage drop across the 100-ohm charging resistor needs to be factored in too. Given that the rectified DC voltage from the NCE PowerCab is around 13.5V I'm puzzled as the Zener will be reverse-biased and non conducting when there is less than 15V across it. Any ideas? Does anyone out there have the knowledge to work out what size the charging resistor needs to be in order for the 16V capacitor to see the full rectified voltage? Would I be best going to 500-ohm, 1k-ohm, higher?? A zener calculator suggests that 100-ohm or less is about right for a charging current of 5mA - but what is a typical charging current for a 16V 2200uF capacitor when powered by regulated DC of around 13.5V?? Link to calculator: http://www.calculatoredge.com/electronics/zener.htm Would I be better using a Schottky diode instead of the charging resistor (and if so, what specification or type no. e.g. 1N4007 - this being an ordinary diode) Cheers, Art Edited January 7, 2019 by Art Dent Link to post Share on other sites More sharing options...
RMweb Premium Izzy Posted January 8, 2019 RMweb Premium Share Posted January 8, 2019 All you need is the charging resistor and diode on the com pos leg. Forget the rest unless you often use a DCC system that outputs more than 16v, ( I know some do). Most such as NCE/MRC(prodigy), are around the 13.5v mark. I have used 2200uf 16v with such as Bachmann 08 + Hornby TTS decoder like this with no problem. But don’t expect the kind of runtimes you are seeking unless you use much larger capacities than indicated. With the plain electrolytic cap type 2200uf I get no more than with much smaller tantalum packs such as those I have posted about in other threads I.e. 660/880uf, fractions of a second. And of course they need much more space...... Izzy Link to post Share on other sites More sharing options...
Damo666 Posted January 8, 2019 Share Posted January 8, 2019 ... and secondly a 3D drawing of how to fabricate the unit. As I know little or nothing about electronics, but wish that I had the capacity to understand it, can I humbly offer the following observation. The schematic drawing shows the 15v Zener and 100 Ohm Resister R2 wired differently to the 3D image, in that they seem to be connected the other way around to the 1N4007 & 100 Ohm Charging Resistor R1. Link to post Share on other sites More sharing options...
Art Dent Posted January 8, 2019 Author Share Posted January 8, 2019 (edited) Hi All you need is the charging resistor and diode on the com pos leg. Forget the rest unless you often use a DCC system that outputs more than 16v, ( I know some do). Most such as NCE/MRC(prodigy), are around the 13.5v mark. I have used 2200uf 16v with such as Bachmann 08 + Hornby TTS decoder like this with no problem. But don’t expect the kind of runtimes you are seeking unless you use much larger capacities than indicated. With the plain electrolytic cap type 2200uf I get no more than with much smaller tantalum packs such as those I have posted about in other threads I.e. 660/880uf, fractions of a second. And of course they need much more space......Izzy Hi Izzy, It is fractions of a second - specifically around 1/4 to 1/2 second that I'm looking for. I considered losing the Zener as I have a NCE PowerCab system which, having measured it previously, outputs around 13.5V I'm just not terribly comfy with the 'headroom' of a 16V cap on a 13.5v system - as you say, there are systems that output more than 16v (indeed the NMRA standard allows up to 22V track voltage). As I know little or nothing about electronics, but wish that I had the capacity to understand it, can I humbly offer the following observation. The schematic drawing shows the 15v Zener and 100 Ohm Resister R2 wired differently to the 3D image, in that they seem to be connected the other way around to the 1N4007 & 100 Ohm Charging Resistor R1. Hi Damo, They are wired in the same direction. Imagine you are travelling down the wire on the left-hand side through the Zener diode. You meet the anode first, then the cathode (striped end). Continue along the wire up past the positive terminal and down through the 1N4007 rectifying diode. Again you meet the anode first - they are wired in the same direction as per the circuit diagram. The sliver stripe on the diodes marks the cathode. Looking at the circuit diagram and the 3D rendition, the anode of the Zener is connected to the negative of the capacitor and the cathode to the positive side of the capacitor. Looking at the 1N4007 rectifying diode, the anode is connected to the capacitor positive and the cathode is connected to the common positive (blue) wire. The diode marked 1N4007 is correctly wired with the cathode connected to the positive rail. In normal operation, when the stay-alive isn't supplying power and the current flows through the 100-ohm resistor into the capacitor to charge it as the 1N4007 diode is reverse-biased.. When the rail voltage disappears, this diode becomes forward-biased and the capacitor discharges and supplies power to the decoder. The 100k-ohm resistor ensures that the capacitor discharges when the loco (and stay-alive) are removed from the track. As far as I can see - both the circuit diagram and 3D rendition match perfectly. Hope this helps, Art Edited January 8, 2019 by Art Dent Link to post Share on other sites More sharing options...
Hamburger Posted January 8, 2019 Share Posted January 8, 2019 Hi Art, 1.) as izzy says, leave away the Zener. A Zener diode is mandatory if you use a tantalum capacitor or maybe a ceramic type, but an electrolytic capacitor is uncritical regarding overvoltage. 10% more is not a problem. You supply the stay alive from the decoder which has a bridge rectifier, so 1.4 volts are dropped there. 16 V + 10% = 17.6 V 17.6 + 1.4 V drop = 19 V. So 19 V DCC should be the possible maximum (or 13,5 V 'normal' DC voltage due to peaks being 1,4 times higher than the nominal voltage). The voltage problem with stay alives is not the DCC voltage but a possible operation on DC ! 2.) the discharge resistor has a too high value in my opinion, should be between 1 and 5 kOhms. 3.) the charging resistor is ok. Eventually a slightly higher value, example 150 Ohms, could make decoder programming possible without disconnecting the stay alive. Wolf Link to post Share on other sites More sharing options...
Art Dent Posted January 8, 2019 Author Share Posted January 8, 2019 Thanks Hamburger I think I will try 1) just removing the Zener and seeing what effect (if any) that has Depending on the outcome of that test ... 2) lowering the discharge resistor value to - possibly 10k and then 5k and 1k - testing each time (assuming the NCE doesn't still complain when the Zener is removed) ... then ... 3) upping the charging resistor to 150 ohms as you suggest. What is puzzling me is that as I understand the circuit, the charging resistance path is over 50k ohms as both the rectifying and Zener diodes are reverse-biassed and therefore have very large resistances and a very large resistance in parallel with a much smaller one effectively halves the fixed resistor value. Why the PowerCab is seeing around 50k ohms as a short is what I don't understand, since the resistance of motor windings is tiny in comparison and the PowerCab is quite happy with that. What I really need to know is what voltag is being presented to the capacitor through this circuit and what the initial current flow into the capacitor is. I don't know the specifications of the LaisDCC Decoder in regards of maximum current flow but as it isn't an expensive decoder, I'm guessing less than 500mA and probably more like 100mA or less. Finally, I'm not sure if my Zener choice is OK - I think it is but it is actually a 1N5351B (14V, 5W) and not a 1N5352B (15V, 5W) as diagrammed. Don't think it will make any difference. Likewise the rectifying diode - 1N4007 - is a 1k-ohm, 1A diode - again I don't think this will cause any issues. Maybe a 1N4933 (50V, 1A) rectifying diode (item #192217098159) or a SB220 (20V, 2A) Schottky diode (item #192167910922) would be a better bet Art Link to post Share on other sites More sharing options...
Nigelcliffe Posted January 8, 2019 Share Posted January 8, 2019 Discharge resistor is also not necessary on modest capacitors - the capacitor will discharge reasonably quickly by powering the DCC chip, even with nothing connected to it. It becomes useful with supercaps circuits for two reasons related to discharge speed via the DCC chip. One is a loco could go back on track and start doing things, and the other is risk of violent discharge if loco dismantled and capacitors are short circuited. Link to post Share on other sites More sharing options...
Art Dent Posted January 8, 2019 Author Share Posted January 8, 2019 BTW, thought I'd posted the fabrication of this unit - just realised I hadn't. 16V, 2200uF capacitor, Zener and discharge resistor ... ... and soldered up ] 1N4007 rectifying diode and 100-ohm charging resistor: Fabricated unit ready for testing The diagonal wire is soldered together and links the cathode of the Zener with the capacitor's positive terminal. Another view. Unit is only 20mm long as you can see so should fit into all but the tightest of spaces. Sadly, it currently isn't working as I'd hoped - so no installation pictures. Art Link to post Share on other sites More sharing options...
Art Dent Posted January 8, 2019 Author Share Posted January 8, 2019 (edited) Discharge resistor is also not necessary on modest capacitors - the capacitor will discharge reasonably quickly by powering the DCC chip, even with nothing connected to it. It becomes useful with supercaps circuits for two reasons related to discharge speed via the DCC chip. One is a loco could go back on track and start doing things, and the other is risk of violent discharge if loco dismantled and capacitors are short circuited. Thanks Nigel. You are saying that this circuit ... ... should be sufficient? EDIT: Just posted the above, thought a bit then said 'That's what small commercial stay-alives have!' - indeed I even posted a picture of this here: ( http://www.rmweb.co.uk/community/index.php?/topic/140556-which-cheap-small-locomotive-decoder-would-accept-a-stay-alive/?p=3409469 ). Looks like a 100-ohm resistor (marked '101') and a diode to me. Art Edited January 8, 2019 by Art Dent Link to post Share on other sites More sharing options...
RMweb Gold RedgateModels Posted January 8, 2019 RMweb Gold Share Posted January 8, 2019 That's all I do Link to post Share on other sites More sharing options...
Damo666 Posted January 8, 2019 Share Posted January 8, 2019 Hi Damo, They are wired in the same direction. ........>>>>>> >>>>>>........As far as I can see - both the circuit diagram and 3D rendition match perfectly. Hope this helps, Art Hi Art, Thank you for taking the time to explain in detail, and clearly. I feel I've learned a little more. Not wanting to dilute your thread with irrelevant posts, but I appreciate the way you took the time to respond and explain. By looking at the way the current flows in your response it made sense, I started from the black wire in the 2d schematic and got myself confused. Link to post Share on other sites More sharing options...
Art Dent Posted January 9, 2019 Author Share Posted January 9, 2019 (edited) Hi Art, Thank you for taking the time to explain in detail, and clearly. I feel I've learned a little more. Not wanting to dilute your thread with irrelevant posts, but I appreciate the way you took the time to respond and explain. By looking at the way the current flows in your response it made sense, I started from the black wire in the 2d schematic and got myself confused. No problem, Damo, you're very welcome. Your post was in the spirit of trying to help and I really appreciated that. I'm glad that my explanation got you sorted. I must confess that it is easy to get a connection wrong, such as an inverted diode, and when checking you overlook it as you don't always 'see carefully' when 'looking'. Often you see what you expect to see, not what is actually there! To this end I drew the diagrams in MS Paint (much preferred to the 'back of a fag packet stetches' IMHO) and went down each wire of the circuit diagram (in different colours) and matched these to the same 'wires' on the 3D sketch - so I was pretty happy that the circuit diagram and the 3D rendition matched. I llike to do a 3D rendition if I can as this helps in visualising how the components will fit together before the fabrication stage as well as during as I can refer to the sketch and check that's how I've soldered things up. Call it overkill (or OCD) if you like but it works for me. When tracing circuits I always imagine that the wire is a pipe with mice running along (aka electrons in the wire) and this usually works - ensuring that what I've got matches the circuit diagram when it comes to wiring them up. Yet to try removing the Zener and/or changing the size of the discharge resistor. Here is the simplified version (as seen above in post #10) - circuit diagram first ... ... and 3D rendition: Will report back! Cheers, Art Edited January 10, 2019 by Art Dent Link to post Share on other sites More sharing options...
letterspider Posted June 22, 2020 Share Posted June 22, 2020 Did the modifications work? Link to post Share on other sites More sharing options...
.thumb.png.c39225d8487a7d0bc9a07f4b50d7436e.png)
Recommended Posts
Create an account or sign in to comment
You need to be a member in order to leave a comment
Create an account
Sign up for a new account in our community. It's easy!
Register a new accountSign in
Already have an account? Sign in here.
Sign In Now