Stay-alive capacitor hacking

Post all your DCC only problems, solutions and discoverys here.
Brianetta
Posts: 190
Joined: Wed Feb 11, 2009 12:49 pm
Location: Jarrow
Contact:

Stay-alive capacitor hacking

Postby Brianetta » Tue Apr 06, 2010 11:07 pm

I have a Hornby Mixed Goods set, which has been incorporated into my shelf layout, Hepton East. The locos are DCC fitted, but they're not exactly top of the range Hornby kit. The Class 08 and the Jinty share the same chassis (which is the same as Thomas the Tank Engine's chassis) and are prone to stall on or near my insulated points.

I have a lot of insulated points.

The locomotives are clean as a whistle, as is the track. It's just plain bad luck that they keep stalling, as the wheelbase neatly manages to equal the dead areas on adjacent points, and the very slightest unevenness causes an immediate stall unless you're doing a reasonable speed.

Reasonable speed, in the case of the 08, is well above prototypical speed. So, I had a look around this here internet to see what other people did about the problem. The answer seemed to be a DCC keep-alive capacitor, especially if one couldn't afford the Lenz Gold and its Power-1 UPS (which together cost as much as my locomotives did).

Can't be too hard, I thought. So, here we go. The 08 was denuded, and its chip exposed. As you can see, it's almost a standard Hornby chip, except that half the wires are missing and there's no 8-pin socket - just a 4-pin thingy. To the left, the bridge rectifier was fairly easy to identify, and comparison with various photos on the 'net gave me a good idea where to expect the blue lead would have been. The rectifier's rectified side is the inside, and the pad just above the red one is the blue one.

Now, this was my first ever attempt at soldering a surface-mounted component, let alone the fact that it was a working piece of kit for which I had no immediate replacement. The wires are AWG-24 solid cores from a Cat5 cable (no surprise there to anybody who follows my threads). The blue one was attached to the blue pad, and the white-blue to the rectifier's negative output. Hopefully, this means I've connected the two wires to the decoder's internal DC power rails:

Placing this on the track, gingerly powering up the DCC controller and watching carefully for smoke, I was pleased to find that the decoder still ran both ways properly on its normal address. I had managed not only to solder two wires to the decoder, but as a special bonus I'd managed not to break it either.

So, now I have these two wires hanging off, and before I go any further (like, for example, buying a capacitor, diode and resistor) I'd like to gauge the opinion of any DCC and/or electronics experts here.

As you can see in the following image, I'm getting 27 volts off these two wires, even under load. That was way more than I was expecting, and everywhere that shows a potential difference (except the track feed and motor outputs when moving) shows that voltage. Is this normal? I was expecting about 12-15. This is twice that. I understand that diodes have capacitance, and so on, and that I might not be getting a correct reading here, but my digital meter agrees with this needle one:

Anyway, it might be all right. It certainly runs fine, but I'm new to electronics in general, and am aware of how much I don't know.

I'd also like suggestions as to the ratings of resistor, parallel diode and capacitor. Most motor-driving stay-alive caps seem to be 1-10 mF, at 16 or 25 volts. I'm worried about big pops from inside my loco; I don't want to under-specify. I already have a 100R resistor and a 1amp capable diode, which should probably be enough. The diode will just be there to bypass the resistor on discharge, and the resistor's just meant to stop the DCC controller panicking at a perceived short as it charges up.

So, there you go. Half a hack. So far. (-:

Brianetta
Posts: 190
Joined: Wed Feb 11, 2009 12:49 pm
Location: Jarrow
Contact:

Re: Stay-alive capacitor hacking

Postby Brianetta » Wed Apr 07, 2010 10:29 pm

I've measured the output of my Select using both my meters, on AC directly and DC over a (different) bridge rectifier. The AC measurement is 16V (slightly over with the analogue meter, slightly under with the digital). The DC measurement is just under 15V. Anybody have any idea how I'm measuring 27V rectified DC on my decoder chip? Will I really need a 40V rated capacitor, rather than a 25V one?

Gordon H
Posts: 214
Joined: Tue Mar 14, 2006 1:08 pm
Location: UK
Contact:

Re: Stay-alive capacitor hacking

Postby Gordon H » Thu Apr 08, 2010 1:05 pm

Brianetta wrote:I've measured the output of my Select using both my meters, on AC directly and DC over a (different) bridge rectifier. The AC measurement is 16V (slightly over with the analogue meter, slightly under with the digital). The DC measurement is just under 15V.

Did you do this with a capacitor attached to the DC output of the bridge?
This would replicate the situation on the decoder and allow a better comparison to be made.

Brianetta
Posts: 190
Joined: Wed Feb 11, 2009 12:49 pm
Location: Jarrow
Contact:

Re: Stay-alive capacitor hacking

Postby Brianetta » Thu Apr 08, 2010 1:50 pm

No. What sort of capacitor?

If I really have to deal with 27V on this decoder (and all the online articles I've found suggest that this is extraordinary) then I ultimately won't be able to put a big enough capacitor into the loco's body to do what I want. I wonder if putting a couple of diodes in series will eat enough voltage to make a 25V rated capacitor safe?

User avatar
Roger (RJ)
Posts: 1488
Joined: Tue May 29, 2007 2:59 pm
Location: Nottingham, UK.

Re: Stay-alive capacitor hacking

Postby Roger (RJ) » Thu Apr 08, 2010 5:37 pm

I would allow a reasonable margin of safety when choosing a capacitor. For 27 volt working I would choose 30 or better still 35 Volts. At 25 volts you are pushing a 25 volt capacitor to its limits.

A bit like running a car too fast. It may have a top speed of 120mph but if you run it at this speed all the time, it won't last very long.

When a capacitor let's go, it can do a fair bit of damage.

Brianetta
Posts: 190
Joined: Wed Feb 11, 2009 12:49 pm
Location: Jarrow
Contact:

Re: Stay-alive capacitor hacking

Postby Brianetta » Thu Apr 08, 2010 7:06 pm

Time to knock this one on the head, then. It simply can't be done. No capacitor rated at that voltage can both hold enough juice to move a loco and fit inside the loco.

User avatar
S_Jay
Posts: 241
Joined: Thu Dec 27, 2007 4:38 pm
Location: Nottingham
Contact:

Re: Stay-alive capacitor hacking

Postby S_Jay » Thu Apr 08, 2010 9:44 pm

Hi Brianetta,

That's a shame I was reading with interest I have the same problem which I'm avoiding right now. It looks like the only real solutions are

1. Replace all your points with electro-frogs (avoids dead zones but tricker to wire, however your efforts above and with the signal lights this should be no problem to a man with your skills)
2. Try replacing your Hornby with Peco (the gaps are smaller)
3. Can you try and increase the number of wheels being used for pick-up?

:)
My Layout, My world, My rules, if that's ok?
My layout with no progress http://www.newrailwaymodellers.co.uk/Forums/viewtopic.php?t=14866
My other layout with progress http://www.newrailwaymodellers.co.uk/Forums/viewtopic.php?f=22&t=18394

User avatar
Roger (RJ)
Posts: 1488
Joined: Tue May 29, 2007 2:59 pm
Location: Nottingham, UK.

Re: Stay-alive capacitor hacking

Postby Roger (RJ) » Thu Apr 08, 2010 10:09 pm

I'm sorry to have thrown a spanner in the works. There may be a simple solution. Try putting a high value resistor across the points where you get the 27 volts. Try 10k for a start and then see what voltage is present. The slight load may be enough to drop the voltage to a more viable level

Brianetta
Posts: 190
Joined: Wed Feb 11, 2009 12:49 pm
Location: Jarrow
Contact:

Re: Stay-alive capacitor hacking

Postby Brianetta » Thu Apr 08, 2010 10:40 pm

S_jay, all six wheels pick up. I can't use electro-frogs in this geometry, and I'm not about to spend that sort of money on this.

Roger, my worry about a resistor was that as the capacitor charges, its resistance approaches infinite, and it would get the full voltage. I did consider a zener diode. Of course, I'm stuffed if the decoder requires that high a voltage.

User avatar
Roger (RJ)
Posts: 1488
Joined: Tue May 29, 2007 2:59 pm
Location: Nottingham, UK.

Re: Stay-alive capacitor hacking

Postby Roger (RJ) » Thu Apr 08, 2010 11:00 pm

Infinite in parallel with 10K = 10k

No Problemo

Brianetta
Posts: 190
Joined: Wed Feb 11, 2009 12:49 pm
Location: Jarrow
Contact:

Re: Stay-alive capacitor hacking

Postby Brianetta » Thu Apr 08, 2010 11:55 pm

Roger (RJ) wrote:Infinite in parallel with 10K = 10k

No Problemo


D'oh, of course. It is a leak; the higher the resistor the better, I suppose. I have some 10k resistors in the house, so I'll give this a go before giving up.

I saw a 4,400µF 35V capacitor today. It was about the size of a C cell. That's silly-huge. I also saw some 100µF ones, which were reasonably small, and I could maybe put as many as I can cram in in parallel... frankly, I'm not sure how much is needed to nudge a loco past a bad patch of track. I'm still trying to get over the fact that there appear to be 27 volts across this decoder's innards, instead of the more usual 12 or 14 that others have found.

Brianetta
Posts: 190
Joined: Wed Feb 11, 2009 12:49 pm
Location: Jarrow
Contact:

Re: Stay-alive capacitor hacking

Postby Brianetta » Fri Apr 09, 2010 12:06 am

S_Jay wrote:electro-frogs (avoids dead zones but tricker to wire, however your efforts above and with the signal lights this should be no problem to a man with your skills)

Sorry, I wasn't able to comment on this last time I replied, being as I was in a pub with a bunch of other goths trying to type on my mobile phone.

A man with my skills! Wow - I've been soldering things for about six months, and pretty much everything I've ever soldered is in a photo on this forum! I'm well chuffed. I'm a New Railway Modeller, and I'm having to learn pretty much all of the skills from scratch. My first train set, my first soldering iron, my first paint brushes, my first scalpel, my first diode, my first resistor, my first LED, my first multimeters. I haven't yet owned a capacitor that wasn't already part of something. This is all theory to me, and I'm putting it into practice for the first time as I go.

I tell you something: YouTube taught me to solder. I remember a bit about capacitors and resistors from school, but my diode knowledge is all courtesy of Google and (to a lesser extent) Wikipedia.

I believe that the only barrier to skill is lack of faith in one's own ability.

User avatar
Bigglesof266
Posts: 927
Joined: Tue Feb 10, 2009 9:59 am
Location: Australia

Re: Stay-alive capacitor hacking

Postby Bigglesof266 » Fri Apr 09, 2010 1:06 am

Brianetta wrote:I have a Hornby Mixed Goods set, which has been incorporated into my shelf layout, Hepton East. The locos are DCC fitted, but they're not exactly top of the range Hornby kit. The Class 08 and the Jinty share the same chassis (which is the same as Thomas the Tank Engine's chassis) and are prone to stall on or near my insulated points.

I have a lot of insulated points.

The locomotives are clean as a whistle, as is the track. It's just plain bad luck that they keep stalling, as the wheelbase neatly manages to equal the dead areas on adjacent points, and the very slightest unevenness causes an immediate stall unless you're doing a reasonable speed.

Reasonable speed, in the case of the 08, is well above prototypical speed. So, I had a look around this here internet to see what other people did about the problem.


Great thread Brianetta, about a most relevant problem I believe everyone who owns those locos and uses Hornby insulfrog points will have experienced. My experiences concur with yours, differing only in that I'd consider it 'bad design' rather than "bad luck". The abysmal slow running characteristics of both of those chassis in conjunction with all the other aspects which might affect operation notwithstanding dictates that even if everything else is in A1 order, the slightest torshional movement in the points as they are traversed or them not having been laid perfectly flat will result in the chassis stalling on them unless running at silly speeds.

My solution like yours was not to 'pay out' on insulfrog points, of which like you I have many both Peco and Hornby (new type), but to point the finger of blame directly to where it really lies in that chassis, gearing and motor combination, and do something about it. I replaced my DCC Mixed Goods 'Railroad' Class 08 with its rubbish Hornby decoder (no apologies, they simply are) with a detailed DCC Ready Hornby 08 in which I put a similiarly inexpensive (ESU) Bachmann 36-553 decoder. Chalk and cheese. The motor and gearing in the detailed Class 08 will run at prototypical speeds and the still short wheelbase chassis makes better contact so that the stalling issues have disappeared completely on accurately laid and properly maintained electro-clip fitted insulfrog points. None of my other locos (more than 20, but much longer wheelbased) suffer stalling on those points. So imperfect -in that they are a compromise- as insulfrog might be suggested to be -the realisim of Standard British Geometry excerpted, other problems which might occur form time to time for which they have become a the convenient 'whipping boy' are IME down to a combination of that geometry in conjunction with spurious other contributing factors which IME can generally be remedied by good preventative maintenance and setup such as ensuring carriage wheels are gauged and free running, pony truck/leading wheels spring tension's are set, plus, .....traversing them at prototypical speeds rather than hitting them like an express train attempting to break the London to Edinburgh record. ie: even my long Hornby 12 wheeler dining car with poxy plastic wheels (gauged) in a six coach rake hauled by Duchess "City of Bristol" has no problem consistantly traversing insulfrog points without derailing. But I digress.

I do like the idea of a decoder mod 'fix', but in the end it's still just a kludge for a fundamentally flawed design. Hence for my money, triage dicates the most inexpensive, KISS and ultimately satisfying result/solution to the same 'Railroad' chassis "Jinty" issue will be similar to my approach to the Class 08 problem. Sell it on FleaBay and replace it with a better designed unit, in this case Bachmann's "Jinty". In the end, as it did with the 08, I do think this will effect the result I'm seeking with "Jinty".

Regardless of our differing approach, I'll still follow your thread with interest. GL.

Brianetta
Posts: 190
Joined: Wed Feb 11, 2009 12:49 pm
Location: Jarrow
Contact:

Re: Stay-alive capacitor hacking

Postby Brianetta » Fri Apr 09, 2010 1:40 am

Thanks, GL. I'm confident that my track is well laid and well looked after. The Jinty actually does a lot better. I will, no doubt, purchase better motive stock in time, but it's out of my reach at the moment. Bits of wire and capacitors are cheap, and if I break the loco's decoder, I break it. I can buy a better one then, and put that into a better chassis when I can afford that. For now, though, this kludge is looking like the most fun!

User avatar
Bigglesof266
Posts: 927
Joined: Tue Feb 10, 2009 9:59 am
Location: Australia

Re: Stay-alive capacitor hacking

Postby Bigglesof266 » Fri Apr 09, 2010 2:28 am

By GL I meant "Good Luck (with your decoder mod project" . Just been around the net so long all those as standard as a smiley :) acronyms are so familiar deployed in common use one considers them universally understood rather than elitist pretension.

In case it came across differently, no aspersion was intended upon your track or track laying abilties by my mention of the issues which can arise derived from my own experience. I've no doubt you were thorough in eliminating track iregularities and in preventative maintenance, easily ascertained given the gist of your thread's initial post.

Completely comprehend your attraction to the 'joy' of the challenge presented by the 'mountain'. Merely pointing out an alternative I chose akin to flying over it instead. I'll remain interested in future developments.

Cheers.
Edit sp.


Return to “DCC Forum”

Who is online

Users browsing this forum: No registered users and 10 guests