DIY Stay Alive

[alangr]

I am interested in setting up a 'Stay Alive' DIY system and have looked extensively at what oythers have done. Many leads see to come back the this site http://www.members.optusnet.com.au/mainnorth/alive.htm which is very helpful. OK, I now have the bits (4700mF electrolytic condenser, IN4001 series diode and 100 ohm resistor) and the positive side goes to the blue wire of the decoder.

 

Now comes the query. How to identify the negative side of the (bridge) rectifier on the decoder to which to connect the other side of the condenser. I have not started yet but aim to put a meter across the blue lead and...somewhere that looks as tough it might be the appropriate place. The real query then is if I get a sensible reading, is it sensible to go the next step and connect the -ve side of the condenser to that point.

 

In other words, am I likely to get any spurious readings and attach a wire in a place that might damage the decoder?

 

OK, I am not asking for guarantees but can anyone give advice on a sensible way to proceed here?

 

Thanks!

[Nigelcliffe]

I'd approach it this way:

 

Decoder in some sort of test holder/harness, so it can be powered (from DCC) and has some outputs working (eg. LEDs + Resistors on motor outputs). Use a current limiting resistor (say 100ohm) in series with DCC supply, arranged so it can be easily by-passed for full power. This might give you some protection from shorts/overloads, then bypass the resistor to confirm readings.

 

Power on DCC, and with a fine-point voltmeter try to find the negative from the rectifier, the blue being the positive.

 

Worst case, the negative is only found at the rectifier in middle of decoder. But, you may find a track somewhere on the edge which is connected to the rectifier.

 

Spurious readings of the negative lead could come from function outputs, so watch for this.

 

No guarantees offered, this is "rip up all paperwork stuff".

 

 

- Nigel

[RedgateModels]

What decoder are you considering?

[Mark K]

First, are you sure that it will be a bridge rectifier and not four surface mounted diodes. I've unwrapped a few different decoders and found it is the latter and usually another diode nearby called a Schottky( don't worry about this one). So you're looking for a cluster of four perhaps five black blocks, they are quite large in comparison with some of the other components and will all have a white band on them and there will be two together with the white bands next to each other. Put your meter set to resistance across the two ends and it should show no resistance. The chances are that this is the -ve. THEN connected the decoder and test.

Surface mounted Bridge rectifiers are usually marked with the 'cycle' symbol on one side and - and + on the other. Look closely as these symbols are quite faint - might need a magnifier. The recifier will be bigger than SMD resistors and will have four terminals. Because of that they are usually easy to spot.

 

Mark K

[kevinlms]

First, are you sure that it will be a bridge rectifier and not four surface mounted diodes. I've unwrapped a few different decoders and found it is the latter and usually another diode nearby called a Schottky( don't worry about this one). So you're looking for a cluster of four perhaps five black blocks, they are quite large in comparison with some of the other components and will all have a white band on them and there will be two together with the white bands next to each other. Put your meter set to resistance across the two ends and it should show no resistance. The chances are that this is the -ve.

 

Mark K

 

Surely not. Two bands together would indicate positive.

 

All too easy to make mistakes when writing basic instructions, for something we know without thinking about it.

 

 

Kevin Martin

[alangr]

Thanks all for the replies/comments. The decoder (which is already installed) is a TCS 4X and used for motor power and lighting. It is installed in a tram which has two double bogies and despite all the wheels, continuity is not all that it could be. I have seen the image from the web link for a TCS T1 so I am hoping to find another bridge rectifier (perhaps something like the one in the image).

 

Nigelcliffe; thanks as this was the kind of approach I was thinking of but probably in situ rather than removed (but I will see).

 

Mark K; Yes, I am aware that a bridge can be built with four diodes so that is certainly a second option to be looking for - and noting the comment from kevinims.

 

I think it is time to head to the trainroom downstairs and get trying. Assuming all does not go up in smoke, I'll let you know how I get on.

 

Thanks again all

 

Alan

[alangr]

Well good news and bad. After almost dismantling the tram and re-wiring (It was early work and very messy) I checked that all worked as before and then stripped off the decoder shrinkwrap. I was quite put off by the size of the contacts or, rather, the lack of size!, but I did find a bridge rectiifier with + and - marked. Set up the meter and, yep, about 15V appeared with polarity as marked. Great!

 

But when I tried to solder to the rectifier arm - this was relatively large and at the edge of the pcb - the solder simply did not take. I scraped the surface and then used some flux but to no avail. Not wanting to heat the unit too much I thought I would see if anyone has an idea of the best way forwards. There is room to twist a wire around the leg of the rectifier and this may be my best bet - but I don't like it! The leg appears top be soldered to the pcb so am I simply using the 'wrong' solder (standard silver - i.e. non lead - solder)?

 

Any help appreciated, thanks.

[Nigelcliffe]

Choice of solder is important, as is soldering iron, choice of flux (what you'd use for sheet or plumbing might not be the best!), and technique. Probably needs practising on something scrap.

 

Iron needs to be very clean, small tip, adequate power, tip tinned with solder in use. I'd be using a 50W temp controlled iron with a tiny tip.

 

I'd use a lead-based solder cream/paste, possibly with a small trace of silver for flow. Use a very tiny amount, applied to the components/board. If possible, I'd quickly pre-tin this onto the board/leg, then apply more before fixing my pre-tinned leads.

 

I know that electronics work should use lead-free solders, but they run at higher temperatures and I find they are harder to achieve really good flow.

 

You may be able to trace around the board and find a track near the edge for soldering a wire to it, fix a fairly fine wire, then attach a larger wire to the fine wire.

 

 

- Nigel

[Mark K]

Surely not. Two bands together would indicate positive.

 

All too easy to make mistakes when writing basic instructions, for something we know without thinking about it.

 

 

Kevin Martin

 

Kevin, Quite right common mistake, the band on a diode is the cathode, ie the -ve. In a bridge rectifier made up of separate diodes, two bands together is the point where the +ve comes off. Same with a capacitor, the band is the negative.

 

Mark K

[Mark K]

I'm glad Kevin picked me up on that. Sorry.

Soldering Surface mounted devices. To comply with standard that board will have been soldered using lead free solder. The trouble is that the lead free stuff melts at a much higher temperature than the old fashioned 60-40, so some irons just don't get there. I've always used the lead version, it mixes with the lead free and on SMD's I use it with a 12 watt Antex. Nigel advice on plumbers solder was not strong enough. Never use in on electronics. You have to neutralise it with an alkaline solution or the whole thing will corrode. Just use an ordinary rosin aka resin cored lead solder, you can still get it easily. Just clean the joint up - a fiber glass brush is good for this - make sure you get rid of all the old residual flux because it cakes. Tin the joint and wire, have a wipe the tip of the iron on a damp cloth, apply a little solder to it for heat transfer and whammo! the thing should join. Good luck.

Mark K

[alangr]

Well, it works and it's just magic!

 

I did not manage to solder to the rectifier leg despite cleaning well with a small grindstone (used careully, of course!); it simply would not tin whatever I did. As I indicated before, the rectifier was quite large in comparison with everything else on the board so I bared the end of a fine wire, pushed it through the gap between the leg and the body (of the rectifier), twisted it and soldered that. A few tests showed that it was making contact OK so, finger's crossed, I started to put things back together. With just directional lighting (i.e. no internal lighting or motor) when picked the tram up from the track the lights died. No immediate cut-off but a definite lag; not for very long, much less than a second.

 

The tram was then assembled and then run with all lights on speed step 1. It did not stop at all although the lights did flicker a bit; not on/off, just on/dimmer.

 

So success? Oh yes but also... The main downside, which is not really a problem, is the fact that the lower floor of the tram just looks full of clutter - a 4700mF electrolytic may look quite small on the bench but in a small space...it's huge! But some false window images inside will help that a bit. Also when putting the tram on the track when the track is powered, some lights seem to flash but that is only momentary.

 

All in all a good exercise. How such a thing would be possible with smaller locos, goodness knows. A smaller condenser might help and still give some reduction in stuttering but, as everyone says, clean track is still the best way out of things.

 

So thanks all and to confirm Nigel I did not even attempt to use plumber's solder/flux; and Mark, I ended up using my 25W Antex with an angled bit - I do have a point bit but have great problems getting it to wet properly, despite cleaning. Anyway, by the time I was using a wire as above, there was plenty of room to solder 'normally'

 

Cheers

 

Alan

[RedgateModels]

I was recently asked to see if I could fit a Stay Alive to a Bachmann 36-553 3 function decoder, so spent a pleasant half hour with a decoder and a multimeter on my bench.

 

The results were very good, the attachment point for the negative end of the capacitor is arrowed. You could solder to the negative end of either of the two diodes but the end of the component as shown is easier. I used a fine bit and my Optivisor to see what I was doing!

 

 

With a 4700uF capacitor the loco ran for an extra 1/2 a wheel turn before stopping on the rollers

[Wogga]

I have had great success with ESU Lokpilot 4 decoders which have purpose connection points for a home made power pack. They recommend a 25v 2200u/f cap, a diode and 100 ohmn resistor. The difficulty is stashing the components. Here are two videos of a Fairburn fitted with the power pack in the coal bunker, the Jinty I managed to squeeze a 16v cap on the cab floor. The question mark at the moment is the 16v too close to the operating voltage of 14.5v.

 

 

 

Both are crawling over a difficult double crossover at speed step 2. before the power pack fit they would only cross at speed step 10 as a minimum..

 

 

 

 

[RedgateModels]

remember the volt drop across the diodes in the bridge rectifier will lower the voltage seen by the capacitor. I use 16V caps all the time as my Powercab is only fed by a 15v DC psu

[Wogga]

Cheers for that, so far so good it was just with the cap tolerance being +/- 20%. I am probably looking at it too deeply.

[Nigelcliffe]

I have had great success with ESU Lokpilot 4 decoders which have purpose connection points for a home made power pack. They recommend a 25v 2200u/f cap, a diode and 100 ohmn resistor. The difficulty is stashing the components. Here are two videos of a Fairburn fitted with the power pack in the coal bunker, the Jinty I managed to squeeze a 16v cap on the cab floor. The question mark at the moment is the 16v too close to the operating voltage of 14.5v.

Solution to lack of space, change the capacitor technology. I've been using 16v ceramic capacitors which are 3.2 mm x 2.5 mm 2.5 mm for 100uF.  Build this into rectangular blocks for the available space and required storage charge.   I've put 900uF into a 2mm scale shunter which is 45mm long over the buffers (pictures/details in blog mentioned in signature block), 900uF into a 2mm scale 14xx tank engine, and 2200uF under the floor of a Y6 Tram (Toby the Tram) in 4mm scale.    The ceramics are not cheap, unless bought in large volume they are over £1 each from UK suppliers, slightly cheaper from Europe.  An alternative to ceramics are tantalum capacitors, available in larger capacitance per block, but I've not found them to be quite as efficient on space.   Surface mount resistors and diodes are also small, so help with space.

 

One solution to marginal track voltages on 16v rated components is to install a 16v Zener diode in parallel to the capacitors.  If the input voltage exceeds 16v, the Zener will conduct the excess voltage away.  Because of the charging resistor/diode, that excess voltage will be current limited through the resistor, so the current in the Zener will be sensible. 

 

- Nigel

[Wogga]

Top stuff cheers Nigel.

[ZiderHead]

I know that electronics work should use lead-free solders

 

OK lets nail this once and for all

 

EU law bans the use of lead in electronic devices manufactured or imported into the EU, with exceptions (and that list is very long). So if you are manufacturing and selling products which aren't exempt you should use lead-free.

 

Otherwise you can use leaded solder (and IMHO you should, just to make your life easier).

 

Lead does not absorb easily through unbroken skin, but it does through mucous membranes - so definitely no eating or smoking while you are handling it, and wash your hands thoroughly afterwards. Use disposable latex gloves if you're really concerned.

 

By far the biggest health threat from soldering is from vaporised flux (eg. rosin is very nasty stuff indeed) so have lots of ventilation. I use a small desk fan the other side of the work to pull the air away from me while I'm working as well as having all the windows open.

 

[Crosland]

OK lets nail this once and for all

Yes, lets.

 

I am involved in manufacturing for sale but also use nothing but lead free for (electronics) hobby work to avoid cross-contamination.

 

Lead-free really is no harder to use than leaded in a hobbyist environment.

 

Andrew

[ZiderHead]

I am involved in manufacturing for sale but also use nothing but lead free for (electronics) hobby work to avoid cross-contamination.

 

That sounds like a sensible precaution, and a sensible idea as you only have to learn the properties of one type.

 

I find leaded much easier to use for a bunch of reasons and I'm sure many others do too, and I just wanted to dispel the rather common myth that leaded solder is banned for sale or use by hobbyists for their own use.

 

Some info from Hakko about lead-free: http://www.hakko.com/english/lead_free/pages/ I'll add that the joints are more brittle, it doesn't flow as well and the flux usually stinks!

[RedgateModels]

Quick update on the Stay Alive on a Bachmann 36-553 - remember to clip the capacitors on the motor if you get erratic running - I had to do this on the Bachmann K3