Jump to content
 

NCE SB5 with AIU and BD20 for block occupancy - need resistor wheelsets


imt
 Share

Recommended Posts

  • RMweb Gold

I expect this is an oft asked and answered question, but I cannot find the answers using Google or RMWeb search - so here goes.

 

I have a layout controlled by JMRI and using NCE products for power supply and other electronics.  I want to move on to Block Occupancy testing so I have acquired the necessary additional hardware - an NCE Auxilliary Input Unit and some BD20 occupancy sensors.  I am at the stage of testing before upending my layout to perform surgery undernesth to add the detectors where necessary.  I have used my stand alone loco programming track to loop in a BD20 on one of the supply wires and to see how things go.  Now a DCC loco is detected and an LED shows on the AIU.  It's a Bachmann 08 and it draws about 0.1 amps when just sitting there.  So that's OK: locos on the layout will be "seen" and the software can react to that.  But of course I want the whole train, or the odd rake of coaches left in a platform, or the whole length of a DMU to register appropriately.  This means resistor/resistive wheelsets and there the problems begin.

 

I know nothing of electronics. I am quite good at other things but sadly not electonics, brain surgery or rocket science.  I have found some YouTube episodes describing how to use SMDs for this purpose.  I bought some but they were so small that I couldn't get on with them - damaged eyesight and arthritiky fingers.  As you will all know, magnifiers are all very well but they insist on magnifying the tweezers too!

 

So I bought some "through the hole" VERO type resistors (10k ohm) which I can see and have nice pieces of wire that I can solder to things. My intention is to CA the resistor to the axle and then solder the wires to the back of the wheels above the insulating bushes. In my test before doing anything permanent  mode I thought, drop one of those accross the rails and see if it's detected.  But it wasn't. Uh? If I put 3 or so on they were. So I thought must have bought the wrong size - multimeter says 10k oms.  Maybe its insufficient contact area. Try holding resistor accross wheel set.  Still doesn't register.  Clean "blackening" off the wheelset - resistor still doesn't register - but I can read 18 volts AC across both of them.  Power OK.

 

What am I doing wrong?  What is it I don't know/understand?  I don't want to waste hours making up wheelsets that won't work.  HELP ME PLEASE - I DON'T CARE HOW MUCH YOU LAUGH AT ME.

 

 

Link to post
Share on other sites

It depends on what the trigger current is for your detectors. As an example a DR4088 will trigger at 1mA which is a resistor of ~15k at 15v track voltage. a 10k would trip a DR4088 nicely. A DR5088RC needs around 10mA to trip and show occupancy which therefore means a resistor of 1k5 at the same track voltage.

 

If you have several resistors of the same value then simply put them in parallel until the occupancy triggers and that will show you what the trip current is

 

10k+10k=5k

10k+10k+10k=3.3k

10k+10k+10k+10k=2.5k

etc...

 

EDIT - once you have worked out the value then buy resistors of the right value like these

 

https://www.ebay.co.uk/itm/SMD-SMT-0603-Chip-Resistor-0R-1M-Ohm-Range/131760464157?hash=item1ead89211d:g:ipwAAOSwQjNW8qkm

 

which you superglue to the axle at the insulated end and use conductive paint 

 

https://www.ebay.co.uk/itm/Electric-Glue-Wire-Silver-Conductive-Needle-0-3ML-1-0ML-Paste-Paint-PCB-Repair/333582960676?hash=item4dab180824:g:YZIAAOSwMLNeoqa~

 

to paint onto the axle and wheel flange from the conductive pads at the end of the resistor

 

works a treat for me :)

Edited by WIMorrison
Link to post
Share on other sites

Reading the BD20 manual....  Three things strike me about it. 

 

1 - sensitivity depends on number of turns through the coil that the track feed wire makes.   So, if not sensitive enough, put in another turn, up to the suggested limits in the manual.  

 

2 - there is a table of current needed to trip the device,  1 turn is 8mA, 2 turns 4mA, 3 turns 2.5mA.   At a nominal track voltage of 18volts, we can apply Ohms law to get approximate resistor sizes.  Resistance = Volts / Amps.   At 1 turn, 18/0.008=  2,250ohm,  at 2 turns = 4,500ohms, at 3 turns =  7,200 ohms.     

 

3 - the manual talks of using 4700ohm resistors, or doubling (in parallel) 10kOhm resistors to give 5kOhm in total.   Those broadly align with your "needs three 10kOhm wheelsets", as that would be 3,300ohms.    If you have two turns through the coil, your result is within specification.  If you only have one turn, then its detecting better than specification. 

 

 

I'd do it as Iain suggests, work out the size needed, use SMD glued to axle, and connect with conductive paint.    Depending on two or three turns, aiming for either 3k3 or around 5kOhm resistance.  

Only difference might be on the size of resistor, the 0603's Iain suggests are fairly small,  0805 or 1206 are physically larger sizes and may be easier to handle.

 

Ceramic tipped tweezers might also be useful, accurate if adjusted well, and don't stick to superglue. 

 

 

- Nigel

 

Link to post
Share on other sites

  • RMweb Gold

So as I read the kind replies, what you are saying is that I am using TOO BIG a resitor at the moment and that's why a single one doesn't work - adding resistors actually LOWER the overall resistance.  Hence using Nigel's simple sums (thanks so much for the simple I need it) I need to buy 4.7k resistors since I have (and intend to keep) 2 turns.  As usual RTFM, but I didn't really understand it when I did!

 

I know you mean well when you keep reccomending SMDs - I really cannot get on with them so I will get some 1206 to try and also order the smallest 4.7k "through the hole" I can get and see which I manage with best.  They are as cheap as chip - ugh!

Link to post
Share on other sites

Yes,  more resistance (bigger value) means less current flows.  The detector needs enough current to "see" something is present.

 

But, I don't think your evidence shows 4.7K will work,  both your experiments, and the manual says you need a lower value. 

 

At two turns, from the BD20 manual, that needs a current of 4mA, which the sums come out at 4500 ohms.  
That is your maximum resistance at 18v at the track.   4700ohms is more than this, so not likely to trigger reliably.  

Just to be safe on track voltage, re-do the sums at 15v, and you get 15/0.004 = 3750 ohms.   

So, to have a bit of margin on track voltage, 3500 ohm might be a best "highest value resistance which will work reliably".  

Standard (and therefore cheap) resistors come in 3,300 ohm, so that's the size to purchase.   

 

Resistors in parallel add up with reciprocals.  Thus  1/R(total) = 1/R1 + 1/R2 +1/R3, etc.. for all the resistors in parallel.   So, three 10K in parallel give you 1/10,000 + 1/10,000 + 1/10,000 = 3/10,000,   turn the answer on its head, and R(total) = 10,000/3 = 3333 ohms. 

( Because the maths drops out this way, if all the resistors in parallel are the same value, the total is R/N  where R is the value of one resistor, and N is the number of resistors, so 10,000/3 gives the answer.   But that only works if all the resistors have the same value.   )

 

 

Small through hole, would be 0.125W.   If we assume 6mA (ie. more than we're expecting to flow) and 18V at the track, then the Watts are 18*0.006 = 0.108W, which is less than 0.125W, so you're OK against the resistor getting hot.    Slightly larger 0.25W resistors would have loads of power (heat) headroom.  

Usually metal film are smaller than carbon film, but these seem pretty small, and carbon is usually cheaper.    Example 0.125W carbon film resistor :

https://uk.farnell.com/multicomp/mcre000043/res-3k3-5-125mw-axial-carbon-film/dp/1700243

 

 

 

 

 

 

 

Link to post
Share on other sites

  • RMweb Gold

Never was much good at physics - but I managed to fly alright without that knowledge.  I have ordered some of those too - so I should have it bracketed.  I'll let you know what transpires.

 

Thanks for all your hlep Nigel - as usual.

Link to post
Share on other sites

  • RMweb Gold

Well thanks Iain and Nigel.  But I have been thinking (again?) and would like a further response.  As I told you I was putting 3 * 10k across the piece of track I was using to test the whole BOD idea, and that was resulting in an "occupied" light from the AIU.  You explained that meant that I needed 10000/3 (special case as all 10k) or 3.3k.  Then I had this horrible thought in the middle of the night!.  If I put 2 * 3.3 on the first and last wheelsets on a coach what would happen - would this mean that the coach would not be recogised because 2*3.3/2 is less than I need?  So I did an experiment this morning with 3 10k close together and they registered occupied as you would expect.  I then slid one half an inch away from the other 2 - still occupied.  But at about an inch not occupied any more.

 

So it seems to me in my ignorant state that a 3.3k ohm resistor at either end of the coach will be ok - but will it?  They will both (and maybe including a loco) be in 1 BOD block.

 

How about short UK waggons when in a rake?

 

Am I just panicking unnecessarily or are there more gotchas to trip over??

Link to post
Share on other sites

You're over-thinking, or thinking backwards.

 

Resistors restrict/resist current flow.  The bigger the value, the more they resist current flow.    The detector (BD20) has a minimum current (4mA) and a maximum current (10A in the two-turns of wire arrangement).

 

As you've seen, putting three 10K resistors in *parallel* on one axle reduces the resistance to 10,000/3 =  3333ohms. 

 

If you put two wheelsets, with 3.3K resistors on each on the track, you have two *parallel* resistors of 3.3K,  so, 3.3K/2 = 1.65K.    That is a lower resistance, so more current flows, which is what is wanted for the BD20. 

 

 

 

Is there another "gotcha",  well, yes, there are two, and they are related:  

 

A  -   too many resistor wheel sets in parallel, and you end up with all your system current through the resistors, so nothing left to run the trains.  
My very rough estimate came to 200 axles with 3.3K resistors uses 1A.   Or,  1000 wheelsets with 3.3K resistors uses 5A.  If you put resistors on every axle of stock, you could hit those numbers on a moderate sized layout.      

In practise, the "sensible" is to put resistors on the vehicles which make up the rear of trains.  So, some coaches, brake vans, and if modern stock, wagons carrying end-of-train lights.  
  

 

B -  the BD20 will have a maximum recommended current flowing through the coil.  Consulting the BD20 manual, and the maximum with two turns of wire through the coil is 10A.   So, provided your system is smaller than 10A, the coil cannot be overloaded.    So, we can discard that for most situations, as we'll not get to 10A from most systems, and if you have power district breakers set at a lower trip current, those are the current limit. 

 

 

As to why your resistors failed when moved apart, probably dodgy contact.    That's something model railways have all the time.   If they made good contact to the rails, then detection would be fine.  

 

 

Link to post
Share on other sites

  • RMweb Gold

OK I'll stop panicking Nigel!  Thanks.  I THINK I understand all that.  The 3.3ks should arrive tomorrow so I can start testing from there.

 

I intend to have a reistor set at each end of the two carriage subsets which make up my rakes.  My maximum length is 5 carriages (6 a real push) and the extra coach would tend to be a an RU or similar in the middle or a BG for parcels on the end.  These extras would probably have a resistor set at each end.  The maximum my fiddle yard and station can hold would be less than 2 dozen "carriages" and that would mean the same number of resistor wheelsets (given that some are 121s (need none) or 117s (need 1 at end) and most are in 2 carriage subsets with 1 resistor each end).  So I think I am in the .25amp load area - or 2 sound locos idling!  I'll think through how to manage the freight later.  My SB5 is 5 amps so I won't hit the maximum, and my 3 power districts and accessory bus have 1.75amp limiters. 

 

Thanks very much, I have learned a lot, and I hope others will find this useful too.  I'll  let you know how I get on.

Link to post
Share on other sites

One more thing - if you have an actual lamp at the end of the train, or perhaps coach lighting then the wagon/coach with the light fitted will not need a resistive axle because the lights should draw enough current to trigger the detection.

 

If you have decoders fitted in coaches or wagons to operate couplings, doors, lights, etc then the same applies - resistive axles are only needed for 'dead' stock which is becoming increasingly rare these days :)

Link to post
Share on other sites

  • RMweb Gold

Thanks Iain.  I hadn't thought of that.  At the moment I don't have lights or additional decoders, though I do keep thinking about it - especially for my 101 and 108. Presently the BOD process is taking up my time and pocket money!  

Link to post
Share on other sites

  • RMweb Gold

 

19 hours ago, WIMorrison said:

Look at using something cheaper then

 

You are quite right Iain, and had I not started travelling down the NCE route 7 years ago I wouldn't be using their gear now (the old "How do I get to Tipperary?  Well I wouldn't start from here" joke) I have quite an investment now and I didn't feel like trying mix and match.  I decide that I'd finish this layout as I started - and at 76 there may never be another.  The problem with NCE is it's all old designs and no development - but at least it does what it says on the tin.  IF I were starting again I would probably go with Digikeijs - European, broadening range and well documented.  But others would no doubt differ.

 

Iain

Link to post
Share on other sites

  • RMweb Gold

Well that was fascinating.  Something new I almost got right, and then it went wrong (again!).

 

The 4.7k ohm through the hole and 1206 smd packages arrived to day, so this afternoon, with much trepidation, I set out to have a go.  Firstly the 4.7k through the hole resistors DO actually register on the NCE AIU, so (even though I intend to use the 3.3k ohms when they arrive, as my mentors Iain and Nigel have advised) I thought I'd see whether I could make up a resistor set. I scraped the paint off the axle and at each end above the insulating bush. My attempts with CA on the through the hole type was not encouraging. they simply would not stick, the CA to metal was OK but not to the resistor.  BUT the 1206 size SMDs looked doable, I could see them and I could pick them up with my "open the wrong way" tweezers. And I managed to stick one at an angle across the insulating bush so it touched the wheel and the axle.  I then used my "Bare Conductive" paint along the axle to the other side and up to the opposite wheel, and then from the top of the SMD to the wheel it was touching.

 

So far so good.  Gosh I may be able to do this! Then I tested the resistance.  It was over 10k but falling (?) and when it got down to about 4 I put it on the track and hey presto the AIU "saw" it.  Go for celebratory cuppa and newspaper read. on return found that the wheelset was very hot - it would have melted a plastic bogie.  Now the resistance is negative - if that's possible.

 

I was generous with the paint.  Is that what I did wrong or is there something else?  Or maybe many something elses!

 

Feeling happy that I can do this, just need guidance to get it right.

IMG_0455.JPG

Edited by imt
to add a photo
Link to post
Share on other sites

2 hours ago, imt said:

......….the 1206 size SMDs looked doable, I could see them and I could pick them up with my "open the wrong way" tweezers. And I managed to stick one at an angle across the insulating bush so it touched the wheel and the axle.  I then used my "Bare Conductive" paint along the axle to the other side and up to the opposite wheel, and then from the top of the SMD to the wheel it was touching.

 

So far so good.  Gosh I may be able to do this! Then I tested the resistance.  It was over 10k but falling (?) and when it got down to about 4 I put it on the track and hey presto the AIU "saw" it.  Go for celebratory cuppa and newspaper read. on return found that the wheelset was very hot - it would have melted a plastic bogie.  Now the resistance is negative - if that's possible.

 

I was generous with the paint.  Is that what I did wrong or is there something else?  Or maybe many something elses!

 

 

I'd guess you have paint from both rims to the axle.  So, you have a short through the wheelset.   It won't be zero resistance, but very low.  (If it were zero, your DCC system overload would have tripped. 

 

The paint may be under the resistor, shorting it out there.  

 

- Nigel

Link to post
Share on other sites

  • RMweb Gold
20 minutes ago, Nigelcliffe said:

The paint may be under the resistor, shorting it out there.  

 

So it's not necessarily too much paint, but probably paint in the wrong place Nigel?  I was trying to put paint onto the silver pads but not below the SMD.  Perhaps I moved too quickly to do it, I tried to let the paint "dry out" a little before I used it so it didn't flow too much.  Presumably make sure I leave it longer next time.

 

I have just looked at the ampmeter - 0.3 amps as soon as I put the wheelset on the track and it heats quickly.  So this is a short?

Link to post
Share on other sites

0.3A is, effectively a short.    0.3A at 18V means a resistance of 60ohms.   Or power (heat) to distribute of about 18*0.3 = 5.4W.  That's quite a lot of power (heat) to get rid of in a wheel set.   (eg. put an old style car side light bulb on a 12v DC controller, and feel how warm it gets). 

 

I can't really advise on how to use conductive paint, because I don't use it.   I'd be able to solder the components to wires. 

 

 

Back at the through-hole resistors.   If you coil one of the leads a couple of turns tightly around an axle, it won't fall off.  Then bend the end of the coiled lead out towards the rim.  Either glue to back of wheel, then conductive paint the connection to rim, or tiniest dab of solder to rim (depends on how good you are at precision and quick soldering). 

Arrange resistor so the other wire does not touch the axle, and bend it so it goes towards the rim, and connect that end to rim. 

To stop resistor moving, small dab of araldite, or superglue, should hold the coiled wire in place. 

 

 

 

 

Link to post
Share on other sites

  • RMweb Gold

I have had a message from the resistor supplier apologising for the delay and saying the 3.3k in two styles will be with me by noon tomorrow.  I'll start with your through the hole idea first because it seems to me to be the less open to my clumsiness.  I'll have a second go at the SMD approach too but that seems more chancy.  I have just had a clumsy quick swing at the task as you suggest using a 4.7k through the hole and it worked quite easily.  its just held in place at the moment but it does the job with no apparent current draw.  I'll keep you informed.

Link to post
Share on other sites

I stich the smd to the axle and the following day I 'paint' on the conductive paint with a cocktail stick and within 24hrs the paint has dried and the wheelset is working correctly - I am working in H0e which means N-Gauge wheelsets and it is fairly easy, would be a doddle in 00 wheelsets :)

Link to post
Share on other sites

  • RMweb Gold

Nigel:

 

Well true to their word, the courier arrived at 11.05 am and this is my first VERY CRUDE attempt with 3.3k ohm through the hole resistors using your suggested method.  Held in place with a crocodile clip whilst one end wrapped round the axle and then stuck to the wheel. I used Bondic, instant curing under blue light.  The other end was cut back and stuck to the wheel with NO contact with the axle.  To keep things secure a drop of fast acting CA was put on the resistor package where it was close to/touched the axle.  I then used Bare Conductive paint and the success of that after maybe 10 mins is in the second photo.  The AIU is top left and the small red smd led can just be seen.  No connection to the products named - they just do what it says on the tin.

 

Now, the result is a bit of a mess, and I will get better, it's crude but it works.  I have a spare pack of metal wheels which I can practise upon. There are some problems in that the led flickers with movement - improved by cleaning the track but I obviously need to get that paint/blackening off the wheel.

 

Iain:

 

Thank you for your continued encouragement.  I understand your method and I have seen YouTube videos showing it.  My problems are physical, knobbly finger joints and damaged sight.  Now I am sure that there are many with worse problems than I and who do much better.  I agree that those who can would find your method easier.  The second part of the delivery is 1206 size SMDs and I WILL try again.

 

Thank you both for your help and encouragement.IMG_0458.JPG.df8cc0b307a729c550ec9ba0a91daf30.JPG

IMG_0457.JPG

Link to post
Share on other sites

Good to see a method is working.   As you said, do a few more and it will get neater.   And local tweaks/improvements will probably appear along the way. 

 

The blackening on the wheels is likely to be chemical, an oxide surface to the metal.  Fine abrasive is the way forward, either fine wet&dry (I'd use that), or a fibre brush (horrible things with glass fibre splinters everywhere, last resort if nothing else will do!) 

 

 

Link to post
Share on other sites

  • 1 year later...
  • RMweb Premium

I too am about to start using AIUs with my NCE system. Quick question that this group might be able to answer. What is the maximun number of AIUs I can have? I have a Power Pro with a Procab and a Powercab attached.

I think the limit is goverend by adress range. My reading is I can attach AIUs certainly to addresses 8, 9 and 10, but possibly also to 4 and 5 since I have only two cabs connected, giving a maximum of 5 AIUs in my case. Can anyone confirm if this is so?

regards

Andy

Link to post
Share on other sites

  • RMweb Gold

Hi! I am currently far away and lying in the sun (sorry!) so I don’t have things to hand. My setup is similar to yours and uses Cabs 4 & 5 for the AIUs for definite. I know you can go to 8 or so but not sure whether it’s 8, 9 and 10.

Link to post
Share on other sites

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 account

Sign in

Already have an account? Sign in here.

Sign In Now
 Share

×
×
  • Create New...