Diode Matrix - Frustrating Problem

[drmditch]

Thank to everybody who has contributed.

 

Yes - the diagram is slightly obsolete. E and F were originally laid as a three way point. But I have had difficulty with remote operation of one of these in the past, so replaced it with two simple turnouts.

 

My conclusion is that if I wish to use the matrix to switch all four roads on 'one button', then it will have to be more equidistant between the two ends of the loop. (Current cable lengths as run are about 8' for D E and F.)

 

This gives rise to a whole issue. This railway (see my construction post .... here ...., is intended as my great work in retirement, and is quite large - the room is about 22' x 11'.

I had intended to have two control positions, but now realise I need to plan and execute these much better.

The traction power (12vDC) uses three ring-mains constructed of heavy duty cable with 'layout-wire' droppers. This (and the necessary switching) all seems to work correctly.

 

Does anybody have any experience of large layouts and how far the 'multi-strand 'layout wire' (7x0.2mm), can be expected to transmit sufficient power?

Also - I have a Maplin's Multi-meter (shown in the first picture above) can I use this to measure voltage drop?

 

I'm happy with the logic of the matrix (after all) but I clearly need to work out power levels and distances!

 

(I may start a new thread for this, as it goes rather beyond the original problem.)

 

So - again with thanks - have some (virtual) cake!

 

[RobjUK]

I've just found an advert for Peco point motors that gives the current requirements - Two amps.

 

7/02 wire has a resistance of 93 milliohms per metre, almost a tenth of an ohm.

Double that when calculating for a two-wire loop; call it 0.2 Ohms in round figures.

 

Based on that, each metre of the loop from source to destination would lose almost 0.4V at 2A

With four point motors in parallel, that become 1.6V per metre lost.

 

Personally, I'd not use 7/02 for any power handling circuit of more than a few inches and at low current; I'd suggest something quite a bit bigger for the points, possible even 1mm^2 flex for the multi-motor cable run.

 

[Donw]

It is not commonly recognised that the wiring used for point motors needs to be heavier than that used for traction power (unless you are using some heavy current 0 gauge models that draw 2 or 3 amps.) Recovered old mains wire could be used. 

 

Don

[steve W]

Reinvigorated with some lovely virtual cake, particularlry liked the lemon one , I have been thinking about this a bit more. The club 00 layout always use 24/02 for runs of three /four feet from the CDU, so in your case of seven feet, my thoughts are that it should be up in that region of size at least. 

The thought that then occured to me is that as you are using an electric pencil/stud contact method of operation you also have to consider complete  distance of CDU output -> control panel -> the diode matrix ->  and then to the point motor and back to the CDU in your length of circuit. Was this included in the 7ft length? 

If it was not then I would seriously consider using 12V DC on the stud contact to operate relays, one per route, the relay contact connects the CDU to the diode matrix  and point motors. The CDU and diodes being located as close to the point motors as possible.  This would be another thread, but happy to advise if neccessary.

cheers

Steve W

 

 

[DCB]

I always use recovered car wiring loom cables much of it recovered from 1970s cars and vans in the 1980s  which I think is 16/0.2 copper rated at 11 amps / 12volts (and 11 amps at 24 volts) 

The layout wire advertised seems to be 16/0.2 which seems to be rated at 3 amp  at 1000V AC

SteveW quotes they use 24/0.2 whoch seems to be rated at 4.5 amps at 1000Volts.

 

I don't understand why the car wire is 11 Amp when the layout wire is 3 Amp from the same size cores unless its just the voltage its quoted for. Pretty bizarre when both are for 12/24 volts.  However mine works with cable runs up to around 8 feet, and its common return so up to 5 points are returning through one 16/0.2.   

I would suggest trying some 11 Amp or more automotive wire sold for cars which has to cope with substantial loads  on a regular basis rather than anything sold for model railway use.  In normal times Car scrapyards sell great chunks of wiring all neatly striped and colour coded for a couple of quid, Vans are the best source, but with the Covid 19 around and the Country in Lock down (Except London which is still in denial )  this option is not available. 

BTW I just realised, Bojo said the country is in lock down, The Cockneys think he means the Countryside is in lock down not the Cities so they can continue to imitate sardines on buses and the tube.

[AndyID]

13 hours ago, drmditch said:

Does anybody have any experience of large layouts and how far the 'multi-strand 'layout wire' (7x0.2mm), can be expected to transmit sufficient power?

Also - I have a Maplin's Multi-meter (shown in the first picture above) can I use this to measure voltage drop?

 

 

You certainly can use your meter to measure voltage drop and you can extend the meter's leads if you are measuring it over a long distance. But the current has to be continuous like the current supplied to a locomotive. To measure the voltage drop associated with a transient current, like a solenoid point motor, you'd need an oscilloscope and it's not all that simple to get an accurate measurement even if you have one.

 

Based on the following table the resistance of 7/0.2mm is about 0.076 ohms per meter which means if it's passing one amp it will drop about 76 millivolts per meter or about 3/4 volts over ten meters.

 

https://www.canford.co.uk/TechZone/Article/MetricAWGWireSizeEquivalents

 

Solenoid point motors have little resistance and do require substantial currents (for a short amount of time) and a couple of volts lost along the wire could possibly make quite a difference.

[AndyID]

2 hours ago, DavidCBroad said:

I always use recovered car wiring loom cables much of it recovered from 1970s cars and vans in the 1980s  which I think is 16/0.2 copper

 

That's because it has a much greater CSA than 16/0.2  Measure the diameter of each of the strands and I think you'll find they are a lot bigger than 0.2 mm. Either that or there a lot more than 16 strands.

[DCB]

3 minutes ago, AndyID said:

 

That's because it has a much greater CSA than 16/0.2  Measure the diameter of each of the strands and I think you'll find they are a lot bigger than 0.2 mm. Either that or there a lot more than 16 strands.

That's what I thought but the specs say 16/0,2 for both types yet the amps are very different. Bit of a mystery.  Mine may be 24/0.2 but I can't check due to Lock down.

[St Enodoc]

15 hours ago, drmditch said:

Thank to everybody who has contributed.

 

Yes - the diagram is slightly obsolete. E and F were originally laid as a three way point. But I have had difficulty with remote operation of one of these in the past, so replaced it with two simple turnouts.

 

My conclusion is that if I wish to use the matrix to switch all four roads on 'one button', then it will have to be more equidistant between the two ends of the loop. (Current cable lengths as run are about 8' for D E and F.)

 

This gives rise to a whole issue. This railway (see my construction post .... here ...., is intended as my great work in retirement, and is quite large - the room is about 22' x 11'.

I had intended to have two control positions, but now realise I need to plan and execute these much better.

The traction power (12vDC) uses three ring-mains constructed of heavy duty cable with 'layout-wire' droppers. This (and the necessary switching) all seems to work correctly.

 

Does anybody have any experience of large layouts and how far the 'multi-strand 'layout wire' (7x0.2mm), can be expected to transmit sufficient power?

Also - I have a Maplin's Multi-meter (shown in the first picture above) can I use this to measure voltage drop?

 

I'm happy with the logic of the matrix (after all) but I clearly need to work out power levels and distances!

 

(I may start a new thread for this, as it goes rather beyond the original problem.)

 

So - again with thanks - have some (virtual) cake!

 

As others have said, given the size of your layout I wouldn't use 7/0.2. 16/0.2 minimum or even 24/0.2 would be my suggestion.

[cliff park]

The current rating of a cable has nothing  to do with voltage. It is determined by the size of the copper mainly. The second consideration is how hot it gets at that rated current, and how that heat is lost. This actually means wires under layouts can run hotter than those in conduit say, because the heat will easily dissipate. The voltage rating of cable is determined by its insulation. Think in terms of higher voltages require thicker insulation. But a cable rated at 10 amps will carry 10 amps at any voltage.

[AndyID]

15 minutes ago, cliff park said:

 This actually means wires under layouts can run hotter than those in conduit say,

 

Which is probably a very bad idea. If you can actually detect a significant rise in temperature in a wire under a layout that is really not a good thing.

[Donw]

5 hours ago, cliff park said:

The current rating of a cable has nothing  to do with voltage. It is determined by the size of the copper mainly. The second consideration is how hot it gets at that rated current, and how that heat is lost. This actually means wires under layouts can run hotter than those in conduit say, because the heat will easily dissipate. The voltage rating of cable is determined by its insulation. Think in terms of higher voltages require thicker insulation. But a cable rated at 10 amps will carry 10 amps at any voltage.

 

For electricians cables are rated at appropriate MAXIMUM current under various conditions free air, in truncking, burried in insulation. It is the heating effect of the current for a specified conductor and the capability of the cable insulation to withstand the temperature balanced against the heat dissipation  in its location that determineHowever the tables for which cable to use for a particular circuit the maximum cable length and hence the voltage drop are taken in to  account. For long lengths a larger cable will be specified to avoid excessive voltage drop.

 Solenoid coils as used in Peco point motors only take a transitary current so the heating effect is not the problem it is the voltage drop that is the problem any resistance in the wire suplying the coil will reduce the current at a given voltage.

 

For coils, motor windings etc it is the number of ampere turns that make the difference to the force generated. I spent part of my working life ensuring that the temperature of the windings did not exceed the limit for the type of insulation. Using a finer wire would generally allow more turns so a lower current would generate the same force and keep the temperature down. Unfortunately it also tends to increase the cost so commercial coils tend to need high currents.

 

If you have access to any recovered house wiring can be useful to layout use 1mm and 1.5mm lighting circuit wire will do the job for us.

 

Don

 

[kevinlms]

33 minutes ago, Donw said:

 

 

 Solenoid coils as used in Peco point motors only take a transitary current so the heating effect is not the problem it is the voltage drop that is the problem any resistance in the wire suplying the coil will reduce the current at a given voltage.

 

Don

 

The transitory current/heating effect is VERY important in the likes of solenoid point motors. Too much current for too long WILL cook the coils in a short space of time.

It is why CDU's are recommended.

[steve W]

Yes they will cook in a very short time, but a much longer time than that neccesary to operate.  With regards to your quoted paragraph, Don's words were in respect of the wire size not the solenoid coil itself  and are entirely correct.

"It is why CDU's are recommended"  Who reccomends? The primary purpose of the CDU is to store and make available more energy than that available immediately from a power source. That enables one or more solenoids to operate, but no better than a very large transformer would. The CDU automatically limits the supply by virtue of it's discharge and is a bonus that protects coils from jammed switch contacts on cheap and cheerful 'passing contact' switches. It also reduces sparking from switches (making them less likey to fail in the first place) for the same reason. The use of a CDU is of course a good thing, but for many different reasons. An adequate wire size though is essential.

[St Enodoc]

14 hours ago, steve W said:

Yes they will cook in a very short time, but a much longer time than that neccesary to operate.  With regards to your quoted paragraph, Don's words were in respect of the wire size not the solenoid coil itself  and are entirely correct.

"It is why CDU's are recommended"  Who reccomends? The primary purpose of the CDU is to store and make available more energy than that available immediately from a power source. That enables one or more solenoids to operate, but no better than a very large transformer would. The CDU automatically limits the supply by virtue of it's discharge and is a bonus that protects coils from jammed switch contacts on cheap and cheerful 'passing contact' switches. It also reduces sparking from switches (making them less likey to fail in the first place) for the same reason. The use of a CDU is of course a good thing, but for many different reasons. An adequate wire size though is essential.

"The CDU automatically limits the supply by virtue of it's discharge"

 

That is the reason they are recommended by almost every writer on the topic over the last 50 years. Everything you have written is absolutely true, of course, but the CDU's function of preventing coil burn-out is the main reason it is used.

[Donw]

9 hours ago, kevinlms said:

The transitory current/heating effect is VERY important in the likes of solenoid point motors. Too much current for too long WILL cook the coils in a short space of time.

It is why CDU's are recommended.

 

7 hours ago, steve W said:

Yes they will cook in a very short time, but a much longer time than that neccesary to operate.  With regards to your quoted paragraph, Don's words were in respect of the wire size not the solenoid coil itself  and are entirely correct.

"It is why CDU's are recommended"  Who reccomends? The primary purpose of the CDU is to store and make available more energy than that available immediately from a power source. That enables one or more solenoids to operate, but no better than a very large transformer would. The CDU automatically limits the supply by virtue of it's discharge and is a bonus that protects coils from jammed switch contacts on cheap and cheerful 'passing contact' switches. It also reduces sparking from switches (making them less likey to fail in the first place) for the same reason. The use of a CDU is of course a good thing, but for many different reasons. An adequate wire size though is essential.

 

Personally I consider any coil that will overheat to be a poor design. I designed fan motors that would not overheat even under stall conditions.

If it can overheat a fusible link is advisable.  That way there is no danger from a coil inadvertently left on.  However under normal conditons the power will only be on for a short time using passing switches or std and probe,  Use of a CDU givesa heftier wack from the usual auxilliary output of a model railway controller than it could deliver on its own.

 

Don

[drmditch]

Again, many thanks for the above contributions.

I do have quantities (at least two half reals) of what is labelled as '1.5AQMM' wiring, which is sold I believe for domestic lighting circuits.

I have so far used it for the 'ring mains' for traction power (not I emphasise for anything to do with any domestic lighting function) at 12vDC.

 

This works well.

 

It is quite stiff, and I now have to work out how I can use this for turnout control.

 

I will also now be able to work out sensible wiring run lengths, and thus the location of the panels etc.

This may be a little difficult, so I am giving it calm, and reflective thought!.

[kevinlms]

4 hours ago, Donw said:

 

 

Personally I consider any coil that will overheat to be a poor design. I designed fan motors that would not overheat even under stall conditions.

If it can overheat a fusible link is advisable.  That way there is no danger from a coil inadvertently left on.  However under normal conditons the power will only be on for a short time using passing switches or std and probe,  Use of a CDU givesa heftier wack from the usual auxilliary output of a model railway controller than it could deliver on its own.

 

Don

Well you missed an opportunity, as Peco redesigned their point motor recently. The old design was presumably out of patent, over the past few years loads of people have come up with alternative lookalikes.

 

I can remember buying some plastic cored versions of the Peco point motor and burning them out fairly quickly - it is then I found out about CDU's.

 

Originally, Peco point motors were recommended for AC usage. I guess that doing that made them so noisy, you couldn't miss it, if the power was left on. DC is almost silent, without the buzzing.

 

I designed fan motors that would not overheat even under stall conditions.

 

Probably you did, but if they were 240 Volt designs, then you MUST, but I don't believe the same restriction applies to low voltage items. Peco must have sold millions of PL10 point motors, I assume that they broke no law by doing so?

[Donw]

10 minutes ago, kevinlms said:

Well you missed an opportunity, as Peco redesigned their point motor recently. The old design was presumably out of patent, over the past few years loads of people have come up with alternative lookalikes.

 

I can remember buying some plastic cored versions of the Peco point motor and burning them out fairly quickly - it is then I found out about CDU's.

 

Originally, Peco point motors were recommended for AC usage. I guess that doing that made them so noisy, you couldn't miss it, if the power was left on. DC is almost silent, without the buzzing.

 

I designed fan motors that would not overheat even under stall conditions.

 

Probably you did, but if they were 240 Volt designs, then you MUST, but I don't believe the same restriction applies to low voltage items. Peco must have sold millions of PL10 point motors, I assume that they broke no law by doing so?

 

I said poor design not dangerous. I haven't heard of a case of one starting a fire. However I have heard of them burning out hence my criticism of poor design. In the days I was doing this work I think the law was much weaker and one had to look for the BEAB kitemark. Having seen an imported fan burst into flame after about 15mins once the fan was jammed it does indicate the seriousness of it. 

We used make fusible links by fixing a piece of solder to the end of one of the coil leads. The girl who used to do this left to have a baby and the others girls said it was too difficult. I was asked to sort it out. I went down to the production line and sat there making them not very quickly I admit. One of the girl said I can do  better than that and took over. The Production Manager said how did you sort it so quickly. Easy I said they didn't want to be outdone by a man.

 

 

Don

[Matt]

Sorry I've not had time to read all the posts here so this may have been covered already... I have similar problems to what you describe...

 

I have a very big CDU and have had no issues throwing up to 11 peco PL10s through a diode matrix even with quite long distances in the wires (20+ft).  However if I have even 1 side mount motor in the mix then things start to go wrong - mostly with the side mount.  I have found that doubling up on wiring to the side mount has helped but is still not 100%.  My only conclusion has been that the side mount is more power hungry than the PL10s or for some reason isn't getting the power it needs... interestingly not all side mounts are affected but I think this may be that some points are tighter to throw than others. I don't have an issue with ANY PL10 thrown points anywhere on the layout.

 

The other possibility I've not explored completely is that the side mounts come with their own weedy wires for a few inches whereas my PL10s are 24/0.2 wire direct onto the contacts - could this be making the route more difficult for the power to get to the side mounts when there are easier routes (you can probably tell i'm not an electrician...).  I did start cutting some of the harnesses short and connecting 24/0.2 wire as close as possible to the motors but I can't say for sure yet whether this has helped. 

 

In days gone by I also had problems mixing Peco and Hornby motors (PL10s and Hornby's similar looking unit) so have tended to avoid this on current setup where possible.  Notably the Hornby motors also come with their own weedy harness...

 

Will be interested to hear if you find a solution as this is an unresolved issue for me at the moment although COVID might give me more time to solve it...

 

M

[Junctionmad]

The car wiring is most likely Thin wall insulation , the current rating is better due to better cooling due to the thinner wall and also the nature of the current specification for car wiring is different to standard hook up wire , the 11A is actually a max figure and really half that is more applicable 

[Donw]

3 hours ago, Matt said:

Sorry I've not had time to read all the posts here so this may have been covered already... I have similar problems to what you describe...

 

I have a very big CDU and have had no issues throwing up to 11 peco PL10s through a diode matrix even with quite long distances in the wires (20+ft).  However if I have even 1 side mount motor in the mix then things start to go wrong - mostly with the side mount.  I have found that doubling up on wiring to the side mount has helped but is still not 100%.  My only conclusion has been that the side mount is more power hungry than the PL10s or for some reason isn't getting the power it needs... interestingly not all side mounts are affected but I think this may be that some points are tighter to throw than others. I don't have an issue with ANY PL10 thrown points anywhere on the layout.

 

The other possibility I've not explored completely is that the side mounts come with their own weedy wires for a few inches whereas my PL10s are 24/0.2 wire direct onto the contacts - could this be making the route more difficult for the power to get to the side mounts when there are easier routes (you can probably tell i'm not an electrician...).  I did start cutting some of the harnesses short and connecting 24/0.2 wire as close as possible to the motors but I can't say for sure yet whether this has helped. 

 

In days gone by I also had problems mixing Peco and Hornby motors (PL10s and Hornby's similar looking unit) so have tended to avoid this on current setup where possible.  Notably the Hornby motors also come with their own weedy harness...

 

Will be interested to hear if you find a solution as this is an unresolved issue for me at the moment although COVID might give me more time to solve it...

 

M

 

Quite obvious when you think about it. Never having used one I sort of assumed it used the same coils but obviously doesn't. Where a number of devices are powered in parallel those with the lowest impeadance will get more power. As they operate the impeadance will rise and their share will reduce, however the output from a CDU will also be declining. A larger capacitor will have more charge available and could be a help. If the impeadances of the two types are wildly different the only answer would be to stick to one type. It should be possible to have two CDUs one for each type but it would complicate the  Diode matrix you would have to have two layers of matrix independant electrically but the selector switches would have to be linked operationally. I can advise if you want to go down that route.

 

Don

[AndyID]

Solenoid point motors - absolutely hate the bloomin' things!

 

I think I still have some Atlas ones. Free to anyone in the US for the cost of postage.

[drmditch]

Again, thank you everybody for all the help.

Partial testing has shown that shorter cable lengths (apparently) remove the problem.

Now re-designing panel and cable runs/thickness.

I think I can do this now!

[DCB]

On 28/03/2020 at 17:41, Junctionmad said:

The car wiring is most likely Thin wall insulation , the current rating is better due to better cooling due to the thinner wall and also the nature of the current specification for car wiring is different to standard hook up wire , the 11A is actually a max figure and really half that is more applicable 

Car wiring generally has  thick wall coverings, not for insulation but because it has to be resistant to oils, heat, water and abrasion. Usually its about 50% thicker than equivalent mains wires.  Unlike mains wiring it does not snap when worked backwards and forwards a few times and comes in a huge number of colours not just Black, Blue, Brown, Red, Green and green/yellow.  It also doesn't get mistaken for a mains lead.