CDU use - advice please!

[clockman]

In my original layout I have used a CDU unit to switch a pair of points together. This works fine.

I am now ‘tidying up’ the whole layout as I embark on the next building phase and am trying to reduce the wire count. So my question is:

‘Can I use a common wire from the point switch (fed from the CDU) to feed both the point motor and the separate relay that switches the frog feeds’.
I will provide a common return from both the point and the relay as well.

I am concerned that
i) the CDU spike might damage the relay and
ii) the CDU might not switch the points with the extra load

The CDU is an old ECM unit with a 3900 microfarad capacitor fed from the 16v a.c. output from my Gaugemaster DF controller. At the moment, the CDU happily switches 2 points together but I use a separate 12v supply for the relays.

Thanks for any input in advance!

[ikcdab]

The cdu delivers a one-off jolt to throw the solenoid then cuts out.  Whereas a relay requires a constant voltage to keep it latched.  So I'm not sure the two are compatible.

If they are, then you need to measure the cdu output with a multimeter and make sure that is within the parameters of the relay.

[clockman]

Thanks Ian. The relays are latching types so once triggered they stay switched. I note your comment on the relay parameters, though!

Martin Ballard

[BR60103]

I've had problems with point motors in parallel that are different makes or even different eras of the same make.  I think one unit may suck all the current from the other ones.

 

[kevinlms]

3 hours ago, BR60103 said:

I've had problems with point motors in parallel that are different makes or even different eras of the same make.  I think one unit may suck all the current from the other ones.

 

Yes, the one with the lowest resistance will take most of the power.

[Enterprisingwestern]

 

Electricity always takes the path of least resistance, so I would put separate feeds in, this will allow for future times when one circuit tires a little more than the other.

 

Mike.

[Suzie]

Yes, you can do it, but you will need to consider a couple of things:-

 

1. The resistance of some point motors can be as low as 4 Ohms (2 Ohms for two in parallel), so you will need to ensure that the wiring and switches do not provide any resistance at all. It is very easy to get a few Ohms of resistance in the circuit which will see perhaps as little as 6V at the point motor at the start of the capacitor discharge (voltage drops after that) - don't be surprised if the relay does not fire if you are using thin wiring.
2. At the other extreme the CDU will give out around 25V when fed with 16V AC, which can be too much for a 12V relay if the point motors become disconnected.

In my experience to get reliability when using latching relays with solenoids I use a 35V DC supply to the CDU (rather than 16VAC) to give a bit more oomph, and use 24V latching relays mounted near the control panel (with a long frog feed). Trying to simplify the wiring like you intend just means you will be using very fat wires. Using the higher voltage allows you to get away with thinner wires. The ECM unit is rated up to 24V AC so should be good for 35V DC.

[clockman]

Thank you all for very useful feedback. I will take it all on board and report the outcome in due course - which at the present rate might be a little while!

[Sol]

Could use one of these http://www.heathcote-electronics.co.uk/point_indicator.html#POINT-INDICATOR-STANDARD

[clockman]

As promised an update on this project:
In short, yes, it works fine.
First my kit:
I’m using old materials recovered from my late uncle’s finescale layout. Track is Scaleway and the custom 5 point matrix is, I believe, Marcway. Point motors are Peco SL10s which I have re-pinned with longer pins to mount below the track board. CDU as stated earlier ECM.
Distance between controller and control panel 2 metres. Distance between control panel and layout board 2 metres.
Issues you raised for consideration:
Matching point motors:
I used a pair that matched to better than 0.1 ohm
Relays:
The relays are Hongfa PCB mount latching signal relays part number HFD2/005-S-L2 mounted on Vero board strips under the board and near the poin t motors. While the spec says 5V coil I have used them in my earlier layout with no problems. The CDU output port measures 23V but without an oscilloscope I cannot measure the spikes. The relays have an adequate switching spec and are available from RS components at under £3 / pair.
Ohms law: 
I have used standard alarm cable (7/.02) but have stripped the wires out of the sheath and doubled up on all signal wires - so now 14/.02. The 25 way D lead between the CP and the track board has 2 pins allocated for the common returns to minimise loss due to the lead. 
The result is a solid 'clunck' switch with no hesitation and I have reduced the signal wire count to a single  25way lead with a separate 6pin Belkin lead for the track and auxiliary 12v lines. 12V wires also doubled to 14/.02. (Belkin was a great little connector now no longer available. Used them in my youth…a while ago!)
So, thank for your inputs. Result achieved and remaining track layout now proceeding! Pictures, circuits and other info can be supplied if wanted.

[Colin_McLeod]

On 01/03/2024 at 19:07, clockman said:

Thanks Ian. The relays are latching types so once triggered they stay switched. I note your comment on the relay parameters, though!

Martin Ballard

A latching relay requires a constant current which the cdu  won't give.  I suppose  it would work with a twin coil relay of some sort  using a second supply to keep it latched. 

[BoD]

1 hour ago, Colin_McLeod said:

A latching relay requires a constant current which the cdu  won't give.  I suppose  it would work with a twin coil relay of some sort  using a second supply to keep it latched. 

Sorry Colin, a latching relay does not require a constant supply once activated - hence the term latching. They are switched on and off by pulses of current.

[melmerby]

On 02/03/2024 at 07:50, Suzie said:

The resistance of some point motors can be as low as 4 Ohms

That's on the high side, a Seep is about 2.6 ohms, I've seen others that are similar, sometimes less.

[Colin_McLeod]

BoD I thought Clockman was  referring to the arrangement where a relay is energised then one of its own set of  contacts maintains the energy to the coil.

Such an arrangement  requires a current  through the coil.  The interruption of that current by pressing a normally closed push button then switches the relay off.

 

 

Are you referring to some kind of mechanical  latching? How is such a relay then switched off?

Edited April 17 by Colin_McLeod

[melmerby]

13 hours ago, Colin_McLeod said:

BoD I thought Clockman was  referring to the arrangement where a relay is energised then one of its own set of  contacts maintains the energy to the coil.

Such an arrangement  requires a current  through the coil.  The interruption of that current by pressing a normally closed push button then switches the relay off.

 

 

Are you referring to some kind of mechanical  latching? How is such a relay then switched off?

Common Latching Relays change state by using two coils, one to operate, one to release.

 

IIRC some have a mechanical latch which is operated by one coil, first pulse operates, second pulse releases, rather like those push - push switches that switch on and off

Edited April 18 by melmerby

[melmerby]

Here's some manufacturers data on 2 coil devices (made for the railway industry):

datasheet-kcd-u200-v1-4.pdf

[Jeremy Cumberland]

11 minutes ago, melmerby said:

Common Latching Relays change state by using two coils, one to operate, one to release.

Or one coil where you apply the reverse polarity to change state - which is of little use with solenoid point motors.

 

Like @Colin_McLeod, until I got involved using miniature relays for operating models, I was used to the term "latching relay" being used for a relay powered via a pair of its own contacts (and a break switch of some description), and this does indeed require a continuous power supply. The problems of terminology.

 

13 hours ago, Colin_McLeod said:

Are you referring to some kind of mechanical  latching? How is such a relay then switched off?

As far as I am aware, they aren't usually latched mechanically, it is just that they don't have a spring, so the armature stays in the position it was last moved to. They are also called bistable relays.

[melmerby]

5 minutes ago, Jeremy Cumberland said:

As far as I am aware, they aren't usually latched mechanically, it is just that they don't have a spring, so the armature stays in the position it was last moved to. They are also called bistable relays.

Here's a mechanically latched relay using a cam:

 

Look mat about 40 seconds in and the operating cam can be seen

[Jeremy Cumberland]

16 minutes ago, melmerby said:

Here's a mechanically latched relay using a cam:

 

Look mat about 40 seconds in and the operating cam can be seen

I had wondered how the single-coil single-polarity version worked.

 

I don't think I'd like to build one into a circuit, not if any kind of automation were involved. What if it failed to set? If the fault were with the coil or latching mechanism, then one pulse ought to set it, but if the fault were with the contacts or somewhere else in the switched circuit, then the pulse would reset the relay. At least with a dual coil or reverse polarity relay you know whether a pulse will set or reset the relay, depending on the polarity of the pulse or which coil you send it to.

 

I wonder if one of the contacts was used just for proving which way the relay was set.

[BoD]

16 hours ago, Colin_McLeod said:

BoD I thought Clockman was  referring to the arrangement where a relay is energised then one of its own set of  contacts maintains the energy to the coil.

Such an arrangement  requires a current  through the coil.  The interruption of that current by pressing a normally closed push button then switches the relay off.

 

 

Are you referring to some kind of mechanical  latching? How is such a relay then switched off?

Others have replied but RS-online has a useful explanation - again, as others have said, it may be a case of the terminology being used.

https://uk.rs-online.com/web/content/discovery/ideas-and-advice/latching-relays-guide

 

[melmerby]

2 hours ago, Jeremy Cumberland said:

I had wondered how the single-coil single-polarity version worked.

 

I don't think I'd like to build one into a circuit, not if any kind of automation were involved. What if it failed to set? If the fault were with the coil or latching mechanism, then one pulse ought to set it, but if the fault were with the contacts or somewhere else in the switched circuit, then the pulse would reset the relay. At least with a dual coil or reverse polarity relay you know whether a pulse will set or reset the relay, depending on the polarity of the pulse or which coil you send it to.

 

I wonder if one of the contacts was used just for proving which way the relay was set.

I can't see why this would be any different to any other latching relay.

If for some reason it doesn't operate properly and doesn't change state, the dual coil and single coil reversible would have the same problem.

You can't reset it if it hasn't set in the first instance.