Diode matrix and mix of point motor types

[Matt]

I've recently started wiring up my large fiddle yard.  In order to set routes I've elected to use a diode matrix arrangement to thump multiple solenoid point motors into position.  I've had experience of this on previous layouts using a large CDU to provide the necessary thump.  On this latest layout, in testing I am finding some points to not throw reliably.  The ladder is a mixture of different point motors (Hornby under mounted, Peco undermounted and Peco surface mounted) out of necessity and it appears it is always the Peco surface mounts that struggle (though interestingly not all of them - some are OK despite being further from the control panel).  I guess I'm asking too much to get one CDU to throw all these points at once and am thinking about options to make these more reliable.  Can anyone make any suggestions?  I was thinking of maybe fitting a point specific small CDU for the sticky points but was not clear on how to go about wiring this in with the rest of the diode matrix so a single button can throw a whole series of points.

 

Thanks in advance for any pointers,

Matt

[kevinlms]

58 minutes ago, Matt said:

I've recently started wiring up my large fiddle yard.  In order to set routes I've elected to use a diode matrix arrangement to thump multiple solenoid point motors into position.  I've had experience of this on previous layouts using a large CDU to provide the necessary thump.  On this latest layout, in testing I am finding some points to not throw reliably.  The ladder is a mixture of different point motors (Hornby under mounted, Peco undermounted and Peco surface mounted) out of necessity and it appears it is always the Peco surface mounts that struggle (though interestingly not all of them - some are OK despite being further from the control panel).  I guess I'm asking too much to get one CDU to throw all these points at once and am thinking about options to make these more reliable.  Can anyone make any suggestions?  I was thinking of maybe fitting a point specific small CDU for the sticky points but was not clear on how to go about wiring this in with the rest of the diode matrix so a single button can throw a whole series of points.

 

Thanks in advance for any pointers,

Matt

I've not come across this problem, as the diode matrixes I've used drive identical point motors. However, any with a lower resistance, will take the lions share of the power. This is simple Ohms Law.

 

The first thing I'd try is to increase the input voltage to the CDU and see if that improves the reliability. If no good, then try a series resistance, for those with a lower resistance and try that.

[Matt]

Thanks Kevin.  I'm using a closed unit DCC Concepts CDU which doesn't give me any option to up the voltage (as it is I am looped across both capacitor banks to give it a big enough thump).  I'll have a play with the resistor idea - one of the point's reliability has improved by doubling up on feed wires suggesting that the differences are quite subtle between them such that this could help.

 

M

[spikey]

12 minutes ago, Matt said:

... as it is I am looped across both capacitor banks to give it a big enough thump ...

Probably a silly question, but this doesn't mean that you've now got two capacitor banks in series, does it?  If so, they ought to be in parallel to increase total capacitance.

 

Whatever, what's the maximum number of points you want to change at once, what sort of length is the wiring from CDU to most remote solenoid, and are you sure that wiring's man enough (if it's still OK to say that)?

[Tony Teague]

I have used a mix of original Peco solenoids and their later surface / side mounted motors PLUS Gaugemaster GM500 relays all fired simultaneously by a single CDU.

If I had a problem with any single point motor it was usually because of poor alignment or a sticking slide bar, rather than an electrical issue.

Where the points and related diode matrix are furthest from the panel (30') I did have initial issues relating to the wire used being not "man enough" for the job as Spikey describes, and so I had to upgrade the trigger and return wires for that zone.

I also suggest using the maximum voltage allowed to fire the CDU and for this to be the largest available if you want to fire multiple points; I found that those mounted on DCC Concepts Sfx boards were too weak to throw more than 2 motors, so I removed the boards (my layout is DC in any case) and I use a chunky CDU from BLOCK signalling (their largest).

The result is that I can easily foire 5 point motors at once, plus 2 or 3 relays as required.

You will probably need to work through each of these issues to get the best result - CDU output, cable size, free movement of points etc.

Good luck!

 

 

Tony

[John ks]

A circuit like this might work

I have never built the circuit so I cannot guarantee it's effectiveness

The CDU's would have to have the (-) connection going from the input to the output (the very basic CDU circuit is shown at the bottom of the drawing to indicate the through connection of the (-) negative

 

The number of CDU's can be increased or decreased as required.

The number of points on each CDU can be changed as required.

 

Hope this helps 

John

 

[Matt]

Thanks Spikey, Tony and John,

 

I'll give all of these suggestions a go.

 

As regards "looped over capacitors" - I followed the instructions with the DCC Supplies CDU where it advised using the outermost terminals to get the biggest punch to the points (and it is more punch than the inner terminal which I started with) so I would assume in parallel as you recommend.  

 

The wiring is quite long and I have used quite beefy wire and then doubled up the return from the sticky motors - this has helped in a couple of cases.  The thing that is odd though is that the furthest PMs from the CDU (which are also side mounted Pecos) have been entirely reliable from first testing so isn't simply a length thing.  I have noticed that the supplied motors have quite long and very weedy wires which in some cases I have trimmed right back - I haven't yet correlated whether these are the reliable ones but will take a look later.  They are only maybe 10cm long so feels unlikely but they are really weedy inside... maybe just adding a little resistance?

 

One of the stickiest PMs I removed the other night and reinstalled to try and ensure that the fit was perfect - couldn't see any binding and did try a little lubrication to no avail... I do agree this also seems a likely cause as my actual points do vary in age quite substantially between new and maybe 15 years old.  Again not correlated precisely but think it is generally newer points at fault... but then I generally fit side PMs to newer points as most of my older ones had the lugs cut off the tie-bars on a previous layout...

 

I didn't mention it initially but I may just be pushing my luck expecting to get so many points firing on one CDU... I do have 11 points working reliably on one ladder (where all PM types the same) throwing at once, but this ladder is giving my grief ALWAYS on the sidemounted PM type.  There is of course the possibility that I am bad at installing side mount PMs...

 

John - appreciate your diagram which makes perfect sense and I will give a go to if all else fails - I was trying to do this one in my head and hadn't quite fathomed it out so much appreciated.

 

I will report back when I solve the issue - may be a while as we have guests coming this week which will restrict play time...

 

M

 

[Broadway Clive]

I prefer to have relays and CDUs fitted on boards and located close to the points for maximum power. Fitted with connectors, the boards can be assembled and tested in comfort on the workbench. Once in position buttons on the control panel and detectors on the track are then used to send 12v+ to operate the relays as required and those circuits control the point combinations and safety features to make operations easier and prevent collisions. 

[Tony Teague]

ELEVEN points simultaneously on a single CDU does sound like a difficult ask!

Is there no way of dividing this circuit in two - otherwise I'd persoanlly doubt that there is an off the shelf CDU available with sufficient power.

I may of course be wrong!

Tony

[DCB]

Eleven points is pushing the boundaries, but if you have a big enough capacitor it shouldn't be an issue, but if you throw 11 points and then just 4 or 5 you risk puling the point blades out of the tie bars. I add extra point motor coils as sort of dummy motors or dead loads to diode matrix so the load is more or less the same no matter how many points are thrown.

[Gordon H]

15 hours ago, Tony Teague said:

ELEVEN points simultaneously on a single CDU does sound like a difficult ask!

Is there no way of dividing this circuit in two - otherwise I'd persoanlly doubt that there is an off the shelf CDU available with sufficient power.

 

 

More to the point, will your push button switches be able to cope with such a large instantaneous current of probably 15A or more?

[spikey]

3 minutes ago, Gordon H said:

 

More to the point ...

 

Hah!  Very good, Gordon ...

[kevinlms]

1 hour ago, Gordon H said:

 

More to the point, will your push button switches be able to cope with such a large instantaneous current of probably 15A or more?

Making the contact isn't normally a problem,  it's whenever you attempt to break it causes contact welding. Holding the button down long enough to discharge the CDU, should prevent that.

[DCB]

21 hours ago, kevinlms said:

Making the contact isn't normally a problem,  it's whenever you attempt to break it causes contact welding. Holding the button down long enough to discharge the CDU, should prevent that.

Using stud and probe "Electric Pencil" stops the switches burning as it does not use switches.  The probe and studs burn but can easily be cleaned with a file.   Alternatively a really heavy duty separate energising switch can be used with two position point switches. That way the heavy duty switch makes and breaks the circuit and the individual route switches are not taking and power when you switch them. I have been switching up to 5 big Hammant and Morgan solenoids with a 1 amp rotary switch but a 5 amp energising switch for over 20 years without any issues.  My panel hinges out if required by generally lies vertical and takes up zero room horizontally, a big consideration for many space limited modellers.  Also the controllers are under the layout again to save horizontal space, space which I then  use for track.

 

On 14/10/2019 at 13:13, John ks said:

A circuit like this might work

I have never built the circuit so I cannot guarantee it's effectiveness

The CDU's would have to have the (-) connection going from the input to the output (the very basic CDU circuit is shown at the bottom of the drawing to indicate the through connection of the (-) negative

 

The number of CDU's can be increased or decreased as required.

The number of points on each CDU can be changed as required.

 

Hope this helps 

John

 

Good idea except if all 3 CDUs are the same the one single point is going to get a much bigger thump than the two point CDUs.  Just add another point motor solenoid to the single point circuit to balance the load otherwise 3 or 5 years down the line the point blades may come loose in the tie bars of the single point

[Gordon H]

22 hours ago, kevinlms said:

Making the contact isn't normally a problem,  it's whenever you attempt to break it causes contact welding. Holding the button down long enough to discharge the CDU, should prevent that.

Two potential situations arise from this which would still cause the problem I was alluding to - switch bounce and too brief a pressing. There is no way to guarantee either of these won't occur, and with the level of current involved it only has to happen once or twice to ruin the switch. Far better to use a solid state solution rated for the job.