Best Power unit for CDU

[RonnieS]

I have a 16v 1 amp power unit feeding a 4 pot (each of 4,700) 18,800 microfarad CDU. The CDU is good and fires 3 points at the same time OK.

I have ambitions of firing 6. The CDU manufacturer claims it should fire up to 10 but recommends a bigger power unit. I was under the impression that when a Capacitor was full it was full and the voltage and amperage were not a consideration. Am I wrong? (I quite often am)

[RexAshton]

Any 24vDC psu will do. Try Gaugemaster.

 

[cliff park]

Make sure you keep well within the voltage ratings of your capacitors.  If the capacitors are rated at, say, 25V then it would be best to not go over 20V.  The amperage side of things will affect the charge time, but only marginally unless you went to extremes.

[RonnieS]

 

I guess my question was phrased badly. What I am trying to establish is would a bigger voltage supply give a bigger “belt” and improve performance? The recharge time is great but I don’t feel there is enough umpf to fire  a double demand. This is the  CDU I am (at present) using https://www.ebay.co.uk/itm/252833225672 

 

I did ask the supplier my question but his answer was a bit ambiguous. I feel the combined experience of rmweb readers is my best bet.

[kevinlms]

The required information is right there on the diagram, which it states next to the AC in terminals - Max 24VAC and output of 35V DC. The latter is the maximum rating of the capacitors, which can be seen in the photo.

 

I suspect that your existing power supply would work fine with more than 3 points. Easiest thing to do is try it out and address the issue if required. The solution would be a slightly higher input.

 

[RonnieS]

This question was in conjunction with a club project I am working on. So last night was another working session and proved the max number of point motors that could be fired at one go with my present set up was 4. I thought I’d post my solution on the off chance that some one else ever had a similar problem.

 

The top sketch shows a rough plan of 5 loops I hoped to control with said set up. Road 1 at the bottom, Road 5 at the top. Ignore the stars and the green numbers, the other numbers are the point motors. The king points each end are 3 ways so we have to fire motors 2,3,6,7  every time.

Road 1 fires 1,2,3,6,7,5

Road 2 fires 1,2,3,6,7,5

Road 3 fires 2,3,6,7

Road 4 fires 2,3,4,6,7,8

Road 5 fires 2,3,4,6,7,8

As the set up only fires 4 motors at once I decided to control the left and right on separate studs just drilling 4 holes and inserting new studs on Roads 1,2,4 & 5.

Edited May 19, 2023 by RonnieS
spelling

[Michael Hodgson]

On 17/05/2023 at 19:15, RonnieS said:

I have ambitions of firing 6. The CDU manufacturer claims it should fire up to 10 but recommends a bigger power unit. I was under the impression that when a Capacitor was full it was full and the voltage and amperage were not a consideration. Am I wrong? (I quite often am)

Yes, when the capacitor is full it's full.  It will fully recharge very quickly after being fired.  The amperage of the supply will only affect the speed of recharge rather than what output ot will deliver, so supply amperage doesn't make much difference.

 

An increase in supply voltage will increase the maxium charge the capacitor can store, and how much kick the CDU can deliver in one go, so it should be able to move more points in one go.  However the voltage needs to be within the limits of the components used, and the capacitors are usually the most vulnerable parts of a CDU.  As mentioned above the capacitors illustrated are rated at 35v. 

 

Unfortunately it's not that simple.  The voltage of a power supply tends to be quoted as the Root Mean Square (RMS), effectively the "average" voltage generated for a rectified AC supply.  However capacitors charge to the extreme peak of the voltage which is RMS voltage x √2.   In the case of a 24v RMS supply that's 34v, which only just within the rating of those capacitors.  As a crude rule of thumb, I would double the RMS rating the of the DC supply to ascertain the voltage rating I would look for in the capacitors to give a good safety margin for fluctuations etc..

 

Personally I don't like excessive power applied from a CDU to solenoid points - they can shoot across with a sound like a rifle shot, but that much force isn't good for the points.  As the available charge is spread roughly equally between the points you operate, you really only want enough charge to move those points.  If you move fewer points in some of your routes, that means more charge is passed through the solenoids on those routes, and they will suffer more wear and tear.

[RonnieS]

3 hours ago, Michael Hodgson said:

Yes, when the capacitor is full it's full.  It will fully recharge very quickly after being fired.  The amperage of the supply will only affect the speed of recharge rather than what output ot will deliver, so supply amperage doesn't make much difference.

 

An increase in supply voltage will increase the maxium charge the capacitor can store, and how much kick the CDU can deliver in one go, so it should be able to move more points in one go.  However the voltage needs to be within the limits of the components used, and the capacitors are usually the most vulnerable parts of a CDU.  As mentioned above the capacitors illustrated are rated at 35v. 

 

Unfortunately it's not that simple.  The voltage of a power supply tends to be quoted as the Root Mean Square (RMS), effectively the "average" voltage generated for a rectified AC supply.  However capacitors charge to the extreme peak of the voltage which is RMS voltage x √2.   In the case of a 24v RMS supply that's 34v, which only just within the rating of those capacitors.  As a crude rule of thumb, I would double the RMS rating the of the DC supply to ascertain the voltage rating I would look for in the capacitors to give a good safety margin for fluctuations etc..

 

Personally I don't like excessive power applied from a CDU to solenoid points - they can shoot across with a sound like a rifle shot, but that much force isn't good for the points.  As the available charge is spread roughly equally between the points you operate, you really only want enough charge to move those points.  If you move fewer points in some of your routes, that means more charge is passed through the solenoids on those routes, and they will suffer more wear and tear.

Super reply! That is a great help and an endorsement to the route I took. It also occurred to me that the diodes in the matrix might blow if the load was increased.

[Michael Hodgson]

49 minutes ago, RonnieS said:

Super reply! That is a great help and an endorsement to the route I took. It also occurred to me that the diodes in the matrix might blow if the load was increased.

Diodes in the matrix will be to individual solenoids, so they don't carry the whole load, only the power to one point motor.

So increasing the number of motors being worked won't stress them. 

 

Don't know what you're using for your diode matrix.  Using an increased voltage could give them a problem - they too are voltage rated.  The PIV rating on a diode is the most that you should ask it NOT to pass - ie to stop it letting a current through the wrong way.  When a solenoid operates, it generates a back EMF, a brief spike that you need to block, but this does tend to be a fairly high voltage - it can give you a kick!

A lot of people get away with 1N4001 for diode matrix operation using an auxiliary output from an old train controller power unit.

If the voltage is increased 1N5401 might be preferable and they are cheap enough if you have to upgrade.

 

[RonnieS]

45 minutes ago, Michael Hodgson said:

Diodes in the matrix will be to individual solenoids, so they don't carry the whole load, only the power to one point motor.

So increasing the number of motors being worked won't stress them. 

 

Don't know what you're using for your diode matrix.  Using an increased voltage could give them a problem - they too are voltage rated.  The PIV rating on a diode is the most that you should ask it NOT to pass - ie to stop it letting a current through the wrong way.  When a solenoid operates, it generates a back EMF, a brief spike that you need to block, but this does tend to be a fairly high voltage - it can give you a kick!

A lot of people get away with 1N4001 for diode matrix operation using an auxiliary output from an old train controller power unit.

If the voltage is increased 1N5401 might be preferable and they are cheap enough if you have to upgrade.

 

Wow! You are extremely helpful thanks. I can use this information on the next panel.