Theory behind CDUs

[OldTom]

My question is why does an AC supply gradually charge up a point motor CDU? If my electrical theory is correct (and I was once wrong LOL !)the charge from an AC supply to a simple capacitor (is a CDU just a capacitor?) should be maximised to positive charge, then discharged, then negatively charged and then discharged in one whole cycle. However AC reportedly can take seconds to ‘maximise charge’.  If a capacitor has a DC supply then the transient is just the standard first order transient.

 

A related but solved question is why does AC operate the solenoids based point motor without it resulting in oscillation at the frequency of the AC supply. The answer to this is clearly due to the core of the solenoids are not magnetised, and both positive and negative poles attract the core. Repulsion and oscillation only occurs if the core is magnetised.

[Bedlington North]

I would have though it was relatively simple to use at least one diode in the AC supply to the CDU to give half-wave rectification. At least....

[Crosland]

It will have a rectifier.

 

The only advantage to an AC supply was that, in the old days, they were cheap (just a transformer or the aux output of a DC controller) and the capacitor would charge to the peak value, rather than the quoted RMS value.

 

Nowadays transformers are (relatively) expensive to buy and to post, and almost everything uses switch node power supplies. There's no reason you can't use DC but you will need a higher specified voltage (approx 1.4x the AC voltage it's replacing) to get the same "oomph".

 

Energy stored in  a capacitor is proportional to the square of the voltage so a small increase in voltage is always worthwhile, so long as you stay within the spec of the capacitors.

[OldTom]

Thanks guys! Yes, basically as you say the CDU must rectify the AC supply before the capacitor. For some reason though, it then seems to be preferred to have the rectifier within the CDU rather than in the transformer, which is of course what you have with the classic model rail dc supply. Perhaps this is because you need a higher voltage (for the reason Crosland indicates) than usually available with a transformer used for running rail DC supply?

Edited December 31, 2020 by OldTom
Additional point

[DCB]

22 hours ago, OldTom said:

Thanks guys! Yes, basically as you say the CDU must rectify the AC supply before the capacitor. For some reason though, it then seems to be preferred to have the rectifier within the CDU rather than in the transformer, which is of course what you have with the classic model rail dc supply. Perhaps this is because you need a higher voltage (for the reason Crosland indicates) than usually available with a transformer used for running rail DC supply?

I don't think that is what Crossland is saying.  To me rectified means full wave DC.  You will lose around 0.7 volts  ballpark in each diode when rectifying AC to full wave DC.  Aa bridge rectifier has 4 diodes that's 2.8, volts more or less  the difference between the rated 16 volts AC and 12 volts DC output on many controllers.    If you don't rectify the AC but put a diode in circuit to run at half wave then the drop is only the one diode in circuit so only around 0.7 volts so the Capacitor will charge an extra 2 volts, quite a useful boost, though it will take longer to get to capacity,   longer to get to 12 volts than the rectified set up.  I have never connected a polarised capacitor to AC but I suspect it wouldn't charge, however I did connect one backwards and after a suitable period to charge up it exploded and blasted vile smelling yellow dye over the shed floor.

[cliff park]

A capacitor can never charge, as such, on AC. If it is non-electrolytic it simply charges and recharges in each direction with each cycle. If electrolytic it will most likely just go bang. (We had a contractor who accidentally connected a large electrolytic in reverse, it went bang and covered him in the yellow liquid. His face retained a yellow tinge for months and he was for ever more called Nanky Poo)

[CourthsVeil]

"A bridge rectifier has 4 diodes that's 2.8, volts ... less" ...

I doubt that, because the four diodes will never be used at the same time.

Two are responsible for the positive pulse, the other pair a hundreth second later for the negative pulse.

And what has this to do with "the difference between the rated 16 volts AC and 12 volts DC output on many controllers." ??

[BR60103]

I assume that CDUs are designed for AC because most model controllers have AC accessory terminals but many do not have uncontrolled DC. Also means you don't have to check polarity going in.

Point motors work on AC because the thingy inside that's attracted to the magnet coil is not in itself magnetized -- just a lump of iron (or equivalent).

 

[kevinlms]

12 hours ago, BR60103 said:

I assume that CDUs are designed for AC because most model controllers have AC accessory terminals but many do not have uncontrolled DC. Also means you don't have to check polarity going in.

Point motors work on AC because the thingy inside that's attracted to the magnet coil is not in itself magnetized -- just a lump of iron (or equivalent).

 

CDU's aren't specifically a model railway item.

 

Any* solenoid or coil based motor will work on AC or DC, they just generate a magnetic force, which attracts the iron core towards itself.

 

On AC, they usually buzz or rattle, because the magnet force is being created & collapsing at 50 or 60 times a second.

 

*There can be specially designed ones with special characteristics, but they are intended for specific purposes.

Edited January 2, 2021 by kevinlms
More info

[AndyID]

22 hours ago, DavidCBroad said:

I don't think that is what Crossland is saying.  To me rectified means full wave DC.  You will lose around 0.7 volts  ballpark in each diode when rectifying AC to full wave DC.  Aa bridge rectifier has 4 diodes that's 2.8, volts more or less  the difference between the rated 16 volts AC and 12 volts DC output on many controllers.    If you don't rectify the AC but put a diode in circuit to run at half wave then the drop is only the one diode in circuit so only around 0.7 volts so the Capacitor will charge an extra 2 volts, quite a useful boost, though it will take longer to get to capacity,   longer to get to 12 volts than the rectified set up.  I have never connected a polarised capacitor to AC but I suspect it wouldn't charge, however I did connect one backwards and after a suitable period to charge up it exploded and blasted vile smelling yellow dye over the shed floor.

 

Only two of the diodes in a bridge are passing current during each half cycle so the voltage drop is more like 1.4 volts.

 

You can make a non-polarized capacitor using two electrolytic caps connected in series but pointing in opposite directions. The capacitance will be half. Caps in series follow the same rule as resistors in parallel.

[Pete the Elaner]

I found a circuit diagram for a CDU on the net some time back which uses what I felt was a neat trick to double the voltage applied to the motor.

It only works on AC because it has 2 capacitors in series but the diodes are arranged so that each capacitor is charged up by a different phase but when they are discharged, you get twice the voltage.

 

http://dccdiy.org.uk/cdu.html

[Crosland]

17 hours ago, Pete the Elaner said:

I found a circuit diagram for a CDU on the net some time back which uses what I felt was a neat trick to double the voltage applied to the motor.

It only works on AC because it has 2 capacitors in series but the diodes are arranged so that each capacitor is charged up by a different phase but when they are discharged, you get twice the voltage.

 

http://dccdiy.org.uk/cdu.html

 

Repeating what i said above, this is good as you get the benefit of the square of the voltage. Doubling the voltage gives you 4x the energy stored, and avaiulable to fire the solenoid.

 

The MERG DCC solenoid decoder works in a similar way.

 

Just make sure then capacitors are correctly rated!