The rate at which you switch with just a psu is pretty much irrelevant,
it is the peak current capability that is the limiting factor; and how
much the voltage sags when you try to draw excess current. PSUs will
recover to their full o/p pretty much instantly once the load is removed.
That is correct in as far as it goes, but a CDU will still fail if there
is significant resistance between it and the motor. All a capacitor is
doing is effectively reducing the output impedance of the source, but
only for a very short period. Once discharged it will present a low
impedance to the psu, and without any current limiting will initially
look pretty much like a short circuit to the psu. Add to that the fact
that capacitors do not like large inrush currents; that is why you will
find ripple ratings on capacitors.
: I think it was my comment about requiring the wiring to
: be "as thick and short as possible" that set Jerry off on
: one :-)
No, it's your utter ignorance which does that, and your continual
refusal to accept that you are wrong -hence why you chose to take
a cheap shot at me- rather than to address the points made by
Oh well, as they say, none so blind as those who chose not to
: The point is that we would not be discussing CDUs if
: the power supply were so over specced (compared to
: what could be used with a CDU) to the extent that it
: could power a solenoid once a second, say.
But most dedicated PSU's could, it is your design of CDU that
can't do that, you just don't get do you, to much Googling and
not enough real world experience is your problem, not helped by
people who wish to play with electronics (as fun as that can be)
more than they actually want top play with model trains. The
problem the OP has is that his PSU can barely supply enough power
for the locos never mind anything else, your idea of a CDU would
actually make his problems worse because it would prolong the
periods of inadequate supply, unless you care to explain were all
this 'free energy' is coming from that you keep whitening on
The OP would have been as well to have ordered a stand alone PSU
(and built a control panel) for his accessories as buy a
different DCC system, other than the fact that he is well advised
to move away from the basic (trainset) end of the DCC market.
: Don't try putting words in my mouth. I'm claiming that
You don't seem to know what you are claiming, that's the problem!
: a CDU allows the use of a power supply that would
: otherwise be inadequate to switch a solenoid motor.
: It's quite a simple concept
Indeed it is, you need a PSU that is adequate to power the
solenoid regardless of if a CDU is fitted, period.
The point you are missing is that what is that "adequate" changes when
you use a CDU.
Assume 4 ohms for the solenoid coil resistance. Assume a PSU of 16V.
That requires a PSU rated at 4 Amps to drive the solenoid directly,
unless you deliberately overload the PSU each time you fire point
With a CDU you could easily limit the charge current to 2A and use a
lower rated PSU.
Perhaps of more interest to the OP, I have 2 peco point motors throwing 2
peco points from a single output of a hornby accessory controller connected
to a Select. No problem with points changing. Another point (yep), by
connecting points that are likely to be changed on a single route to
different accessory controllers there is less work for any single
Finally is a few seconds between point changes that important - lets face it
how long did it take a real signalman to set up a route ?
Just for reference a lot ex BR PSB interlockings use a battery bank to
do a similar job as the CDU does for the modeller so that when a route
is set all the points can move at the same time, at least on Southern
where I worked on the S&T.
For my fiddle yard I use a CML DAC10 that allows me use inputs to set
routes and allows you to adjust CDU pulse time for each motor and
recharge time afterwards so when I set a route I see the points
switching one after the other. It has Loconet interface for feedback to
controllers or computers connected to the network if you need that.
: The point you are missing is that what is that "adequate"
: changes when you use a CDU.
Err, MBQ, you are missing is the *fact* that it does not, as even
Jeff has pointed out, and indeed a CDU can actually make problems
My keyboard is working pewrfectly !!!!!!!!!!!!!
No, you are missing the point, if there is little resistance in the wire
and the psu is capable of supplying the peak current required then a CDU
will not improve things.
If that is not the case and, for example there is significant wiring
resistance, then you could just add the capacitor without the series
resistor. As long as the wiring between the resistor and the motor is
*short & low resistance* you will see the same effect as having the
resistor there *except that there will be a large inrush current to
re-charge the capacitor, which will be limited only by the wiring
resistance and the psu o/p capability.
You example of throwing multiple points from one of this type of CDU is
a red herring since for one CDU to operate many points the wiring from
the CDU to the points *must* be low resistance and because of multiple
points it cannot be close to all of them. So you might as well throw
away the CDU and just make sure that you have low resistance and a psu
that can provide the peak current.
Of course the above does not apply if the CDU runs a higher voltage that
the normal rating of the motor, and provides a short pulse at a higher
voltage. In that case more series resistance can be tolerated.
No exactly the opposite if you are talking about DC, stepping down
normally wastes the excess voltage as heat, very inefficient, unless you
use some form of switching regulator. If you are talking about AC & a
transformer, or voltage doubler, there is no significant penalty as
conservation of energy applies, less any small transformer loss, whether
you are stepping up or stepping down.
That is no different to them operating at 12 or 18V and using 24V
: You example of throwing multiple points from one of this type
of CDU is
: a red herring since for one CDU to operate many points the
: the CDU to the points *must* be low resistance and because of
: points it cannot be close to all of them. So you might as well
: away the CDU and just make sure that you have low resistance
and a psu
: that can provide the peak current.
Err, I think that *was* my point! Lets get back the basics of why
CDUs became popular in the 1960s, it had nothing what so ever to
do with wiring size as most people routinely used wire more than
adequate for the task, it was to (try to) prevent solenoid
burn-outs due to poorly laid and/or 'sticky' points -hence my
previous comment about giving the solenoid a "kick up the arse"-
and as you point out to add some protection to the coil windings
should the power continue to be feed to the coil due to the unit
not throwing over.
: No exactly the opposite if you are talking about DC, stepping
: normally wastes the excess voltage as heat, very inefficient,
: use some form of switching regulator. If you are talking about
AC & a
: transformer, or voltage doubler, there is no significant
: conservation of energy applies, less any small transformer
: you are stepping up or stepping down.
Best you tell that to the electric utility companies then mate,
they will love that revolutionary discovery! Free, or at least
paid for, lunches all round...
: > It would actually make more sense for these point solenoids
: > made to work at a (continuous rating) of ~ 6 to 8 volts and
: > power them for short periods at 12 to 16 volts, without any
: > for a CDU.
: That is no different to them operating at 12 or 18V and using
Indeed, and no CDU in sight, except that most trainsets/layouts
do not have a 24v supply, hence the suggestion of 6 to 8V coils.
The electricity companies use the principle all the time, if it were not
the case then there would be horrendous losses in the supply system.
Stepping the voltage up and down with transformers is essentially a
lossless process, apart from the small losses in the actual transformer
The National Grid is stepped up to EHT voltages, up to 400kV, because
the resistive losses in the lines will be lower for a given power
consumption. However, the *POWER* will remain the same, and when the
EHT is stepped down to lower voltages the current will go up, *but* the
power will still remain the same, (less any losses in the transformers).
This works for both stepping up and stepping down in voltage, the
current changes but the power must remain the same due to conservation
DC is a different matter as you can't use transformers, so in order to
step DC *UP* you must convert it to AC (as in a switching
regulator/psu). To step it down you can dump the excess voltage, but it
has to go somewhere, and that is in heat, as in a simple linear
regulator, or even a series resistor; or you can again use a more
efficient switching regulator.
When you step up to a higher voltage in a CDU the step up is not
'inefficient' the additional current is needed because you are putting
more current through the coil (for a short period), which is you aim in
order to make the point move quickly. There is no, or at least, very
little penalty in the transformation process itself.
The situation when you charge a CDU capacitor is entirely dependant on
the charging circuit. If you have a 12V supply & capacitor voltage you
need some series element to protect the psu from seeing a low impedance
as the capacitor discharges, and also to keep any recharge current
within the capabilities of the psu, & to prevent high in-rush currents
into the Cap. & to prevent a high continuous current into the coil.
If you step up to a higher voltage you either need a a higher voltage dc
psu, or some means of transforming a lower voltage up. An AC psu and
transformer would be the easiest, but a voltage doubler circuit would
do, but are generally capable of lower current outputs (which would
slow the re-charge time) and are somewhat more inefficient. A switch
mode regulator would be efficient but is more complex.
Either way you still need current limiting for the reasons given above.
If you pulse the motor at a higher voltage the peak current will be
higher through the coil (the whole point of the exercise), so more
energy will be taken from the cap, so for a recharge will either take
longer for a fixed input power or more power and a higher current will
be required, however, the transformation to the higher voltage is not
'inefficient', the higher current is required merely because you are
putting more current through the coil.
It has 2 very useful benefits: once discharge the current flowing is
limited so it is less likely you will burn out a point motor; Once
discharged the current flowing is limited and you will minimise damage
to the switch contact when breaking the current flow .
 Connecting a solenoid to a power supply through a switch means
that the full current determined by the solenoid coil resistance all
the while the swutch remains closed. This is a DC current. Breaking
the current by opening the switch leads to arcing between the swicth
contacts which damages the contacts. Given the cheapskate nature of a
lot of modellers I can guarantee many people are using switches that
are not really up to the job.
Look at the "current blocking" design "A resistor to limit the maximum
charging current to a reasonable level".
A better example would be http://www.merg.org.uk/merg_resources/dcc/download/acc2bsch.pdf
where you can see the classic two transistor current limit circuit.
Stating the obvious.
The amount of current it draws during the recharge phase is under
control of the CDU designer. It's total charge that matters, which is
proprtional to both current and time.
Yes, it is designed to discharge very quickly into a low impedance
load such as a solenoid point motor. That discharge takes of the order
The whole point of a CDU is to buffer the main supply from these very
large, but brief current pulses. To give a "discharge" that could not
be supplied by the main supply. A *well designed one* does that by
recharging more slowly.
As an example A few amps for a few milliseconds into the solenoid can
be replaced by recharging in a few tens of milliseconds (10x the time)
at a few hundred milliamps (1/10 current).
No it iosn't. You really do need to go back to basic theory.
Doubling the voltage in the charging circuit quadruples the energy
stored in the capacitor so you can give an even bigger kick to the
solenoid. The discharge current will initially be double due to simple
application of ohms law. The charging current and charging time are,
again, under control of the designer.
Completely different, and irrelevant to this discussion, scenario.
High voltage is used in the grid to reduce resistive losses.
WTF has this got to do with CDUs and point motors?
: > Does it, looks to me that it limits current draw on
: > the recharge side *when the CDU is discharging*,
: > thus protecting the recharging side from possible
: > overload.
: Look at the "current blocking" design "A resistor to
: limit the maximum charging current to a reasonable
: A better example would be
: where you can see the classic two transistor current limit
So there is an alternate power path to by-pass the CDU should the
capacitor(s) be discharged? By limiting the charging current one
also limits the total current draw through the unit at all times
and thus acts as overload protection. A CDU doesn't need a
'trickle' charge, it needs to recharge as quickly as possible,
ready for the next operation (or should do if of decent design).
: > A CDU has to recharge in seconds, that is because
: > the next operation could (and probably will) be in
: > seconds,
: Stating the obvious.
So why do you claim otherwise, with your 'trickle charge' effect!
: The amount of current it draws during the recharge
: phase is under control of the CDU designer. It's total
: charge that matters, which is proprtional to both
: current and time.
No, only time is in the hands of the designer, current is a fixed
known, and unless you are prepared to allow the unit to fail
completely due to no fail back should the capacitor be discharged
even time isn't really in the designers hands ('time' is governed
by the track plan and or control panel design).
<snip most of MBQs ignorance>
: The whole point of a CDU is to buffer the main supply
: from these very large, but brief current pulses.
Utter claptrap. A CDU is not like a cars ignition system, it's
more like the cetral locking system, think about it....
: You really do need to go back to basic theory.
Stop talking about yourself MBQ, you do not have a clue as to how
and why a CDU is used/works.
: Doubling the voltage in the charging circuit quadruples the
: energy stored in the capacitor so you can give an even
: biggerkick to the solenoid.
Yes, but you don't get that effect for free, which is the point,
the OP doesn't have the spare amps to use in that way.
: The discharge current will initially be double due to
: simple application of ohms law.
No one is disagreeing about the discharge or current provided.
: The charging current and charging time are, again,
: under control of the designer.
Again, only time is, assuming s/he is willing to trade
reliability and/or usability.
: > national grid distributes at a HIGH volts/watts and
: > then converts down to *lower* volts/watts (ultimately
: > to 240v, typically, for
: Completely different, and irrelevant to this discussion,
: scenario. High voltage is used in the grid to reduce
: resistive losses.
Not at all irrelevant, to your claims that one can step-up
voltage/current without penalty.
: > It could be done but it would actually be more dangerous,
: > one of the reasons why the USA use duel phases @ 110v
: > (with a centre tap return) to allow the provision of a SP
: > 220v high power supply for cookers. water heaters etc.
: WTF has this got to do with CDUs and point motors?
It doesn't, directly, it was in reply to your claim that one can
magically get more current from using less, if that was possible
don't you think that every electricity utility company in the
world would be doing so! Lets use another example, most older CRT
TVs and computer monitors have very high power consumption
figures, not because the electronics on the PCB need such a
supply but because the CRT needs a high voltage/current supply
that is produced by stepping-up the input 240v supply, you seem
to be claiming that one could design a CRT PSU that provides the
same constant high current output power supply but which draws a
very much lower current on the intake side - no doubt you have
also been out to another one of your free lunchs today...
: > You're missing the point MBQ, but no surprise there,
: > yes these units exist but they actually draw more
: > current *for the same work*, OR, draw the same
: > current but do less work.
: Or vary the recharge time.
Only if you trade reliability/usability, I want to be able to
switch more than a couple of points sequentially at a time, as do
most I suspect! Yes your suggestion might work in a route setting
matrix, given a suitably large CDU but even then one would have
to wait before switching the next route!...
I wouldn't call it a trickle charge. The circuit is designed so that
the current is sufficient to recharge the capacitor in time for the
next operation. That minimum re-use time will be a design parameter.
I don't. All I have claimed is that the recharge time can be longer
than the discharge time. It will still "recharge in seconds".
Use of "trickle charge" is *your* terminolgy. It suggests too low a
recharge current, to me, so I don't use that term.
Charge or discharge current? Fixed by whom?
The time before the CDU can be used again is a design parameter, and
needs to be less than the time between operator interaction with the
control panel. Once that is determined you can calculate the maximum
charge current required in order to ensure the CDU is recharged
So why bring it up?
By this point the discussion had moved on to a general discussion of
CDU operation. The OPs situation is not relevant to that discussion.
It's called thread drift.
I made no such claim. [see below]
Again, I made no such claim. I claimed you can get more *energy* in
the discharge pulse by charging the capacitor to a higher voltage. The
higher voltage will result in a higher discharge current (simple ohms
law). There will be penalties in the form of changes to the required
current and/or charge time.
If you are so clever, do a SPICE simulation of a voltage doubling CDU
and see how it works.
Polytechforum.com is a website by engineers for engineers. It is not affiliated with any of manufacturers or vendors discussed here.
All logos and trade names are the property of their respective owners.