Interesting Powerpack Differences (DC)

[tehachapifan]

I recently purchased a newer power pack to replace one that was about 20-25 years old. I replaced it because I was concerned it was playing a part in a couple motor and aftermarket light module losses. All my locos' motors seem to love the new pack as all my engines seem to run much quieter and smoother with it. My Kato P42's in particular had a very audible hum using the old pack and they are silky quiet with the new pack. On the down side, all my locos now start to move well before the lights come on and the lights don't get as bright with the new pack. Not sure of the physics at play here. Ideas?

[robwill84]

What kind of power pack did you have previously? Back when "pulse power" was new, some power packs had it as a permanent feature, without a switch to turn it off. This was good for older, higher amp draw motors that were hard to get started, but newer motors don't benefit as much from this feature.

[peteski]

Really old power packs were simply a transformer with a full-wave rectifier, and a rheostat (variable power resistor) wired in series with the rectifier output.  That resulted in a unfiltered (pulsing at 120 Hz) DC being feed to the locomotives.  Pulse power was an option which removed a pair of diodes out of the circuit, resulting in a half-wave rectification.  That feed a pulsating DC (at 60 Hz) to the model.  The buzzing would have been audible either way.

Next generation replaced the rheostat with a transistor circuit (transistor is really an electrically-controlled variable resistor).   But the voltage supplied to the model would still have the same characteristic as in the previous example.

Newer throttles might be using a filter capacitor installed on the output of the bridge rectifier. It smooths out the pulsating DC voltage to more closely resemble pure DC voltage (less, or no buzzing).  Size of the capacitor (its capacitance) determines how well it filters the pulsating DC (which can result in more or less audible buzzing of the motors).  If those throttles have a pulse setting, that is done by injecting the AC pulses into the transistor circuit which controls the output voltage.  If the pulse is turned on, these will also cause buzzing of the motor.

[Chris333]

Off topic.  I surprised how much used DC packs go for considering everyone is DCC now.

[mmagliaro]

Your description fits in well with what I would expect from an old pack vs a modern one that doesn't use a rheostat.
Your old pack probably did put out half-wave pulses that were lighting up your headlights, but with a low enough
RMS (average), they probably didn't move your motors yet.
Now, you are playing nice with a pack that has no pulse, or gentler pulses.  So the lights done come on as early.
And I'm guessing that your engines are good enough (good motors and smooth drivetrains) that they start moving
just fine without pulses and never needed them.

It would help to know what kind of pack that "old" one was.

Personally, I use a home-made throttle based on simple discrete transistors and a capacitor filtered supply (like Peteski
described).  BUT, it feeds some of the half-wave pulses into the transistor chain after clipping and shaping them
a little with a resistor and capacitor.  The end result is a gentle ripple that rides on top of the smooth DC.
I haven't run a motor yet that doesn't run better on this than on smooth DC.  Even the engines I have with
beautiful stainless coreless motors and gearheads will start and run just that much slower and better with the pulses
enabled.  And because they are not brutish pulses, there is no heat or motor damage. 
I went into this long screed to make the point that although pulses have gotten a bad rap over the years,
when they are done right, they are wonderful.

Chris, I suspect the reason that DC throttles still bring high prices is twofold.   For one, the remaining people who
use DC really want DC.  I mean, by this time, anyone who has not converted probably never will.  They have
made their choices and they want what they want.  So they will pay whatever they have to for a good
DC throttle.  Since there are fewer options in throttles these days, but there is still some demand, the price goes up,

And the second reason, heh, is obviously that not "everyone is DCC."  There is a strong market for DC
stuff, even if it's small.   

[Kev1340]

Off topic.  I surprised how much used DC packs go for considering everyone is DCC now.

You think? Not everyone has been sucked in. For me, and plenty of other, the benefits of DC outway the expense of DCC.

Cheers,

Kev

[mecgp7]

So, which power source did you have and what did you replace bit with? I am in the market for new power sources too.

[randgust]

My first experiments with using DCC have not gone well at all; bad enough that the decoders were now removed and locomotives converted back to DC.  The dramatic increase in wheel dirt, erratic performance, and 'spontaneous rebooting' were enough to take away my enthusiasm if not set me against it entirely in the future.   I've never, ever, ever had these kinds of problems with DC and any power supply I've had.  And the tests using 'same locomotive on the same track' removed all the variables down to strictly power supply and decoder.  Nothing else.  And I went back to the deocoder supplier for diagnosis, and even replaced the decoder once under warranty; and I'm not kidding, the problem was finally down to 'your locomotives are simply too light to work reliably in DCC - put the decoder in a different locomotive'.   That's not a solution for me.   I'm a tiny locomotive guy and have quite a fleet built up now.

So, as a DC guy, I've had multiple power supplies, including the ones I've built myself and several transistor throttles.  As we find out what you had before and what you've got now, there will be an answer.   I quit using non-transistor 'plain' packs back about 1978, but still find that they run some of the Faulhaber motors and other tiny gearmotors better - if you can get the voltage and control range right.  DC is not DC is not DC either.

[LV LOU]

You think? Not everyone has been sucked in. For me, and plenty of other, the benefits of DC outway the expense of DCC.

Cheers,

Kev

No interest in DCC here..

[DKS]

The dramatic increase in wheel dirt, erratic performance, and 'spontaneous rebooting' were enough to take away my enthusiasm if not set me against it entirely in the future.

Same here.

These days, my preference is PWM (pulse width modulation). For those unfamiliar with this term, it's a method of chopping DC into pulses (but not "old school" pulses) of varying widths to control motor speed. Sophisticated supplies allow one to vary the frequency of the pulses, as well as mix them with pure DC in varying amounts, to produce the optimum power for a given motor.

[tehachapifan]

Thanks for all the replies. I replaced an MRC Tech 2 2800 dual power with an MRC Tech 4 280 dual power. One thing I should note is I'm not using the power pack to control 2 separate blocks on the same track, but rather 2 separate tracks entirely. I've seen some postings that the 280 is supposed to be used for separate block control with a "common rail" but nothing that came with the power pack itself mentions anything about having to do this whatsoever.

[Chris333]

I have a tech 2 2800 that I use all the time on 2 separate mainlines. I also have a Control Master 20.

http://forum.atlasrr.com/forum/topic.asp?ARCHIVE=true&TOPIC_ID=50606

[peteski]

I've seen some postings that the 280 is supposed to be used for separate block control with a "common rail" but nothing that came with the power pack itself mentions anything about having to do this whatsoever.

If your tracks are totally separate then this is irrelevant.

[randgust]

We may be talking apples and apples here.   I've got two hand-held throttles, one very old and rebuilt Cama train control transistor one that was 'buzzy', so I reduced the input voltage with a rheostat, and discovered that changing that input voltage to the applied load made an enormous difference; you match power to load; much less for a light switcher and crank it up for a four-unit consist pulling 30 cars.   I've got a pair of PSI 550 cabtrollers, still can be found occasionally on Ebay, same deal works, reduce the input voltage and wonderful things happened, they quiet down, no sparking or track carbonizing, still great control.   I have a straight DC rheostat-only one I built to expiriment on coreless motors.  And have the original MRC rheostats with additional resistors added to get the control range where I want.  The logging railroad module is strictly an old throttleback with additional resistance added.  Very quiet, dependable with small motors and locomotives.

I 'think' that what I've done to reduce the input voltage to the transistor throttles to alter maximum wave voltage is almost what David is doing; certainly sounds like it, and it was one of my better accidental discoveries.  The biggest 'oh, wow' impact using some kind of wave-based DC is that it tends to kick over multiple units in synchronization rather than letting one slip under load like conventional DC.  I can get four or five units with similar gearing to slip at EXACTLY the same time this way, dramatic change compared to resistance-only control.

[tehachapifan]

Thanks for the feedback but I am nowhere near being able to build or modify my own throttles/power packs and am barely following some of the comments. From the best that I can tell with my extremely-limited knowledge of electronics, I am faced with the following choices at this point:

a) Stick with the new pack which makes the locos run extremely smooth and quiet but causes dim headlights on many locos (my Atlas MP15DC's headlights barely come on at all with it and even then only well after these locos start to move).

b) Go back to the old pack which gave me very bright lights before anything started to move but caused the locos to run noisier and somehow less smooth.

c) Try another pack entirely and see what happens.

d) Try adding some more track feeders. I have very few track feeders (ok...I have 1 feeder on this line right now, which is essentially a loop isolated at one point opposite the feeders due to having Electrofrog switches). This never really caused problems on the old pack but maybe it's effecting light brightness....even though the loco speed is where I want it(?).