First, welcome to the list.
> I have a question on Z scale voltages.
Oh boy. Your first question will no doubt have quite a response. <grin>
This is not the easiest of questions. Brace yourself for several different
> I have both Marlkin and MT engines. The documentation
> for both say that you should not exceed 8 volts.
Aaaaahhh. Read that again. MT says not over 10 volts. Marklin, for 24
years, said nothing over 8 volts. Marklin recently brought out a new power
supply and now says no more than 10 volts. But mind you, the new Marklin
pack is not a "DC" pack, its a processed waveform pack.
> Also the MT says not to run them on pulse power.
That is a correct statement of what the label says.
> I own 2 Marklin power packs (the large white ones
> that cost alot of money and aren't as good as some
> MRC power packs that I have for HO). I put a dc volt
> meter on them (actually I tried 2 different ones)
> and both power packs put out a maximum of 12.5
> volts. So what I do is only run them up to 3/4
> maximum. Am I missing something?
One thing at a time.
First, if the power supply in question uses a resistor to vary the output
voltage, and you measure its output with no load, you will see the full
rectified output of the transformer. Why? Because if there is no current
across the variable resistor, you will not have any voltage drop between the
transformer and your point of measurement. It will always show you a false
reading, no matter the knob setting. If you really want to see what it puts
out, put a automotive stop light bulb across the output, so some real current
flows. Then measure it.
If the power supply is a solid state unit, measuring its output may or may
not meaning anything.
The problem is that almost any meter a hobbiest can afford is an RMS meter.
That simply means the meter expects to encounter re-occuring and identical
sine waves and give you a reading on the "heating effect" of that sine wave.
In other words, what DC voltage would be equal in "power" to the AC waveform
you are measuring.
But if the supply puts out a pulse or processed waveform, and all bets are
off as far as the meter telling you anything meaningful. I have a chart I
picked up from an engineering magazine that gives you a multiplier if all you
have is a "affordable" meter and a strange waveform. The multipliers run
from 0.2 to 3.8 depending on the waveform you think you are connected to.
> Also why does MT say not to run them on pulse (do they have coreless
The fatal enemy of any motor is heat. The purpose of any pulse supply is to
force more enegy into the motor than it would otherwise accept at a given
speed. While this results in a more responsive motor, the result is also
In the case of MicroTrains, their motor is totally encapsulated inside a
metal loco weight. There is no place for the heat to go. I guess is, the
motor runs at a safe temperature margin, but at the top of the range.
Anything beyound that and Poof !! Neither Marklin or MicroTrains has a
thermal path for heat to escape the motor.
So, what should you do.
Well, if you want to be very conservative, stay with pure DC, and insure you
buy a pack that does not put out too much voltage (Marklin or MT). Or epoxy
a "stop" to block the knob of a 12vdc pack so it cannot exceed the rated
voltage. Or put diode pairs in its output leads to "kill" the extra voltage
(that's another email).
If you elect to try pulse, you can use solutions others have found to be
acceptable over the years. Use some caution in taking "advice" and ask more
than a few questions. Its YOUR train that will burn up.
Or you can monitor the temperature of your motor every 30 seconds for an hour
of operation to see if "this loco" and "that kind of pulse pack" is a good
combination. Folks on the Nn3 list write about Marlin motors going up in
smoke in 60 seconds on some exotic pulse-width type pulse packs.
You see a lot about the 2800 packs. Folks I have a high level of confidence
in use them. It is NOT a pulse pack, rather it has an interesting processed
and complex output waveform. But plan to block them at less than full
rotation. At full output, it puts out a full 12.5vdc RMS on the rails. Far,
far too much. Measuring it with a "normal" meter is only accurate near the
high end of the speed control. At the low end, your meter will lie, without
I actually own a "true" RMS meter. One that is not bothered by waveforms.
Thank heavens for oil company surplus equipment auctions, otherwise I'd never
own one ($3,000usd). Internally, it uses the input voltage to heat a
thermocouple, then measures the temperature of the thermocouple and reads in
"volts" of effective DC. If I can get "roundtuit" I'll hook a 2800 up to the
meter and see where the 8 and 10 points are. You are the second person I've
promised that to. Guess I'd better get busy.
So what do I use. I make my own throttles from LM317 integrated circuits.
Pure 8 VDC. Works for me. Very similiar to this circuit (not my web page):
Clean rails, wheels and drivetrains are more important than exotic throttles
in this scale anyway, or at least I think so.
> I installed a Digitrax DCC in one of the
> Marklin engines. When I run it on DCC I set the VMAX
> to 8 volts (but don't know if I have to).
Now I'm jealous. And yes, 8 volts is typical. Remember, DCC puts pulse
power to the motor, not DC. Watch the heat. If you are really into DCC,
some DCC motor overheating discussions can be found at:
and the full decoder project described at:
>dcc decoder </A>
Glad to have you with us.