is there a maximum voltage that locomotive motors are designed to?
for dcc, or any pulse controlled throttle, this will be the amplitude of the pulses, even if the locomotive is set for slow speed.
greg
Yes, the NMRA standard for DC was set at 12 volts for HO and N. With DCC the peak amplitude of the PWM motor drive output will be near the peak track signal amplitude less a couple of didoe drops in the H bridge drivers, so 15V tot he track is common for HO on those systems that offer an option, 12V is commonly used for N scale, and it’s pushed to 18V for O and G. Other systems don’t have a specific setting but you may have access to a potentiometer to fine tune the track voltage. NMRA DCC specs state the decoders have to take 22 volts, many are listed as handling up to 25 to give some cushion but others follow the spec closely and are damaged if the voltage to the track goes above 22.
The only system in the past known to put too much voltage on the track was the Atlas Master/Lenz Compact when used witht he recommended power supply. A string og back to back diodes was used to reduce the voltage, which you could build yourself. MRC sold a premade one, so perhaps one of their older systems also tended to be too high a voltage to the track - probably that honking 8 amp booster they made.
–Randy
Randy, that NMRA standard voltage is RMS, 70.7% of the VPP that the OP asked about.
I suspect that the coil wire insulation of our little motors will take upward of 50VPP before failing, but the motor would fail due to overspeed long before reaching 33V RMS.
High peak, low RMS (as in variable-pulse DC throttles) which cause the motor to develop high torque at low RPM will also cause motor overheating if run much over 18VPP. The prototype assigns 1-hour ratings to DC traction motors. If anyone wants to assign 5-minute ratings to the multiplicity of model traction motors, I would like to see the research. (My own operations involve start-to-stop runs of less than 5 minutes.)
If in doubt, err on the side of caution. None of my power supplies can deliver more than 18VPP, although it required me to install a couple of extra diodes to tame my ancient Tenshodo after I swapped a silicon bridge for the original selenium stack rectifier.
Chuck (Modeling Central Japan in September, 1964)
i’m thinking about building a pwm throttle reusing some old wall supply. i believe i can build a circuit that can handle voltages a lot higher than the motor can handle.
but i’m thinking about good low speed performance. I’m assuming that even though the pulse duration may be small, low speed performance will be better the higher the supply voltage. If this is true, then using an 18v dc supply would be better than a 12 v supply, while either supply would be good for higher speed performance. is this true?
If the 12V rating is the after rectification and filtering, then it will liekly be too low once fed through the pulse generation circuitry. If that’s 12V AC then after a filter stage you’ll have a higher voltage and it should be good. 15V is probably the sweet spot, 18 will likely result in peaks that are a bit too high. Not likely to damage the motor other than perhaps a particularly sensitive one, sinc eunless you are also building in things liek torque compensation the pulse frequency will probably remain low which is death on coreless can motors. High frequency PWM without compensation results in poor slow speed performance - it’s all a trade off.
–Randy
If by an 18vDC supply you mean a wall wart meant for charging 18 volt cordless tools, be aware that the no-load RMS voltage pushes 20V, and the VPP is about 1.5 the maximum recommended for our nominal 12VDC (RMS) motors. OTOH, if you have a readily-available supply of cheap motors…
Even the 12vDC battery chargers put out more like 14 V RMS and just under 20 VPP.
I sometimes assemble my own power supplies. Simply stated, for powering motorized rolling stock I wouldn’t touch an 18V battery charger with YOUR 3 meter pole! (I do use one to power some recalcitrant switch machines - twin-coils, momentary contact, operating on half-wave DC.)
Chuck (Modeling Central Japan in September, 1964)
i’m referring to voltage across an operating motor
can you explain torque compensation, and how freq affects performance?
Around 12 VDC is what many are designed for.
Below is a link with a lot of good info.
http://home.cogeco.ca/~rpaisley4/CircuitIndex.html#Throttles
I suspect you are not sure what you are doing with the way the question is asked.
I made one some years ago by Thorne from from a MR article. The Pacematic Throttle. I used a LM555 IC for pulses.
Rich
i believe this page, http://www.members.optusnet.com.au/nswmn1/BEMF_dither_comp.htm,[](http://www.members.optusnet.com.au/nswmn1/BEMF_dither_comp.htm) explains the noise and torque effects of PWM frequency, and how some randomization of the pulses helps minimize noise to allow use of lower frequency longer duration pulses to improve low speed performance. It would be nice to see the algorithms for this in the decoder controllers.
i think this page, http://home.cogeco.ca/~rpaisley4/Throttles.html, was being suggested. It shows various circuits but I didn’t see much discussion about the designs. The Variable Width Pulse Type (VWPT) Throttle seems the most refined. (I would have thought that there should be a reverse biased diode across the output, instead of or in addition to the 5uf cap, to allow current to continue to flow when the pulse is turned off).
It seems that pulses at about 4 times the wheel rotation of the actual locomotive, with some momentum, would be most realistic. This would require adjusting the pulse rate as the locomotive speed changes. This would make use of the pulse rate adjustment on the VWPT throttle. (I believe some decoders have axle rotation sensors).
it would be interesting to know what the best waveform shape and amplitude is for low speed performance. Maybe something more like a saw tooth with a fast attack and slow decay. Does anyone know?
thanks for info
Starting at the basics, on straight DC, a permanent magnet motor’s RPM is controlled by the voltage across the motor. For most motors, the voltage to RPM relationship is fairly linear until approaching the maximum RPM. Model railroading takes advantage of this relationship by very simply controlling the speed of our trains through varying DC voltage.
At very low motor RPMs there are several physical phenomena which make simple voltage variations less than ideal. Motor cogging is caused by the uneven forces of attraction as the motor rotates. The science experiment 2 pole motor is the ultimate in cogging - it can stall, and not be able to restart regardless of voltage. Current required for low RPM running is usually lower than the current needed just before rotation because of the need for more force to overcome starting friction. As a result, most small DC motors have a minimum RPM in the low hundreds.
Pulse trains can be used to lower the minimum RPM. A very narrow, high amplitude pulse can give lower speeds than a wider or smoother pulse. And the lower the pulse frequency, the slower the motor can be “kicked” with the pulses. 60Hz was a common pulse frequency because it was easily generated from house mains, no oscillator needed. The problem with the low frequency pulses is that they cause vibrations within the motor in the audible range, often know as growl.
The other problem is that pulses generate much more heat in the motor than filtered DC. A coreless motor does not have an iron armature to help it dissipate heat. Narrower, higher amplitude pulses generate more heat than smoother, smaller amplitude pulses. The heat problem does not go away with PWM until the pu
Marcus’s site describes the issue of torque compensation pretty well. The variable pulse width method is probably the most advanced DC motor control system (at least that’s afforable by mere mortals). It’s both highly effecient ont he driver side and also allows for excellent motor control. The trade-offs are discussed in some of the linked articles, and can usualyl be worked around. The efficiency reason was used for DCC decoders, allowing some awfully tiny chips and transistors to stand 1.5-2 amp loads. But it’s also used in top notch DC controls like the Aristo Train Engineer.
–Randy
I think the OP is interested in the pulse voltage in a DC power pack, not PWM used with decoders. If it is really a PWM DC power Pack, I would be interested in seeing what the schematic is like,
Generally the DC power packs and DIY DC power packs had the pulse, ONLY, for kick startling the motor. Early ones picked off the low voltage AC as a form of pulse. Eventually, some used a transistor multi-vibrator to develop pulses. Many back years ago had open frame three pole and open frame five pole motors.
Some like mine used a LM555 timer IC for pulses. The pulse level was usually at the level of the DC in the power pack which could be between 12 volts to maybe 19 volts depending on the DC supply.
The pulse was fixed width, again, only for kick starting the motor. Mine could have had pulse width adjustment for fine tuning of the speed but the motor was primarily controlled by the DC voltage. I did not have room for another pot. The pulse was usually always present.
The filter capacitor in these had the speed pot directly across the cap so the voltage
I think this is fairly consistent with Rich’s reply. In his tat-iv article from 1969, linn wescott reported that 2.5ms 18v pulses 40 pulse/sec substantially improved low speed performance, but that it should quickly be removed and replaced with DC because above low-speed, pulses resulted in overheating and noise. Even though the maximum DC voltage is 12V, he said the pulses should be 18V.
my guess is that PWM is used in decoders because the decoder processors, such as PICs, have built in PWM outputs and PWM, as Randy said, PWM doesn’t require high-power components.
i’m surprised that Wescott’s observations and designs still hold up today.
Skipping over all the electrical engineering data, I use the Aristo Craft Train Engineer throttles and have them set in the Pulse Width mode. I do not know the exact frequency specs of the TE, but I do have that info stored away some where.
That said, I power my Trailn Engineer throttles with 13.8 volt regulated and filtered power supplies typically used to power CB radios from 120VAC sources.
They provide a final maximum track voltage 13.5 VDC at full throttle. This voltage seems to provide nearly perfect prototypical top speeds with a vary large precentage of the newer (last 20 years) models on the market today. This allows full use of the throttle range during operation and allows very fine speed adjustment.
Example - Original early run Proto2000 EMD E units, typical prototype top speed, 88 mph, on 13.5 volts, 90 smph.
Myself, and several other modelers I know who use the Train Engineer throttles have never experianced any motor overheating issues using the the TE in pluse width mode at these voltages.
NMRA guidlines suggest that 12 volts should provide full/top speed operation.
Sheldon
Keep in mind PWM motor control is NOT the same as injecting pulses of AC on top of the DC variable track voltage. Pulse power was originally accomplished by blending in some half wave DC to produce 60hz pulses. This ideally should fade out as the throttle is increased as it can cause overheating. Once transitorized controls became practical, experiments were done by Westcott and others in various pulse frequencies and duration, tot he point of some of the more sophisticated power packs had easily accesible adjustments for these settings, since settings for best performance tend to vary among different motors and differences in the drive trains.
Still none of this is PWM, not even close.
With PWM, there is a constant amplitude pulse of varying duration, or width. This is NOT AC, it doesn’t change poalrity around a common references. As the pulse duration increases, the area under the curve increases - this is the mean voltage seen by the motor. The longer the “on” time, the higher the voltage and the faster the motor turns. At full throttle there would be no or very little “off” time and the output is flat DC at the pulse amplitude - so you don;t want the output to be 30 volts, or when runnign at full throttle, your loco motos would see 30 volts.
One way to think of it is like having a a battery attached t the track via a normally open pushbutton switch. If you tape the switch, the trainwill move - tap it fast enough and any jerks in motion will not be seen. The longer youhold it closed on each tap, the faster the motor will go as it tries to speed up to whatever speed it iwll turn on the supplied voltage. Don;t hold it down as much and the motor will be unable to rech as high a speed, the voltage it sees is less than the full supply voltage., Hold th ebutton down without tapping, and you get full power, 100% on time, or 100% duty cycle, and the full supply voltage is delivered to the motor.
&nbs