Author Topic: MOTOR FULL LOAD CURRENT  (Read 29721 times)

Alan

  • Conductor
  • ****
  • Posts: 1073
    • LK&O Railroad
Re: MOTOR FULL LOAD CURRENT
« Reply #15 on: August 02, 2018, 07:27:33 AM »
... With the dual log slope waveform, how much harder is it to work out what the average voltage the motor sees is compared to a square waveform (which is easy)? Given the "1000 speed steps" we have the 2 to 20Hz duty cycle mentioned in the Patent doesn't sound like it'd give that.
- Tim

Finally, a use for that Integral Calculus class you took at university years ago!

calc2.png

Close enough to a sine wave to use RMS or 0.707.

Why do you think we can't have 1000 steps? Cycles per second don't have to be integers. I would lay odds it is actually 1024 speed steps since it is a digital system.

« Last Edit: August 02, 2018, 07:32:57 AM by Alan »
Alan

LK&O Railroad website

When I was a kid... no wait, I still do that. HO, 28x32, double deck, 1969, RailPro

G8B4Life

  • Signalman (Global Mod)
  • Conductor
  • *****
  • Posts: 1247
  • I'll think of a catchy tag line one day
Re: MOTOR FULL LOAD CURRENT
« Reply #16 on: August 02, 2018, 09:01:24 AM »
Finally, a use for that Integral Calculus class you took at university years ago!

You have the wrong idea of how far I made it through schooling.

Quote
Why do you think we can't have 1000 steps? Cycles per second don't have to be integers. I would lay odds it is actually 1024 speed steps since it is a digital system.

I'm not saying that I think we can't have 1000 speed steps.  I was saying that with my very limited knowledge of PWM the face value of the information presented to me made it sound like it was not possible given the duty cycle range, not that it wasn't possible.  Now that I know that cycles per second doesn't have to be an integer (I've never seen any reference to it being anything but an integer in any PWM related information I've ever looked at before, it's always shown and written as whole numbers) it now makes sense that there can be 1000 speed steps within the given cycles per second range. In my effort to understand this I've probably got a bunch of other things wrong so I'll just stop now.

- Tim

Alan

  • Conductor
  • ****
  • Posts: 1073
    • LK&O Railroad
Re: MOTOR FULL LOAD CURRENT
« Reply #17 on: August 02, 2018, 09:58:58 AM »
Neither frequency nor pulse width are limited to integer values. Either or both can be real numbers.

Think of it this way... FM radio signals are similar to PWM ( I said similar  :) ) except sine instead of square. The FM signal is modulated to represent the audio spectrum (or at least some of it). In this case there is a near infinite number of possible frequency divisions. It is just a matter of how many places to the right of the decimal point you can measure.

Clearly this is the case with RP. As you note, with a bandwidth of only 18Hz the "steps" must be real numbers else with integers we would be limited to 18 speed steps. Which BTW sounds eerily similar to the old 14 step DCC.

As this conversation drifts deep into the weeds it is worth mentioning there are two controlling parameters to square wave PWM when also used to convey control signals (DCC track) - pulse frequency and pulse width (percentage). Obviously at 0% and 100% pulse width there is 0 frequency thus no control signal conveyance. But for everything in between the pulse width conveys digital info while the frequency and width combined determine the total power delivered over a given time period. At the motor output of a module we have no need to convey information so frequency and/or pulse width modulation will do just fine.
Alan

LK&O Railroad website

When I was a kid... no wait, I still do that. HO, 28x32, double deck, 1969, RailPro

Tom

  • Engineer
  • ***
  • Posts: 79
Re: MOTOR FULL LOAD CURRENT
« Reply #18 on: August 02, 2018, 06:14:56 PM »
I am happy to learn that manually setting the MFLC to anything will not harm an HO locomotive motor.

I should say I have had only a very few problems when consisting multiple locomotives, mainly uncoupling in certain circumstances due to one pulling or pushing slightly harder than the other.

Having said this, it is still demonstrable that linked but separated locomotives, mine anyway, will not stay the same distance apart by using the automatic MFLC.  I got all locomotives to run close to the same speed, that of about a 45-50 MPH scale speed, by trial and error starting with 3 locomotives and manually adjusting the closest two and then getting each additional locomotive in sync with one of those.

Having accomplished the above it is still noticeable that some locomotives move sooner and start faster than the others and travel at slightly different speeds at different throttle selections.  Overall I think they run more smoothly together, but would not want to do what I did with too many locomotives.

Most of the complex electrical comments in this thread are beyond my electrical level of competence, which is close to zero.  I still do not fully understand the differences between DC, DCC (never used DCC) and PWR-56.  Can someone provide a short lucid tutorial in laymen’s terms: from above comments, a PWR-56 is DC current, DCC is AC current, and Ring modules will run on both DC and AC powered track.  But do not run ring modules using a DC power pack with the throttle set at full speed.  Never mind for now the signals running through the tracks for DCC control.

Alan

  • Conductor
  • ****
  • Posts: 1073
    • LK&O Railroad
Re: MOTOR FULL LOAD CURRENT
« Reply #19 on: August 02, 2018, 06:47:21 PM »
A picture is worth a thousand words, right? Perhaps this illustration from an old thread will help you.

The top row illustration is a conventional power pack. Notice the constantly fluctuating bumpy output. That is bad for modules. Hence the recommendation not to use power packs with LMs.
The middle row illustrations are DCC at various points in the current path. Left = signal on DCC powered track, Center = rectification inside the module, Right = smoothing inside the module.
The bottom row illustration is the output of a PWR.

Yes, the horizontal axis for zero volts is missing on the last DCC graph. My bad.

waveforms.png
« Last Edit: August 02, 2018, 06:54:34 PM by Alan »
Alan

LK&O Railroad website

When I was a kid... no wait, I still do that. HO, 28x32, double deck, 1969, RailPro

KPack

  • Conductor
  • ****
  • Posts: 787
Re: MOTOR FULL LOAD CURRENT
« Reply #20 on: August 02, 2018, 10:37:48 PM »
Regulated DC for the win!!!

TwinStar

  • Conductor
  • ****
  • Posts: 513
  • Modeling a 1961 Rock Island Twin Star Rocket
Re: MOTOR FULL LOAD CURRENT
« Reply #21 on: August 02, 2018, 11:38:30 PM »


Having said this, it is still demonstrable that linked but separated locomotives, mine anyway, will not stay the same distance apart by using the automatic MFLC.  I got all locomotives to run close to the same speed, that of about a 45-50 MPH scale speed, by trial and error starting with 3 locomotives and manually adjusting the closest two and then getting each additional locomotive in sync with one of those.

Having accomplished the above it is still noticeable that some locomotives move sooner and start faster than the others and travel at slightly different speeds at different throttle selections.  Overall I think they run more smoothly together, but would not want to do what I did with too many locomotives.



Tom:

I think you are completely misunderstanding the key difference between load sharing and speed matching. I wouldn't expect any uncoupled and linked locomotives to ever maintain a given distance between them. Real locomotives don't start in uniform unison. They bang and clang against each other until they get a 'feel' for who is doing what. Speed matched RailRro locomotives, unless I'm completely in error, is just wrong. The locomotives need to be coupled to each other, they need to be linked, and they need a good MLC to operate correctly. If your lead Atlas unit is dragging at 20% power it's going to communicate to the trailing P2K to give a little more. And similar adjustments will be made throughout the power band. The locomotives need to feel, I think through BEMF, and communicate to operate correctly.

I might be wrong but I think the rest of us are running RailPro correctly while you're trying to DCC Speed Match them. More informed comments will be coming from others shortly.
Jacob Damron
Modeling late 1950's Dallas Union Terminal in Free-mo+ modules

Texas Railway Modeling and Historical Society trmhs.org
trmhs.org

Alan

  • Conductor
  • ****
  • Posts: 1073
    • LK&O Railroad
Re: MOTOR FULL LOAD CURRENT
« Reply #22 on: August 03, 2018, 06:47:23 AM »
I concur with the group - speed matching and load sharing are two different animals. The module knows the load because it can measure motor current. It has no way of knowing speed. There is no rotation sensing of motor or wheels going on. No speedometer.

Although........... at 10:15 in the video. I suspect prior to making the video the locos were speed matched like Tom's locomotives. Marketing, not to be confused with reality.  :P

Alan

LK&O Railroad website

When I was a kid... no wait, I still do that. HO, 28x32, double deck, 1969, RailPro

Dean

  • Conductor
  • ****
  • Posts: 211
Re: MOTOR FULL LOAD CURRENT
« Reply #23 on: August 03, 2018, 09:41:04 AM »
Almost all my RailPro engines are in 4 or 5 unit consist. When they first start moving there is a few seconds where they appear to 'hunt'. After that, they all run with tight couplers. The controller needs feedback before it can start controlling the engines. Running them uncoupled is fruitless. The current draw per engine is so low it would be extremely hard to control them.
If you have engines that have large differences in full load current, one or the other could be out of the control range of the controller. My Atlas Trainmasters and Stewart 'F' units are that way. On top of that, the Stewarts are 20 years old and the Trainmasters are ~5 years old. I can run them together but it seems that it takes longer to get them to settle down. Slow speeds are worse. Once they come up to speed they run fine.
The engines that always run together I have tuned to make them run even better. I set the accel rate and the decel rate the same on all engines. I found through some experiments the controller overrides these settings but having them set the same makes them easier to control...my opinion. I make sure the top speed and current settings are all the same. But, these engines are the same make and model and the same age.
 
Dean

G8B4Life

  • Signalman (Global Mod)
  • Conductor
  • *****
  • Posts: 1247
  • I'll think of a catchy tag line one day
Re: MOTOR FULL LOAD CURRENT
« Reply #24 on: August 03, 2018, 10:08:56 AM »
I think Jacob might have it with the understanding of load sharing and speed matching. RailPro is designed to share pulling power when MU'd, not to go at the exact same speed as each other (though they go roughly at the same speed when MU'd). As someone previously mentioned this is likely why load sharing doesn't work as well with MU'd light engines as opposed to hauling a train, with light engines there is nothing to share as the locomotives are each pulling only themselves.

I pulled this info (tidied it up a bit) from the Patent tonight. It relates to load sharing.

"Direct bi-directional RF capability allows Multiple Unit consisting of locomotives that can be controlled as one while being coupled together. With that said, the lead locomotive can be using open loop control with just a reference command signal to adjust to. The locomotive motor power can be monitored by the control module in the locomotive and sent to follower locomotives. All follower locomotives would be configured for closed loop power regulation to match the lead locomotive. All locomotives in the consist will then pull with the same power and self regulate for variables that change load on the consist; such as overcoming static friction during movement initiation, trailing model railroad cars forming a train that maneuvers through vertical and horizontal curves, change in track voltage at various locations of the layout, etc. The designated lead locomotive can transmit the information directly to the followers, transmit to them through repeaters, or transmit the info back to a controller that can then pass the information to the followers."

and

"The power values of the motor loads are defined or measured by a motor current. One of the controller and the at least one control module execute a predetermined logic including at least one of a drop power calculation by way of multiplying motor current squared by resistance and motor losses represented by power loss in windings of the motor. At least one control module is configured to use the full load power value at full slip of a respective vehicle as a close equivalent of a full power value of a respective motor. At least one control module is configured to receive a preselected power value so as to maintain tension between couplers of the at least pair of rail vehicles."

and

"When considering the static friction of a consisted, multiple locomotive train, getting it started in motion takes substantially more power than it does to keep it moving at a specified speed. With that, an advantage exists when using bi-directional communication between the locomotive control modules and sharing the power value between the consisted locomotives with only one known parameter stored in memory, that is the full power value at full slip. Full slip is defined when maximum control system voltage is applied to the locomotive motor through the control module and it is held from movement with only the locomotive weight affecting the vertical load on its wheels. When the locomotive is in the "full slip" state, a motor "full power value" is stored in memory thus characterizing that particular locomotive motor. When used in practice, the designated lead locomotive would be sent a reference command signal from a controller in which a locomotive motor power value would then be measured from the designated lead locomotive's motor by the control module and relayed to the follower locomotive(s). By doing this continuously, the follower locomotive(s) will follow the lead locomotive's power reference and automatically regulate to share load about evenly through the breaking of static friction, curves, turnouts, and any other cause of load changes affecting the model train. As an example, when the lead locomotive is sent a reference command signal to begin movement, its motor power will increase until static friction is broken, then the motor power will reduce in order to maintain a steady roll. Since the follower locomotives are continuously receiving the lead locomotive's motor load value, they too are continuously adjusting for a good match in pulling power. This match is simply a percentage of full motor power for each individual locomotive as defined when they had their full load motor value stored in memory during their full slip calibration. This series of control steps allows for better and more accurate control during very low velocity train movement as well as it self-compensates for consisted locomotives with different motor characteristics, differing gear ratios, differing internal and external drag resistances, etc. In cases where the locomotives are very closely matched and couplers between consisted locomotives are loose, a slightly modified motor value command signal can be transmitted to the followers in order to keep the couplers tight and sharing load as evenly as possible."

I'm not so sure that Ring "speed matched" the locos in the video. I could be wrong but I imagine that Ring would not have bothered going through that lengthy rigmarole to try and speed match them using the MFLC.

- Tim

KPack

  • Conductor
  • ****
  • Posts: 787
Re: MOTOR FULL LOAD CURRENT
« Reply #25 on: August 03, 2018, 11:37:45 AM »
Thanks for the info there Tim.  That clears up one of the questions that I had regarding the locomotive modules communicating with each other.  Initially I thought the lead locomotive communicated with the controller, and then the controller sent the signal out to the followers.  It looks like that is only one of the paths, and that the locomotive modules communicate directly with one another.  Love it!

-Kevin

Alan

  • Conductor
  • ****
  • Posts: 1073
    • LK&O Railroad
Re: MOTOR FULL LOAD CURRENT
« Reply #26 on: August 03, 2018, 11:45:08 AM »
Quote
As an example, when the lead locomotive is sent a reference command signal to begin movement, its motor power will increase until static friction is broken, then the motor power will reduce in order to maintain a steady roll.

Confirmed. I saw this on the scope. Anytime you move the throttle off of 0%, even if you only move it a little bit, say to 5%, there is an immediate huge current spike that lasts only milliseconds. Following the spike there is relatively high current that quickly trails off. Reminds you of the lights in the garage when the compressor kicks on.
Alan

LK&O Railroad website

When I was a kid... no wait, I still do that. HO, 28x32, double deck, 1969, RailPro

Tom

  • Engineer
  • ***
  • Posts: 79
Re: MOTOR FULL LOAD CURRENT
« Reply #27 on: August 04, 2018, 10:48:54 AM »
Alan,

Thank you for the great picture.  I am assuming that “RECTIFIED FILTERED DCC INSIDE THE MODULE” is, at least roughly, the same as “REGULATED DC ON THE RAILS AND INSIDE THE MODULE.”

Since I have no familiarity with DCC, one last question: is a DCC decoder doing essentially the same thing as a Ring Module (the middle row) for power, in addition to picking up command signals from the rails instead of over the air?

Alan

  • Conductor
  • ****
  • Posts: 1073
    • LK&O Railroad
Re: MOTOR FULL LOAD CURRENT
« Reply #28 on: August 04, 2018, 10:55:11 AM »
Since I have no familiarity with DCC, one last question: is a DCC decoder doing essentially the same thing as a Ring Module (the middle row) for power, in addition to picking up command signals from the rails instead of over the air?

Yes.
Alan

LK&O Railroad website

When I was a kid... no wait, I still do that. HO, 28x32, double deck, 1969, RailPro

TwinStar

  • Conductor
  • ****
  • Posts: 513
  • Modeling a 1961 Rock Island Twin Star Rocket
Re: MOTOR FULL LOAD CURRENT
« Reply #29 on: August 26, 2018, 12:56:01 PM »
Tim:

Where did you find the patent info? I'm got several Free-mo guys trying to tell me there's no way Ring can have no 'radio interference issues' like Digitrax. I've tried to explain how I've stood next to a Dgitrax duplex user who had lost their signal while my RailPro was running fine.

Jacob Damron
Modeling late 1950's Dallas Union Terminal in Free-mo+ modules

Texas Railway Modeling and Historical Society trmhs.org
trmhs.org