Coreless motors weak at low rpms on PWM controllers

[JeffB]

Hey guys,

  Question for those more knowledgeable than I...

I've been using some coreless motors (8mm and 10mm) in a few custom mechanisms the last six months or so.  I test run them using a home built DC throttle that uses a simple ebay sourced PWM control and a DPDT switch wired for reversing.

Even in smooth running mechanisms, I've noticed that the motors don't appear to have a lot of power (even with considerable gear ratios/compounding) at low rpms. 

Is this the result of the PWM control voltage being mostly "off" (very short duration square voltage pulse) at lower speeds? 

Or is this inherent in coreless motors? 

I was always under the impression that coreless motors were more powerful than iron core motors, but could be wrong. 

I don't see the problem on a rod driven diesel mechanical locomotive I built.  It uses a similar gear ratio and coreless motor, but it's running on DCC with an ESU V4 "Micro" decoder.

Just curious...  It's also driving me a little batty. 

One mechanism in particular has 6:1 compounding (with a final ratio of 90:1).  It should be a "torque monster" but at low motor rpms its pretty weak.  The mechanism is super smooth when I pull the motor and run my fingers over the drivers, with no apparent binding at all.

JB

[wvgca]

in my experience, coreless motors have -slightly- more torque than iron core motors at low rpm, and slightly less at high motor rpm [[rpms being roughly equal]

[peteski]

I have never looked into this very deep, but the construction of the coreless motor's rotor is quite different than standard motors with iron-core armature.  That is why DCC decoders (which exclusively use PMW to drive the motor) have to be adjusted to those motors.   Properly adjusted, DCC decoders will produce good amount of torque with coreless motors at low RPMs.

I'm sure the impedance (AC resistance) of the coreless motor winding is quite different than conventional motors, and it is likely that the frequency of the PWM pulses has a lot to do with it.  Possibly the PWM throttle you use is designed to work best with conventional motors (or even LED lighting).

I don't have any personal experience or specific data about this, but I'm not surprised.

[Jim Starbuck]

I have many stock, custom built mechs and remotored locomotives using coreless motors. I use mostly 7x16 mm and stock gearing on the remotor projects.
All of them are controlled by ESU decoders however so an apples and oranges situation here but I haven’t experienced a lack of power in any of them.

Jim

[JeffB]

I have never looked into this very deep, but the construction of the coreless motor's rotor is quite different than standard motors with iron-core armature.  That is why DCC decoders (which exclusively use PMW to drive the motor) have to be adjusted to those motors.   Properly adjusted, DCC decoders will produce good amount of torque with coreless motors at low RPMs.

I'm sure the impedance (AC resistance) of the coreless motor winding is quite different than conventional motors, and it is likely that the frequency of the PWM pulses has a lot to do with it.  Possibly the PWM throttle you use is designed to work best with conventional motors (or even LED lighting).

I don't have any personal experience or specific data about this, but I'm not surprised.

Thanks for the input Peteski...  I'm aware of the structural/construction differences between iron core and coreless motors.  But I'm not super familiar with the electrical characteristics.  I've looked at the output of the PWM controller on an O-scope and "as advertised" it controls speed by varying the pulse width of the full voltage output.

It might just be that being a generic and very simple PWM circuit/control, it doesn't have any logic that allows it to adjust to different motors or make adjustments through feedback from the motor.  Not really sure, but my guess is that it does not. 

I do know that after adjusting the motor related CV's on the ESU decoder, there doesn't appear to be any issues with the coreless powered mechanism I use it with.  So that gives me some hope that the motors will work out in the end.  Just have to get more decoders!

I would use a conventional DC throttle, but my last "Troller" power pack gave up the ghost about a year and a half ago.  Using DCC now days, I have no desire to buy another DC throttle, so I'm stuck with the simple one I have for testing.

in my experience, coreless motors have -slightly- more torque than iron core motors at low rpm, and slightly less at high motor rpm [[rpms being roughly equal]

That was my recollection as well, but it was unconfirmed...  I know that coreless motors with integral gearheads are very powerful (due to gear compounding/torque multiplication), but I wasn't sure on coreless motors without gearheads.

Jeff

[JeffB]

I have many stock, custom built mechs and remotored locomotives using coreless motors. I use mostly 7x16 mm and stock gearing on the remotor projects.
All of them are controlled by ESU decoders however so an apples and oranges situation here but I haven’t experienced a lack of power in any of them.

Jim

Thanks Jim...   As I said in my last post, I don't have any issues with the one mechanism I have that uses an ESU decoder.  Just not sure if it was the motors or the PWM controller.

BTW...  I'm using 8mm and 10mm coreless motors from Tramfabriek, though the one in the ESU controlled mechanism is a Maxon 1217.  But as a comparison, other Maxon 1217's I have don't fair too much better with the PWM controller, and may be worse.  On the one mechanism I'm working on now, it originally had a Maxon 1217, but I rebuilt it with the Tramfabriek 1020 and it wasn't much better.

Jeff

[peteski]

None of the PWM variable voltage motor drivers I have ever worked with (mostly inside DCC decoders) have any sort of fully automatic adjustment of pulse frequency to "tune" themselves to the motor they driving. There are several CV registers in decoders which adjust multiple PWM and BEMF parameters, but those have to be tuned manually.  Some decoders (like ESU) will "auto-tune" (with more or less success) some of those parameters, but only when directed to do so.

I doubt that the inexpensive PWM variable voltage devices have any sort of feedback (BEMF or otherwise) incorporated in them. I suspect they are simple devices, tuned to specific fixed PWM frequency. I also suspect that they are derived from, or geared towards LED dimming circuits, and since LEDs are not inductive loads, the only thing designers are considering is that the PWM frequency is high enough for the human eye not to notice any flickering.

[JeffB]

Did a deep dive on my stockpile of junk yesterday after work and found one last functional (very old) Troller power pack (Transpak 2.5 for anyone who remembers them).

Tested three scratchbuilt mechanisms that had the Tramfabriek coreless motors...  All three performed exceptionally well.  Gobs of torque!  Even at the lowest speed setting on the throttle, steady and moderate finger pressure on one of the drivers wouldn't stop the motor completely, only slow it down. (It must be noted that all three drives have double compound reduction with anywhere from 4.5:1 up to 7.2:1 of compounding).

Aside from gear noise and a tad bit of vibration, which I'm hoping might lessen with DCC control ("silent mode"??), the coreless motor powered mechanisms run great on a conventional DC power pack.

So the problem was that very cheap (about $5 each on ebay or Amazon) PWM module.  They work Ok to get something moving, but they're not ideal to judge ultimate performance.  Now I just have to buy a few more ESU micro or nano decoders.

Thanks for your input guys...

Jeff

[mmagliaro]

You said you put the output of that "cheap" PWM controller on an oscilloscope.  What is the frequency of the PWM output?

Faulhaber (I hunted around the web) explains that their motors require very high PWM frequencies (they suggest 100 kHz) because a coreless motor has such a low electrical time constant. 

Looking  that up...

The electrical time constant Te, is the time it takes the motor to reach 63.2% of its stall current with the armature locked.

So apparently, a coreless motor takes a much shorter time than a conventional motor to hit its stall current when you apply a voltage to it.  I would guess this has something to do with the low mass of the rotor.   Coreless motors can spin up much faster than an iron-rotor motor, and (I'm just guessing here, admittedly), even when the shaft is locked, their magnetic field probably also reaches full power faster.

----
My experience with those sub-$10 PWM eBay controllers is that they run at about 16 kHz at best, and many run at a lower frequency.

Does anyone know what frequency a DCC decoder is capable of sustaining on its PWM output, or if it is intelligent enough to actually change its PWM frequency to suit the motor?

[peteski]

IIRC, the PWM pulse frequency can be around 40kHz (in ESU and ZIMO decoders).  it is usually mentioned in the decoder's literature.

Without revisiting my AC theory materials, I think the iron core is the important factor for slowing the current through the windings.  It has to do with the winding's inductance.  An indictor resists any change in the electrical current (opposite of capacitors).  How much resistance to the change is related to the inductance. Coreless motor's windings have air as their core.  Standard motors are wound on an iron core.  The iron core concentrates magnetic flux which in turn increases the inductance.

[JeffB]

You said you put the output of that "cheap" PWM controller on an oscilloscope.  What is the frequency of the PWM output?

Faulhaber (I hunted around the web) explains that their motors require very high PWM frequencies (they suggest 100 kHz) because a coreless motor has such a low electrical time constant. 

Here's a photo of the O-scope display with the PWM control at one particular throttle setting.

PWMT_2 by jeffB, on Flickr

If the Δt (or T or Period) = 34µS, then frequency = 1/Δt (or 1/T), which would be ~30kHz.  Is that correct?

Jeff

[mmagliaro]

I think that's right.  Is that M 20 uS at the bottom indicating a horizontal 20 uS per division?  That would also be consistent with a
period of 34 uS. 
When you run it at a low speed, so the pulses are very narrow, does it maintain that same frequency, or does it change (it shouldn't change, but it would be good to know.)

While we're at it, what originally led you to say the motors were weak at low rpms on PWM?  Is it just observing how well the engine runs?
And what coreless motors are these that you are using again?

My experience with Faulhabers and Maxons is that once the motor gets over about 200 rpm, good luck trying to stop the shaft with your fingers!  And 200 rpm is a pretty low motor speed when you consider how slow the gear train will bring that down.  I will go try some of these on one of my eBay PWM controllers and see what it does on my coreless motors compared to running them on straight DC.

[JeffB]

I think that's right.  Is that M 20 uS at the bottom indicating a horizontal 20 uS per division?  That would also be consistent with a
period of 34 uS. 
When you run it at a low speed, so the pulses are very narrow, does it maintain that same frequency, or does it change (it shouldn't change, but it would be good to know.)

Yes, here's the evidence.  Three different throttle settings, note the period (time) for each cycle.

PWMT_1 by jeffB, on Flickr

PWMT_2 by jeffB, on Flickr

PWMT_3 by jeffB, on Flickr

While we're at it, what originally led you to say the motors were weak at low rpms on PWM?  Is it just observing how well the engine runs?
And what coreless motors are these that you are using again?

Yes, weak/poor performance at lower rpm/voltage settings.  I had a basic understanding of PWM control and some of the potential issues with it, which led me to think it was the controller after testing multiple different motors and mechanisms.

I tested multiple Tramfabriek coreless motors (816, 1015 and 1020) and a few Maxon 1217's.  Both in mechanisms and stand alone.  The poor torque was more evident in mechanisms.  It must be added that the mechanisms are relatively free running (for double compound reduction at least).

My experience with Faulhabers and Maxons is that once the motor gets over about 200 rpm, good luck trying to stop the shaft with your fingers!  And 200 rpm is a pretty low motor speed when you consider how slow the gear train will bring that down.  I will go try some of these on one of my eBay PWM controllers and see what it does on my coreless motors compared to running them on straight DC.

Yes and no...  Depends on the size of the Faulhaber or Maxon.  I have several Maxon 816 and 1016 motors and they're not very powerful by themselves.  Coupled to a gear head or double reduction gear arrangement (in a mechanism), then sure, they're plenty powerful, but the top continuous speed is pretty low. 

Here's the PWM control I used...

PWM_controller by jeffB, on Flickr

Here's it installed in a simple frame, with a DPDT center off switch set up for reversing...

PWMT_5 by jeffB, on Flickr

Jeff

[peteski]

Jeff, I have a chuckle when I look at your photos of the scope screen showing the 2019 date.  It is like time traveling.

I also see that the pulses are only 5V in amplitude. Shouldn't they be 12V for 12V motors? 5V will not produce full torque.

[JeffB]

Jeff, I have a chuckle when I look at your photos of the scope screen showing the 2019 date.  It is like time traveling.

I also see that the pulses are only 5V in amplitude. Shouldn't they be 12V for 12V motors? 5V will not produce full torque.

Yeah, it's been a while since I built that unit and tested it!  If you want time travel, you should see the size of the O-scope I used, it's about the size of a small suitcase!

I'd only recently noticed issues with running using the PWM controller.

You are correct about the voltage...  I used a 6VDC "wall wart" power supply to power the PWM control for this unit.  I have another one that used a 12VDC wall wart, which was outputting closer to 16VDC.  It didn't produce markedly better results.  It all makes sense now why when running a proven mechanism on my modules using the PWM control, I was getting less than acceptable results. 

On DCC though...  The 1217 Maxon motored mechanism runs great.  I used an ESU V4 Micro decoder, with the CV's set for a coreless (Faulhaber setting) motor.  Whatever ESU does with those settings works great. 

Same using the old Troller power pack...  The torque of the mechanisms with 4.5:1, 6:1 and 7.2:1 compounding ratios is such that even at the lowest speed setting, the motor won't stall. 

Jeff