Side-effects of running massive endbell timing in a low-power application?

[fyrstormer]

I have a Tamiya XV-01 that I built as a FWD car. Because it has limited traction, I'm running a 30-turn crawler motor with the highest gearing that will fit in the car, plus 24 degrees of timing advance, to get the speed I want. (for the record, it still spins the front tires on takeoff.) It works fine, but I find myself wondering: What are the side-effects of running massive endbell timing *specifically* in a low-power vehicle?

As I understand it, the best power efficiency is found at near-zero timing, because each armature coil powers-up when it's in just the right position for its EM field to interact as productively as possible with the permanent magnetic field, resulting in the highest torque per watt -- but not the highest wattage, and thus not the highest top-speed. Advancing the endbell timing powers-up each armature coil before it's in ideal alignment with the permanent magnets, which causes the causes the armature's EM field to deform the permanent magnetic field, pushing it further away from the armature. This temporarily reduces the strength of the permanent magnetic field in the vicinity of the armature, making it easier for the armature to rotate through it at high speed, because a weaker permanent magnetic field can't induce as much reverse-current in the armature coils, so there is less parasitic drag. This reduced reverse-current, in turn, exerts less resistance to amperage flowing through the armature coils (similar to what happens when the motor is stalled), allowing more amperage to pass through, which increases the motor's power consumption, power output, and top speed. But higher amperage also means more heat, and running hot is the most obvious tradeoff for running massive endbell timing.

Having said all that, I'm specifically running a low-power motor in a vehicle much lighter than it was originally intended for, so overheating is not a concern in my specific circumstance. Soooo...what's to stop me from increasing the endbell timing to truly ridiculous settings? I know if I rotate the endbell more than 90 degrees, the motor will just run in reverse (a handy thing to know on the rare occasion that you buy a motor that was assembled incorrectly), but what about running 48, 60, 72, or even 84 degrees of timing? There must be some side-effect of running massive endbell timing, even if the problems of overloading and overheating are eliminated, or else it would be much more common.

Does anyone know what other disadvantages there are to running massive endbell timing, besides the risk of overheating?

 

[EddieO]

In general, many of the standard motor theories in rc motors gets thrown out the window. Most of the engineers will tell you its not possible to do what we do with them.

Zero timing has rarely been the best efficiency on a motor. Typically, at least a few degrees of timing is required to hit it. IN 1/12th scale pan cars, efficiency was everything back in the day of Nickel based batteries. We would spend hours on the dynos setting mod motors to peak efficiency so we didn't dump before the end of the 8 minute race. I think the lowest I ever found for peak efficiency was around 4 degrees, 6-14 was the normal though.

You rotate an endbell 180 degrees to reverse it, not 90....you have to reverse the relationship of the brush to the magnet, at 90, both brushes would be in between the magnets...a motor at 90 degrees will not likely even run, it will draw tons of power very quickly and the ESC will go into shut down mode. 180 resets at zero, and you adjust timing from there....in the opposite direction for a reversed motor.

Way back when, stock motors did not have a preset timing limit like they did in the later years. Companies came out with motors with up 55 degrees of timing (I think there may of been a few at 60)...they all lasted 1-2 runs before either blowing up or needing full rebuilds (comm lathes were very rare at the time). ROAR put the 24 degree max in place, motors last much longer. Slot car motors run up to 55 or so as well, but they achieve it by adjusting the comm during winding and it cannot be changed after, though some have the ability to add 10 degrees or so in the endbell as well.

You can't go much higher than 55 or so as well, because the endbell will more than likely be at the tabs, so it will just fall off. YOu can overcome this by redrilling new holes in the timing ring. We used to do this with the ultrabird 19t at snowbirds so we could put 55 degrees or so in them. One run motors...things needed full rebuilds even if they survived.

You can't really consider a 30t that low of power though, as Stock motors were 27t...even at 24 degrees locked and 7.2 volts, we were still killing them fairly easy by overheating, which either was a new motor or a full rebuild. Now, those were stock arms versus a full stack crawler arm, but its not gonna be the world of difference.

So, as you increase timing
More RPM
More Power
Less Torque
Less Efficient (Typically once past the sweet spot for efficiency)
Less Brakes
More Heat
More Comm wear (typically from additional arching and increased RPM).

Add in the fact when you start getting into truely nutty timing, gearing low enough may be an issue. ESC damage is possible as well...

If you want to go faster that bad....stick with a reasonable timing amount and add more voltage.

Later EddieO

 

[fyrstormer]

In general, many of the standard motor theories in rc motors gets thrown out the window. Most of the engineers will tell you its not possible to do what we do with them.

Conventional wisdom is the bane of the tinkerer.

Zero timing has rarely been the best efficiency on a motor. Typically, at least a few degrees of timing is required to hit it. IN 1/12th scale pan cars, efficiency was everything back in the day of Nickel based batteries. We would spend hours on the dynos setting mod motors to peak efficiency so we didn't dump before the end of the 8 minute race. I think the lowest I ever found for peak efficiency was around 4 degrees, 6-14 was the normal though.

As I understand it, inductive resistance is to blame for this. As each armature coil is powering-up, the increasing EM field generated by the coil temporarily increases the coil's resistance as well; that effect doesn't dissipate until the coil comes up to full voltage and the EM field stops increasing in strength. So setting the endbell for zero timing actually results in a slight negative timing in real life; advancing the endbell timing slightly counteracts this effect.

You rotate an endbell 180 degrees to reverse it, not 90....you have to reverse the relationship of the brush to the magnet, at 90, both brushes would be in between the magnets...a motor at 90 degrees will not likely even run, it will draw tons of power very quickly and the ESC will go into shut down mode. 180 resets at zero, and you adjust timing from there....in the opposite direction for a reversed motor.

Well yes, I was saying if I set the timing to more than 90 degrees then the motor would reverse direction. I wasn't actually saying 90 degrees of timing would work, just that 90 degrees is the dividing line between 0 and 180. If I turned the endbell 120 degrees, for example, the motor would run in reverse with effectively 60 degrees of timing (180 - 120 = 60) in the reverse direction.

Way back when, stock motors did not have a preset timing limit like they did in the later years. Companies came out with motors with up 55 degrees of timing (I think there may of been a few at 60)...they all lasted 1-2 runs before either blowing up or needing full rebuilds (comm lathes were very rare at the time). ROAR put the 24 degree max in place, motors last much longer. Slot car motors run up to 55 or so as well, but they achieve it by adjusting the comm during winding and it cannot be changed after, though some have the ability to add 10 degrees or so in the endbell as well.

Interesting history lesson. What voltage and C-rating were you guys running your stock motors on? I'm surprised that you were able to wear them out in a couple runs when all you had to work with were wheezy little NiMH and NiCd batteries.

You can't go much higher than 55 or so as well, because the endbell will more than likely be at the tabs, so it will just fall off. YOu can overcome this by redrilling new holes in the timing ring. We used to do this with the ultrabird 19t at snowbirds so we could put 55 degrees or so in them. One run motors...things needed full rebuilds even if they survived.

True, the timing ring won't tolerate that much adjustment without modification. Drilling new mounting holes is easy, though, so I wasn't considering that an obstacle to running massive timing.

You can't really consider a 30t that low of power though, as Stock motors were 27t...even at 24 degrees locked and 7.2 volts, we were still killing them fairly easy by overheating, which either was a new motor or a full rebuild. Now, those were stock arms versus a full stack crawler arm, but its not gonna be the world of difference.

Well...30x1 is low-power compared to the 12x2 brushed motors I run in some of my 4WD touring cars and rally cars. Relative to a 55x1 crawler motor, 30x1 is high-power, but relative to a 12x2 motor, 30x1 is low-power. They're actually the lowest-power motors I run in any of my vehicles; the only 55x1 motor in my fleet is the one powering my comm lathe.

Funny you should mention stock armatures; I ran a stock motor once, the kind with the large air-gap between the two halves of the armature, and I hated it. The speed was good, but the braking was terrible. Clearly the air-gap in the middle of the armature reduced parasitic drag and improved efficiency when running, but it reduced braking so much that it was comparable to coasting to a stop with a normal armature.

So, as you increase timing
More RPM
More Power
Less Torque
Less Efficient (Typically once past the sweet spot for efficiency)
Less Brakes
More Heat
More Comm wear (typically from additional arching and increased RPM).

I expected more RPM, more power, lower efficiency, more heat, and more comm wear thus more arcing. Less torque surprises me, though; I thought higher amperage meant higher max RPM *and* more torque at low RPM? At least, that's how it works when you switch to a lower-turn armature. What is it about endbell timing that causes the motor to behave differently than it would if you switched to a lower-turn armature? I also didn't expect braking to be reduced; do you know why that happens?

Add in the fact when you start getting into truely nutty timing, gearing low enough may be an issue. ESC damage is possible as well...

Why would it damage the ESC? If you start with a high-power motor, I could see the ESC being overloaded by higher power requirements from massive endbell timing, but I'm running a 30x1 motor on 2S with an ESC that's rated for a max of 12x1 at 3S, so that's not likely to be an issue.

If you want to go faster that bad....stick with a reasonable timing amount and add more voltage.

Conventional wisdom strikes again. Sure, there are ways that make more sense, but I'm curious about why you can't just keep cranking the endbell more and more to get more speed out of a slow motor.

 

[EddieO]

Conventional wisdom is the bane of the tinkerer.

As I understand it, inductive resistance is to blame for this. As each armature coil is powering-up, the increasing EM field generated by the coil temporarily increases the coil's resistance as well; that effect doesn't dissipate until the coil comes up to full voltage and the EM field stops increasing in strength. So setting the endbell for zero timing actually results in a slight negative timing in real life; advancing the endbell timing slightly counteracts this effect.

Well yes, I was saying if I set the timing to more than 90 degrees then the motor would reverse direction. I wasn't actually saying 90 degrees of timing would work, just that 90 degrees is the dividing line between 0 and 180. If I turned the endbell 120 degrees, for example, the motor would run in reverse with effectively 60 degrees of timing (180 - 120 = 60) in the reverse direction.

it won't reverse at 90...I for giggles tried it on all my motor checkers and even straight to a battery....all of them went into shutdown, motor did not move....even on the battery. But yes, if you went 120 in the correct direction, it would reverse with 60 degrees of timing....


Interesting history lesson. What voltage and C-rating were you guys running your stock motors on? I'm surprised that you were able to wear them out in a couple runs when all you had to work with were wheezy little NiMH and NiCd batteries.

We were still running 1400 SCR NIcads back then...they did not use a C rating, thats a term that was not used by battery companies on Nickel based batteries. C rate did not even show up as battery rating until Lipo...kinda like KV, never used on brushed motors in RC until Brushless showed up.

True, the timing ring won't tolerate that much adjustment without modification. Drilling new mounting holes is easy, though, so I wasn't considering that an obstacle to running massive timing.

Well...30x1 is low-power compared to the 12x2 brushed motors I run in some of my 4WD touring cars and rally cars. Relative to a 55x1 crawler motor, 30x1 is high-power, but relative to a 12x2 motor, 30x1 is low-power. They're actually the lowest-power motors I run in any of my vehicles; the only 55x1 motor in my fleet is the one powering my comm lathe.

Funny you should mention stock armatures; I ran a stock motor once, the kind with the large air-gap between the two halves of the armature, and I hated it. The speed was good, but the braking was terrible. Clearly the air-gap in the middle of the armature reduced parasitic drag and improved efficiency when running, but it reduced braking so much that it was comparable to coasting to a stop with a normal armature.

I expected more RPM, more power, lower efficiency, more heat, and more comm wear thus more arcing. Less torque surprises me, though; I thought higher amperage meant higher max RPM *and* more torque at low RPM? At least, that's how it works when you switch to a lower-turn armature. What is it about endbell timing that causes the motor to behave differently than it would if you switched to a lower-turn armature? I also didn't expect braking to be reduced; do you know why that happens?

Timing is a much different animal in how it changes the motor running than different construction techniques, like winding. Though they can be done in construction, how it functions is just different.

While I am a guilty of saying it too, mainly just because it was accepted as "right", higher turn arms don't really make more torque overall, they simply make more torque per amp. In the case of a crawler motor, keeping that torque in a much smaller power band with the ability to produce it on less amperage is why we mainly use higher turn arms. (trying to keep this as simple as possible, as I am not a fan of getting super technical in posts like these.).

Why would it damage the ESC? If you start with a high-power motor, I could see the ESC being overloaded by higher power requirements from massive endbell timing, but I'm running a 30x1 motor on 2S with an ESC that's rated for a max of 12x1 at 3S, so that's not likely to be an issue.

Excessive amp draw can damage an ESC...plain and simple. Years ago, while at a big race, people were cranking comms left and right. Some idiot went way overboard, cranked it so much he broke the comm lock and tried to run it....his ESC melted down from the amp draw. The motor was drawing 20 amps at 4 volts on our dynos. Went I looked at the motor, he probably had 30 degrees in the comm...plus 24 on the endbell. Was just a stock motor too.

Conventional wisdom strikes again. Sure, there are ways that make more sense, but I'm curious about why you can't just keep cranking the endbell more and more to get more speed out of a slow motor.

Later EddieO

 

[JohnRobHolmes]

Just have a moment here, but there is an easy way to look at timing changes. They effectively adjust the motor Kv via field weakening, and thus torque per amp goes down as speed increases. Since the resistance of the motor is fixed, this means that more heat is created for a given load at specific rpms. This is why very low timing is the sweet spot for efficiency, the amperage needed for torque is lowest, and amps make heat.


The magic combo is really the same as a crawler. Add voltage (or increase kv) and use as much reduction as possible to stay within your speed needs. 27t (2100Kv more specifically) is about where brush (and other copper outside of the coils) losses begin to dominate and torque falls, so if you can increase voltage instead of more timing or lower turns, it will run cooler and longer. The sweet spot is 3s, 6s, and 12s for escs, so using these voltages and working backwards for turn count to reach desired speed tunes the full system for optimal performance.

 

[fyrstormer]

The sweet spot is 3s, 6s, and 12s for escs

I don't understand what this means. Why would *all* ESCs run better on 3S, 6S, or 12S than on other voltages?

 

[JohnRobHolmes]

I don't understand what this means. Why would *all* ESCs run better on 3S, 6S, or 12S than on other voltages?

Mosfets have 15v, 30v, and 60volt ratings. Power capacitors margin of saftey wrangle this into 3s max, 6s max, and 12s max typically. Maximizing a system is strongly tied to using the lowest resistance FET, which is achieved by using the lowest voltage rating possible. I'm talking ideal setups and not what real world may dictate. But going ideal, we would choose voltage based on ESC rating and then work the motor Kv backwards from this to hit desired rotor speed. This is how electric car and bike systems are designed.


In the real world, we have batteries already on the shelf. Kv selections are not always what we want for a specific voltage. So we work with what we have. Maybe you have an ESC that is 6s capable but you run 3s. And then you want more power, so you buy a 4s pack. Life tends to make a mess of ideal situations.

 

[fyrstormer]

"In theory, theory and practice are the same. In practice, theory and practice are different."

I didn't realize MOSFETs were only available with such a limited selection of voltage ratings. That does make it hard to match the MOSFET to a pre-packaged combination of 4.2V cells.

The highest rating on any of my ESCs is 4S. I've never needed more than that. I only have one vehicle that runs on more than 2S, and that's my Summit which runs on 4S. What you just explained about MOSFET voltage ratings makes me wonder how those ESCs have a 4S rating instead of a 6S rating; either they're built with other components that can't handle 30 volts like the MOSFETs can, or they're pushing 15-volt MOSFETs harder than they're rated and hoping the heatsinking will make up for it.

 

[JohnRobHolmes]

Esc's with 2s and 4s limits typically have analogue voltage regulators, or digital ones that are hastily designed. It's just cheaper.

 

[fyrstormer]

Where do you learn all this stuff? You must sleep with electrical-engineering books under your pillow or something.

 

[Waitwhatsthat?]

Where do you learn all this stuff? You must sleep with electrical-engineering books under your pillow or something.

It's kind of his job to know this stuff. You know since he makes some of the best esc's and motors around.

Sent from my XT1585 using Tapatalk

 

[fyrstormer]

Sure, but lots of people suck at their jobs. ;-) So it's nice that he's good at his job, and I marvel at the in-depth knowledge he has.

 

[JohnRobHolmes]

Sure, but lots of people suck at their jobs. ;-) So it's nice that he's good at his job, and I marvel at the in-depth knowledge he has.

:ror: Thanks. It really is just a passion that turned into a living. When I'm interested in something, I absorb a lot of information. Lots of reading technical documents from TI, Atmel, ST, NASA, Infineon, along with my involvement in the ebike world. I've been aiming at my own fully engineered speed controllers for a while, but turns out that it's a full time effort for 2 years or more to get there. I'm only able to put in about ten hours a week to study, but I've got the hardware side down pat. Just need to tie together the last bits of software and start compiling code.

 

[Highmark]

I love a good post with a bunch of EddieO and Holmes replies, it usually means I'm going to learn some stuff. :ror:

 

[84yoda]

I love a good post with a bunch of EddieO and Holmes replies, it usually means I'm going to learn some stuff. :ror:

Seriously. Plus now having robert/thomas from castle, always learning from all of them."thumbsup"

 

[fyrstormer]

You're writing your own ESC firmware? Wow. Most people can only handle the hardware or software side of things.

 

[JohnRobHolmes]

I'll never get exactly what I want without understanding both sides. Hardware done right can eliminate much software.

 

[fyrstormer]

I'll never get exactly what I want without understanding both sides. Hardware done right can eliminate much software.

Hardware done right can eliminate *all* software, at least in the way people generally think of software. Case in point: Mechanical chronograph watches. To the extent that they have software at all, it's carved into the mechanical design of the parts. In the RC world, I suppose the closest comparison is an old-fashioned mechanical speed controller containing contact pads connected with resistors, and the resistance values of the resistors describe the speed controller's throttle-response curve. It's reconfigurable if you know how to use a soldering iron, but most people wouldn't recognize its mechanical design as "software" in a familiar sense.

The opposite end of the spectrum is a PC -- the hardware is designed to be reconfigurable on-the-fly with different software packages that are copied from non-volatile storage into general-purpose hardware capable of doing damn near anything. It's not as mechanically or electrically efficient, but its flexibility makes up for that in a lot of circumstances. Most ESCs are just single-chip computers connected to signal amplifiers, that rely on replaceable software packages loaded into their memory to govern their operation; they could just as easily control anything else if you could find a software package to make them do it.

In both cases, software in its purest form is just the configuration of hardware that causes it to perform a useful function; this is what distinguishes software from data. The software of a book, for example, is the physical configuration of having pages bound to a spine and the ability to retain markings on each page; the markings themselves are the data, and don't contribute to the book's functionality as a book, which is why blank diaries are still usable as books even though they don't have any markings on the pages yet.

So I guess, in the end, I'm agreeing with you that hardware done right can eliminate *most* software, but software as a concept is more abstract than 1's and 0's stored in memory. People think it's limited to 1's and 0's because the term "software" didn't exist until computers were invented, so it was never used in any other context. You could design a speed controller that doesn't use any binary codes to control its operation (and for decades, people did just that), and it would still have software in the form of its mechanical design. If you could eliminate sliding contacts from the design, it would probably be *even more* efficient than a highly-optimized ESC, too. Whether people could live without a million easily-adjustable settings, in an age where we expect everything to have a million easily-adjustable settings, is a different matter.

Sorry for the philosophical tangent, I'm a software engineer by training. Most of my professors used to work at Bell Labs back when UNIX was being created (and Multics before that!), and we spent weeks' worth of classes talking about topics like this. People tend to see software as a convenience item, but one of my professors had a story he liked to tell about being awarded the contract to write software for a fighter jet; the Air Force officer overseeing the project congratulated his team, and then told them if the plane crashed they would get to explain what went wrong to the pilot's family. Software is serious business in a lot of applications you'll never see on sale at Best Buy.