load sharing between 2 perm mag DC motors

[strantor]

This post is for the electric go cart project I'm brewing up (bought a rolling frame today, it's UGLY 50$!)
I think I want about 20hp worth of motor on it. I can get 2 10hp motors for much cheaper (and lighter weight) than I can get one 20hp motor.

I've seen already read http://forum.allaboutcircuits.com/showthread.php?t=57339 that mechanically coupling and electrically paralleling brushed DC motors results in one motor (the one with the least electrical resistance) taking the bulk of the electrical power, to the point of melting. I am wondering, Assuming 2 identical (not perfect world "identical", but same model bought from the same place at the same time) motors are coupled to a shaft and provided the same PWM duty cycle; if by using 2 seperate controllers (or 2 seperate banks of FETs in the same controller) for 2 motors, and mechanically coupling them, would they share a load evenly? Well, this makes me think that if they are both receiving identical PWM, then functionally they are still in parallel, just with seperate switches between the motors and ground and would encounter the same problem.

Another idea: what if on one motor (call it master motor) the PWM duty cycle is controlled by the accelerator, and the the duty cycle for the second motor (call it slave motor) is controlled to maintain a match of the current measured on the first motor?
Anybody have any wisdom to share about load sharing with DC brushed motors? I have a feeling it's doable but probably a lot trickier than first thought.

one question, on the theory of it; which of the following 2 statements are the closest to being true?
1. If the 2 motors aren't putting out the exact same torque at the same time, then the lower torque motor is actually acting as a load on the higher torque motor and slowing it down, not helping it. (I.E. motor 1 is putting out 100 ft*lbs of torque, motor 2 is putting out 75ft*lbs; total torque is 75ft*lbs) Chain is only as strong as it's weakest link.
2. If the 2 motors aren't putting out the exact same torque at the same time, then the lower torque motor is adding to total torque, but to a lesser degree than it could be if they were matched.(I.E. motor 1 is putting out 100 ft*lbs of torque, motor 2 is putting out 75ft*lbs; total torque is 175ft*lbs)

 

[bigmoose]

Strantor welcome to the forum. To answer your question, the easiest way is to wire the motors electrically in series (if they are brushed) and locking their shafts together. Run twice the battery voltage that you were intending.

One could also use two controllers that run in current controlled mode and power them in parallel. Because of the high temperature coefficient of copper resistance, if both motors are paralleled and run off of one controller, and the motors are not perfectly "matched", one will run away thermally to destruction.

Do a search on agni and TTXGP and you will see some of the experience of our members that are running two Agni's in parallel in the TTXGP race bike. You will see how much effort they go through to match the motors, and even then they can get failures.

 

[liveforphysics]

Parallel almost never works, unless you really know what you're doing with the setup.

What does work though, it's 2 separate controllers, or as Moose said, running them in series at twice the voltage is the best option.

 

[strantor]

Parallel almost never works, unless you really know what you're doing with the setup.

What does work though, it's 2 separate controllers, or as Moose said, running them in series at twice the voltage is the best option.

What does someone "who really knows what they're doing" know? in other words, what what needs to be know to do it properly? So far I've come up with brush timing, but I know that's not the whole story. What else is involved? length of motor leads possibly...? and no, being a noob, I'm not asking so that I can go read a wikipedia article and presume to know what I'm talking about afterwards, just curious.

Strantor welcome to the forum. To answer your question, the easiest way is to wire the motors electrically in series (if they are brushed) and locking their shafts together. Run twice the battery voltage that you were intending.

One could also use two controllers that run in current controlled mode and power them in parallel. Because of the high temperature coefficient of copper resistance, if both motors are paralleled and run off of one controller, and the motors are not perfectly "matched", one will run away thermally to destruction.

Do a search on agni and TTXGP and you will see some of the experience of our members that are running two Agni's in parallel in the TTXGP race bike. You will see how much effort they go through to match the motors, and even then they can get failures.

thank you!
The series idea I have encountered as well. It seems like a less than ideal solution for a lightweight vehicle, as you would need twice the batteries. Also I would think that (to a lesser extent) you would suffer the same (actually the opposite) problem if one motor had less resistance, less voltage would be dropped across it. I don't think I'm seeing the whole picture though.
I searched TTXGP and I found Jonescg's voltron build. read all 25 pages of it. his bike and motor failure is actually what sparked my question. from the build thread, I never found the answer as to why his agni melted. someone suggested the brush timing may not have been perfectly matched, but that was never confirmed or denied and no other possible answers were given.

 

[strantor]

FYI I'm the reason for this line of questioning is that I'm building my own controller. I have found that from the electronics perspective and the components available to me, it would be more advantageous to me to build a low voltage, high amperage controller, than a high voltage low amperage controller. This go cart build is only in support of my controller build. I need a test platform. If low voltage, high amperage means multiple motors, then I want that to be a major design feature of my controller. So, when I ask theoretical questions about the best way to parallel motors, try to think of it less from the perspective of what's possible using off-the-shelf controllers, and try to think of it more in theory, "what is actually best". That's why I asked about one motor receiving a speed command, and the other motor receiving a current command to match the current of the first motor. That's my idea for a dual motor controller; do you think it would work?

I am branching out from my electronics hobbyist background into this high power EV stuff, so bear with me.

 

[bigmoose]

No problem, and it is great that you are making a high current brushed controller. Just design your algorithm to measure current and make that your control variable rather than PWM period and you are set. Build 2 power stages, two current monitors and one "brain" and you are done.

If you are capable, I would size your build to run two Agni's and you might have some customers in the TTXGP and small car crowd!

 

[strantor]

No problem, and it is great that you are making a high current brushed controller. Just design your algorithm to measure current and make that your control variable rather than PWM period and you are set. Build 2 power stages, two current monitors and one "brain" and you are done.

That's the idea. I'm pretty sure it will work (from my electronics perspective) but just wanted to hear it from the mouths of the guys who have the motor experience. BTW, is there anybody out there making dual motor controllers? and if so, are they affordable?

If you are capable, I would size your build to run two Agni's and you might have some customers in the TTXGP and small car crowd!

You know, that's exactly what I was thinking. I was wondering how I would get my first customer, because I can't afford 2 agni motors myself, and if I did have 2 of them, I probably wouldn't let someone else test their homebrew controller with my motors. I can't sell it untested, so I would be at an impass. Anyways, I'm putting the cart before the horse, as I haven't even ordered the parts for the controller yet

 

[John in CR]

If you're building a new controller, then I'd request you go with brushless motors, because high power controllers are the missing ingredient for some incredible and economical high power to weight ratio motors. OTOH, I might be able to build one for a brushed motor and I know next to nothing about electronics.

 

[strantor]

If you're building a new controller, then I'd request you go with brushless motors, because high power controllers are the missing ingredient for some incredible and economical high power to weight ratio motors. OTOH, I might be able to build one for a brushed motor and I know next to nothing about electronics.

Yes building a brushed motor controller seems like it should be fairly straight forward; I will soon find out if that's true or not. Nothing is holding anyone back from trying. I have ideas for BLDC controllers and induction motor controllers as well, but I'm taking baby steps. I want to get some practical knowledge & experience with high power applications and see some success before getting in over my head with technology I don't fully understand. baby steps.

 

[liveforphysics]

Parallel almost never works, unless you really know what you're doing with the setup.

What does work though, it's 2 separate controllers, or as Moose said, running them in series at twice the voltage is the best option.

What does someone "who really knows what they're doing" know? in other words, what what needs to be know to do it properly? So far I've come up with brush timing, but I know that's not the whole story. What else is involved? length of motor leads possibly...?

The the axial motors, inconsistant air gap is to be expected. Very much so on Agni's. Just tweaking carrier timing to match KV doesn't help you if one is retarded or advanced to compensate for the gap difference.

Brush resistance. It varies way more than you would ever expect it to on Mars, Perm's and Agni's.

Cabling. (duh, this one is easy though)

Airflow through the carrier and the rotor being ~equal on both.


I had Cedric Lynch himself (creator of the Lynch, Mars, Agni motor flux design), digging from a large collection of Agni parts setup a twin Agni race bike for use in parallel. He had access to our dyno for as long as he wanted, access to thermal imaging, current data, ESR meter, and a fully instrumented dyno system. The best he could do was +-20% or so after a long time of dicking around.
Then on the racetrack, the airflow was all different, and once one got warm they didn't share worth a damn again.



Series is so easy, because no matter what, the current through both motors HAS to be equal, if it wants to or not.


When making a controller, even something as simple as a brushed one, anything you can do to reduce current levels to get the same power output will save you so many headaches in FET killing spike clamping, dI/dT related FET control issues, cap ripple, PCB melting and package limit melting etc...


Your battery does not need to be twice the size. I could be half the size or smaller, or 10x bigger or whatever you want. You would just adjust the capacity accordingly to suit your needs.


Electric karts are fun. Or at least semi-fun, as the ones we built a few months ago only have been used a handful of times since building them. lol

For a point of reference, we used Perm132 motors, 60v batteries, and 450amp controllers. It made for a mild, but pleasing powertrain for the karts.

[youtube]YazSVi36_X8[/youtube]

 

[John in CR]

I was thinking of just some metal bars with different resistances as my brushed motor bullet proof controller and throttle. 8)

Maybe get fancy and use some contactors to change the voltage by connecting at different portions of the battery pack if I wanted to get fancy.

 

[jonescg]

Hi Strantor, I have already replied to your PM, but I'll put my reply up for all others to see:

I'm running two Agnis in parallel, fixed at the shaft. The battery pack voltage is nominally 105.6 V, which is a bit high for the Agnis, but the controller is set to deliver only 90% of top motor speed (which is about the ~84 V maximum for them anyway).

Now, we smoked the first Agni motor because we put too much current through it AND it was out of balance. The controller I have is capable of delivering a stupendous amount of power, so when it was set to 90% power it didn't disappoint. Until the motor smoked. So we scrambled around to find a replacement motor, installed it that night and tried to dial the power back to 70%. This worked reasonably well, getting me second place in the first race. However, come the second race I was harder on the throttle and things started to overheat again. The problem this time was that even 70% was too high (70% of 1200 A max is 840 A, which split between two motors is still 40 A overboard) and the motors couldn't be balanced. The volt per rpm value of each motor was most likely different by a couple of volts, meaning one did all the work and the other was just holding it back.

Winton raceway was a much better weekend, as I got Arvind to send me two perfectly matched motors where the volt per RPM was within 1.0 of the other. In addition, the brushes were grouped into piles of similar resistances, and all 8 (four per terminal, for both motors) were installed on the positive, and the other 8 were on the negative. This meant the currents would be as close as practical and they would hopefully wear evenly. The Kelly controller was dialled back to 60% power output so out of turn 9 we were pulling 600 A max, battery side so the motors were well within spec. Each time I'd come into the pits, Abraham and Tony would run out with a multimeter and check the resistance of the NTC thermistors of each motor and they were never below 3 kOhm. The absolute don't-exceed-or-you'll-melt-it limit is 2 kOhm. The motors were run gently after a race or a hard session to keep them from sinking too much head into the brushes and as they cooled the resistance began to rise. Within 10 mins we were up to 7 or 8 kOhm. Do bear in mind that these conditions are race-pace, and regular toodling around would never load the motors like this. But on a go-kart, that's different

I disagree with Luke that you 'need to know what you are doing'. I don't know shit about electronics and I've come this far :lol: You need two matched motors, and there's 90% of the battle won. But, never set and forget though. Always inspect the brushes, and make sure you don't make any brash changes without remembering what you did.

Cheers,
CHRIS

 

[liveforphysics]

I disagree with Luke that you 'need to know what you are doing'. I don't know shit about electronics and I've come this far :lol: You need two matched motors, and there's 90% of the battle won. But, never set and forget though. Always inspect the brushes, and make sure you don't make any brash changes without remembering what you did.

That's a fact mate. If you get to start with matched motors, matched brushes, setup the controller conservatively, regularly inspect and balance brush wear, and ensure both have equal airflow/cooling, then you don't need to know what you're doing at all. But... if you're doing those things, you're kinda defining "knowing what you're doing" in respect to running parallel motors IMHO.

 

[John in CR]

I disagree with Luke that you 'need to know what you are doing'. I don't know shit about electronics and I've come this far :lol: You need two matched motors, and there's 90% of the battle won. But, never set and forget though. Always inspect the brushes, and make sure you don't make any brash changes without remembering what you did.

That's a fact mate. If you get to start with matched motors, matched brushes, setup the controller conservatively, regularly inspect and balance brush wear, and ensure both have equal airflow/cooling, then you don't need to know what you're doing at all. But... if you're doing those things, you're kinda defining "knowing what you're doing" in respect to running parallel motors IMHO.

All that conservatism for racing, why not just run 2 controllers? Are they just too big and heavy to fit?

 

[strantor]

I disagree with Luke that you 'need to know what you are doing'. I don't know shit about electronics and I've come this far :lol: You need two matched motors, and there's 90% of the battle won. But, never set and forget though. Always inspect the brushes, and make sure you don't make any brash changes without remembering what you did.

That's a fact mate. If you get to start with matched motors, matched brushes, setup the controller conservatively, regularly inspect and balance brush wear, and ensure both have equal airflow/cooling, then you don't need to know what you're doing at all. But... if you're doing those things, you're kinda defining "knowing what you're doing" in respect to running parallel motors IMHO.

All that conservatism for racing, why not just run 2 controllers? Are they just too big and heavy to fit?

or one controller with dual outputs

 

[liveforphysics]

Yep, a pair of controllers is very easy.

 

[strantor]

Hi Strantor, I have already replied to your PM, but I'll put my reply up for all others to see:

I'm running two Agnis in parallel, fixed at the shaft. The battery pack voltage is nominally 105.6 V, which is a bit high for the Agnis, but the controller is set to deliver only 90% of top motor speed (which is about the ~84 V maximum for them anyway).

Now, we smoked the first Agni motor because we put too much current through it AND it was out of balance. The controller I have is capable of delivering a stupendous amount of power, so when it was set to 90% power it didn't disappoint. Until the motor smoked. So we scrambled around to find a replacement motor, installed it that night and tried to dial the power back to 70%. This worked reasonably well, getting me second place in the first race. However, come the second race I was harder on the throttle and things started to overheat again. The problem this time was that even 70% was too high (70% of 1200 A max is 840 A, which split between two motors is still 40 A overboard) and the motors couldn't be balanced. The volt per rpm value of each motor was most likely different by a couple of volts, meaning one did all the work and the other was just holding it back.

Winton raceway was a much better weekend, as I got Arvind to send me two perfectly matched motors where the volt per RPM was within 1.0 of the other. In addition, the brushes were grouped into piles of similar resistances, and all 8 (four per terminal, for both motors) were installed on the positive, and the other 8 were on the negative. This meant the currents would be as close as practical and they would hopefully wear evenly. The Kelly controller was dialled back to 60% power output so out of turn 9 we were pulling 600 A max, battery side so the motors were well within spec. Each time I'd come into the pits, Abraham and Tony would run out with a multimeter and check the resistance of the NTC thermistors of each motor and they were never below 3 kOhm. The absolute don't-exceed-or-you'll-melt-it limit is 2 kOhm. The motors were run gently after a race or a hard session to keep them from sinking too much head into the brushes and as they cooled the resistance began to rise. Within 10 mins we were up to 7 or 8 kOhm. Do bear in mind that these conditions are race-pace, and regular toodling around would never load the motors like this. But on a go-kart, that's different

I disagree with Luke that you 'need to know what you are doing'. I don't know shit about electronics and I've come this far :lol: You need two matched motors, and there's 90% of the battle won. But, never set and forget though. Always inspect the brushes, and make sure you don't make any brash changes without remembering what you did.

Cheers,
CHRIS

Thanks, good to hear from the guy who got me into this, lol. Do you think the second motor actually working against the first one, or do you think it was helping, just to a lesser extent? That's one of the things I'm hung up on; if the speeds are what need to be synchronized or if it's the torque.

The "set it & forget it" thing is what I'm trying to accomplish with the dual output controller. no more need to do all this painstaking stuff to "perfectly" balance the motors. The second motor just looks at what the first motor is doing, and does the exact same thing, no matter how much current it takes to do it (all the way up to the current limit)

 

[jonescg]

I doubt the other motor was doing anything except adding load. We were trying to balance them by adjusting the brush advance, but it would become a generator if you took it too far. We weren't testing it under load, mind you.

If you have a barge of a frame, like a Dailem, then two controllers is easy. But I just don't have the room on my bike for two. If I had known about the things I've encountered, I would have gone with a big single motor, rather than two. But if you were stuck with a pair of Agnis, then I'd say series with a higher pack voltage is the way to go. I don't think Kelly makes controllers that run supremely well at 180 V, but it would certainly be a robust system.

 

[bigmoose]

Stantor, if you terminate your PWM period based on measured current and not time, both motors will near perfectly share the load. It's in the math of the control algorithm. If you base your control algorithm on PWM time period, it will be better than two motors paralleled on one controller, but they will not share the load equally. The benefit is that when you set your red line current trips, the controller will not let the deviant motor go beyond that limit, that is for two PWM period controllers and 2 motors.

Torque scales with current. If you want equal torque, you need equal current. Let the equivalent voltage be what is needed to drive the current... and it will be different for the two motors. In the extreme case when they heat up and deviate the PWM period will be different in the two controllers to achieve the same current.

 

[strantor]

Stantor, if you terminate your PWM period based on measured current and not time, both motors will near perfectly share the load. It's in the math of the control algorithm. If you base your control algorithm on PWM time period, it will be better than two motors paralleled on one controller, but they will not share the load equally. The benefit is that when you set your red line current trips, the controller will not let the deviant motor go beyond that limit, that is for two PWM period controllers and 2 motors.

Torque scales with current. If you want equal torque, you need equal current. Let the equivalent voltage be what is needed to drive the current... and it will be different for the two motors. In the extreme case when they heat up and deviate the PWM period will be different in the two controllers to achieve the same current.

I read this post several times and never understood what you were saying. Went and read some other stuff, purged my head of some stupid ideas, came back and read it again, and now it makes sense to me. I'm a little slow sometimes. what you are saying is exactly what I've been trying to say. If my controller functions as I conceive it, the voltage (well, PWM duty cycle rather) to each motor may or may not be the same at any given time. The duty cycle to the "leader motor" will be controlled by the analog voltage signal from the throttle pot. the duty cycle to the "follower motor" will be controlled to achieve the same current that is measured on the "leader motor". This will ensure that the 2 motors always have the same current and share the same load.

So, now I am wondering, does "same current" always equal "same torque"? I've heard that permanent magnets lose magetism with heat, so lets say the cowl on your bike blocks just a little more air going into one motor than it does the other, so now X amount of amps going through each motor doesn't necessarily equal the same Y amount of NM/ ft*lbs as it used to. Is this scenario possible? or is it that if the magnetism drops, the current will also drop?
Thanks for your time

 

[bigmoose]

Ok, now make the next big step. Do not think that the throttle sets the PWM period. Think that the throttle terminates the PWM period based on current measurement.

Yes there will be motor to motor variations based on Kt. But IMHO they do not need to be 100 matched, just independently controlled.

Play around with the following:

You can use these handy equations to calculate steady state motor performance. If the Torque constant is not supplied by the motor manufacturer, you can measure the motors no-load RPM/Volt and use the following equations to calculate the torque constant.

Torque constant: Kt=Kv x 1.345
Current draw of motor: I = [V-(Kv x kRPM)]/Rm
Torque output of motor: J = (Kt x I) - (Kt x Inl)
RPM of motor: kRPM = (V - RmI) / Kb
Power output of motor: Po = (J x RPM)/1345
Power input: Pi = V x I
Motor efficiency: Eff = (Po/Pi) x 100
Current at peak motor efficiency: Ie max = Sqrt [(V x Inl)/Rm]

Symbol Definitions:
Eff = Efficiency
I = Current
Iemax=Most efficient current
Inl = No load current
J = Torque (oz-in/A)
Kv = Voltage constant (Volt/1000 RPM)
Kt = Torque constant (oz-In/A)
Pi = Power input (Watts)
Po = Mechanical power output (Watts)
Rm = Terminal resistance
RPM = Revolutions/minute
V = Voltage