Motor & Controller for Efficiency Racer!

jimcroisdale

jimcroisdale

10 mW
Joined
Mar 13, 2024
Messages
24
Location
Leeds, West Yorkshire
Hi all,

I'm part of a team that's designing an electric single seater "efficiency" style racer - 3 wheel reverse trike formation, super low drag body - you know the type. There are many final technical decisions/choices yet to be made, but we have a lot of solutions "circling the drain", ready to fall into place.

Our car will need to be able to run at a constant speed of around 35mph, using around 650w. Battery pack will vary from 94v at its highest (right off the charger), quickly dropping to 86v at first use, and then gradually tapering off to a minimum of around 83v.

Wheel speed for 35mph will be approx 500rpm, so low kv motors are needed, to avoid reductions ratios higher than 10:1. We will most likely be using spur/pinion gear drive, as it is most efficient, and ratios can be easily adjusted.

We reckon 15Nm at the wheel will do it.

There are some very expensive solutions out there (motors that the big solar car teams use) but they are waaaaay out of our budget. Our research has led us to the outrunner RC motors of Astro Flight, Neumotor, Plettenberg and more - and they look to have some good fractional horsepower options with high efficiency, but we're wondering if there's anything else out there that we're missing. We're aiming/hoping for 92%+ - seemingly not easy at such low speeds.

We're also going to need an efficient (and bulletproof!) controller too!

Thanks for reading - I welcome your thoughts! 🙃

Jim
 
Well, I'll leave it to others to lend insight on the motors - that is a bit out of my realm. Though overall, I would say focusing on aerodynamics could yield the biggest benefit if you are running 35mph.

As an example - if you are looking at 92% efficient motors or drivetrain, a lot of work/money may get that up to what...95%? Or even say it went nearly ideal to 98%. Then you would look at the output side. 650 watts x 0.92 = 598 watts to the ground, 650 x 0.95 = 617 watts to the ground and 650 x 0.98 = 637 watts to the ground.

So going from 'very good' to 'extreme efficiency' to 'near superconducting' levels gets you somewhere between 19 and 39 more watts to the ground.

A great deal of research has gone into aerodynamics on the human powered cycling side. You can find a ton of links, but here is one with a table showing some watt savings vs aero improvements.

www.roadbikerider.com

Cycling Aerodynamics: Going Aero Without Going Broke

By Kevin Kolodziejski Whether you race or not, whether you ride 10,000 miles a year or four-tenths of that in two, one thing’s for sure. It’s part of your cycling DNA. It’s the continual craving (except on recovery rides) to go just that little bit faster. What’s not so certain is how you...
www.roadbikerider.com www.roadbikerider.com

You start seeing things like "shaved legs" saved 10 watts and aero/time trial helmet vs standard helmet saved 23 watts. ...and this is at 27 mph, the savings would be considerably more at faster speeds. So very minor aero improvements in your racer would save way more than costly motor upgrades.

Also, not sure if your 650w is based on calculations or some on-road experience? Some of the fastest human cyclists can hit 85+mph with a claimed "1 horsepower", or ~745 watts. (and faster still in the near decade since this record was set - but this is one article I noticed mentioning the power)


Obviously these are ultra-refined aerodynamic shells. But, again, serves to show what can be done with good aero vs more physical power.
 
I appreciate that given this is the internet, and this is my fist post on here, you don't know my background, or perhaps more specifically, the background of the team (I'm just the "Chief Generalist" at this point :-D ).

Let's just say that aerodynamic efficiency isn't an area that we're worried about at this stage. We have people that have that part well covered. Having said that, we do also have experience when it comes to electric motors etc (Greenpower being one), but as that was a spec series, choosing motors was not really part of it. What we DO know though, is that every percent that we can gain is going to matter, because our goals are lofty despite our small budget. There WILL be areas where we inevitable give efficiency away to the competition because of budget (and of course experience/expertise), but there's not a lot we can do about that at the moment.

So while I agree totally that aero is important, assuming that our drag figures are going to be as low as is possible for anyone regardless of budget, we need to maximise all other areas for the overall performance of the car to be in the right window for its intended application. Based on our maths at this point and the data that we have available, these few percent could make ALL the difference.
 
Well as a former Manager of a herd of MiT engi's Have you rounded out your group with a good fab guy? Almost always overlooked until you realise you need a generalist in actually making weird stuff.

Oh, and a wizard of an electrician. Doesn't matter which side of the tree they fall on (high voltage, low voltage, long run etc) the ones that got into it because they like the popping noises...
 
Not a lot of high efficiency motors exist in the 1kw range outside of the Astro motor models.

In a larger motor with higher efficiency, at 650w you'll be way out of the peak efficiency band :/
 
neptronix said:
Not a lot of high efficiency motors exist in the 1kw range outside of the Astro motor models.

In a larger motor with higher efficiency, at 650w you'll be way out of the peak efficiency band :/
Click to expand...
Yes, this certainly looks to be the case. Hence none of the common hub motors are on our list, despite their transmission efficiency.

Our other problem is pack voltage. Because we're working at around 86v, all of the 32xx series Astro motors will rev too high, at least in standard spec. The kv gets lower as the motors get more powerful (339kv for the 3205 vs 137kv for the 3220) but then I would imagine we're back to the issue of using a 4kw motor well below it's efficient range (and a 23:1 reduction ratio for final drive isn't great either). We'd probably need the 3205, but with a lot more winds to get the kv down significantly.

This is what led us to looking at the Neumotors 4619:

There are options from 20kv to 990kv, so in theory at least, we can select a motor that takes 650w with a bit of headroom (working on the assumption that this will be the motor's most efficient range), and still have revs at 86v that allow for reduction ratios less than 10:1. However, we've currently got no idea about the efficiency of these motor as no information is on the website, and my team mate has received no replies to his emails. :-/
 
Last edited:
DaLanMan said:
Well as a former Manager of a herd of MiT engi's Have you rounded out your group with a good fab guy? Almost always overlooked until you realise you need a generalist in actually making weird stuff.

Oh, and a wizard of an electrician. Doesn't matter which side of the tree they fall on (high voltage, low voltage, long run etc) the ones that got into it because they like the popping noises...
Click to expand...
Yes, we have a mechanical engineer who is exceedingly handy. Plus we all have some crossover experience through various DIY disciplines. What we are missing really at the moment is a hardcore electronics person. Although what do they say about there being an idiot in every room, and if you can't spot them, that means that the idiot is YOU. :D
 
Hey, thats me!

I was once the only guy in a startup without:
A degree from MiT
A degree in mathematics
A PhD
and ::coughs:: at the time a degree in a solid field. I was dual enrolled med/premed and kinda got the wobblies, bailed out and had to take what was next up.. So, dual majore Biz/journ (I had a *LOT* of units at the time) So yeah. I became the common sense quotient in the company for a while.. then they made me take the big chair because I was good with paperwork...

It was *not* a promotion.

I don't know any sparkies in your back yard.. now I know some amazing guys in Southern Jersey...
 
Well, sent emails out to maybe 10 manufacturers and to be fair, they have mostly all replied. Some don't do one-offs of any kind (even just custom winding) and that's fair enough. Some DO offer custom winding etc but charge a lot for it (obvs not part of their business model). Best options so far though are:

Astro 4010, wound down to 60kv
Neumotors 6521/53kv
Lehner Torqstar 1 (7024/40winds) which is about 69kv under load

What they've pretty much ALL said (both directly or indirectly, through their recommendations) is that a BLDC motor is generally at it's most efficient when run at about 30% of it's maximum power, so these are motors that are more powerful than what we originally looked at. Lesson learned there.

Not sure I have any questions for you all at the moment but I wanted to keep the thread updated - I hate finding old forum threads that just go cold without a conclusion when i'm doing my research! :-D I suppose we'll be onto compatible controllers next, which seems to be a tricky affair due to us running around 84v.....
 
jimcroisdale said:
We will most likely be using spur/pinion gear drive, as it is most efficient, and ratios can be easily adjusted.
Click to expand...
I don't believe this will be true in this case, while I think it depends on load and precision cut gears with ideal tooth profiles can be very efficient even at high loading you won't be at high loads and I'm not sure you will be able to get the gear profiles so good while also being able to change them easily. Even using off the shelf gears or gearboxes, these often are not designed for maximum efficiency, rather for durability.

High efficiency timing belts assuming you don't install an undersized pulley or use an inefficient tensioning system will be very efficient with little effort and it's very hard to mess them up. Also changing the ratio will be much easier without losing efficiency from throwing various gears at it. Even a bike chain with a low friction EP filled wax lube will be extremely efficient and again be trivial to change ratios with little change in performance.
 
Vbruun said:
Is weight a concern?
Click to expand...
Well, weight is always a concern I guess, but if there's something that's efficient yet weighty then we can always plug the numbers into the spreadsheet and have a look and see how it fares. Most of what's been recommended already is in the 2kg bracket i think, so we're not expecting featherlight (which I think people assume we must be, given our overall low power requirement).
 
scianiac said:
I don't believe this will be true in this case, while I think it depends on load and precision cut gears with ideal tooth profiles can be very efficient even at high loading you won't be at high loads and I'm not sure you will be able to get the gear profiles so good while also being able to change them easily. Even using off the shelf gears or gearboxes, these often are not designed for maximum efficiency, rather for durability.

High efficiency timing belts assuming you don't install an undersized pulley or use an inefficient tensioning system will be very efficient with little effort and it's very hard to mess them up. Also changing the ratio will be much easier without losing efficiency from throwing various gears at it. Even a bike chain with a low friction EP filled wax lube will be extremely efficient and again be trivial to change ratios with little change in performance.
Click to expand...
This is a very interesting point, thanks!
 
jimcroisdale said:
Yes, this certainly looks to be the case. Hence none of the common hub motors are on our list, despite their transmission efficiency.

Our other problem is pack voltage. Because we're working at around 86v, all of the 32xx series Astro motors will rev too high, at least in standard spec. The kv gets lower as the motors get more powerful (339kv for the 3205 vs 137kv for the 3220) but then I would imagine we're back to the issue of using a 4kw motor well below it's efficient range (and a 23:1 reduction ratio for final drive isn't great either). We'd probably need the 3205, but with a lot more winds to get the kv down significantly.

This is what led us to looking at the Neumotors 4619:

There are options from 20kv to 990kv, so in theory at least, we can select a motor that takes 650w with a bit of headroom (working on the assumption that this will be the motor's most efficient range), and still have revs at 86v that allow for reduction ratios less than 10:1. However, we've currently got no idea about the efficiency of these motor as no information is on the website, and my team mate has received no replies to his emails. :-/
Click to expand...
You can model the efficiency of a motor via Grin motor simulator if you can get the necessary information.

Also, you can try playing around with the settings to get a feel for how things influence each other.

Motors generally aren't most efficient when run at close to max power, so if you are thinking about the neumotor, maybe consider going a bit bigger. Maybe you should contact them as well and hear how they think about the issue?
 
Could the battery pack be reconfigured to 47V so as to ease the kV constraints on your motor?
100% throttle might be optimal but how much does efficiency vary with throttle setting x kV? (The Grin simulator does simulation sets.)
How will the motor be cooled and how much power will that take?
Do "efficiency" style racers win by being first over the line or by being the most efficient at a given speed?
 
Vbruun said:
You can model the efficiency of a motor via Grin motor simulator if you can get the necessary information.

Also, you can try playing around with the settings to get a feel for how things influence each other.

Motors generally aren't most efficient when run at close to max power, so if you are thinking about the neumotor, maybe consider going a bit bigger. Maybe you should contact them as well and hear how they think about the issue?
Click to expand...
Yeah, if you read further down, the manufacturers all came back with suggestions that were higher power. We've got some motor sim software (my team mate has won at Greenpower racing in years past) but the manufacturers can't always supply all the data - windage, friction and details of BEMF etc.
 
AnthonyC said:
Could the battery pack be reconfigured to 47V so as to ease the kV constraints on your motor?
100% throttle might be optimal but how much does efficiency vary with throttle setting x kV? (The Grin simulator does simulation sets.)
How will the motor be cooled and how much power will that take?
Do "efficiency" style racers win by being first over the line or by being the most efficient at a given speed?
Click to expand...
No, pack voltage can't be reduced unfortunately. And yes, 100% throttle is most efficient as you don't have any PWM switching losses in the controller. So basically, we're designing to be most efficient at full throttle - around 30-35mph @650w. Motor will be air cooled. The particular event we're looking at is a 24 event, so it's most laps in 24 hours.
 
Anyway, we're getting some good numbers now - around 92-93% efficiency at the torque we're going to need. Just a couple more bits of data to get from the manufacturers and then we can start to bother them with questions about controllers :D
 
jimcroisdale said:
No, pack voltage can't be reduced unfortunately. And yes, 100% throttle is most efficient as you don't have any PWM switching losses in the controller. So basically, we're designing to be most efficient at full throttle - around 30-35mph @650w.
Click to expand...

This is not true.
A bldc motor still needs to have commutation via PWM even if run at full throttle/voltage.

If the duty cycle is low and the motor is low inductance, some losses will increase, but I would need to see data to believe that this is a significant issue.

Freeing yourself from very narrow KV restrictions would also make it much easier to find a suitable motor and controller.
 
Yes you still have commutation, but that's unavoidable isnt? I meant that you don't have losses from voltage reduction etc. PWM always has losses - we did look at going from the 85v down to around 24v at one point using a dc-dc converter (before the motor controller), and even the most expensive and efficient units weren't much better than 90% efficient at that range. I know the percentage loss is a function of reduction ratio though.

We've got a few options for motor now, that are correct for kv and in the 92-93% efficiency range so I think we're doing ok. Higher Kv motors are of course more plentiful, but looking at the data sheets, they aren't necessarily any more efficient, but will put the transmission at well over 10:1, which so far i'm led to believe is less than ideal for spur gear drive (wheel rpm will be about 500, so 60kv is an ideal maximum)

Like all these things, you have to be careful not to gain a couple of % in one area and then lose that or more in another! :D What we CAN'T do is drop the pack voltage, for reasons I won't bore you with.
 
jimcroisdale said:
Higher Kv motors are of course more plentiful, but looking at the data sheets, they aren't necessarily any more efficient, but will put the transmission at well over 10:1, which so far i'm led to believe is less than ideal for spur gear drive (wheel rpm will be about 500, so 60kv is an ideal maximum)
Click to expand...
How big is the driven wheel? If it's large enough you could fit quite a large pulley for a belt or chain drive and get that reduction while still having both cogs large enough to be efficient in a single stage. Obviously a larger wheel means more reduction but the idea is to keep the motor pulley as large as possible.

A single speed bike chain can be something like 97% efficient and that probably could be improved slightly with EP wax lubricants and careful chain tension.

If you're willing to believe gates' marketing material their timing belts can be 98-99% efficient but I'd bet this is in ideal conditions. Ideal conditions though are kinda of what you have, you don't need much shock loading or range of power transmission overall so you can run a small belt at a low tension, maybe like a narrow 5m or a 3m pitch and as long as you can make that motor pulley reasonably large. You'll want the belt sized as correctly as you can to the load, the smallest belt you can get away with although hard to know what that will look like, a number of variables here.
 
Wheel diameter is approx 600mm, depending on tyre choice.

Average Nm at the wheel will be 12 ish on the flat, although will be higher going up the gradients (which are as high as 3.4%)

Id love to believe that belt drive could be as efficient as spur gears!
 
You must log in or register to reply here.

Similar threads

P
help me chose a motor for my electrathon racer
Replies
20
Views
4,282
ProEV
P
M
drone pancake motor low reduction mid drive with regen
Replies
16
Views
5,830
Voltron
V
tigcross
Introducing the Electrom
2 3 4 5
Replies
112
Views
26,008
tigcross
tigcross
QSMOTOR
QSMOTOR,0.5-12kW Electric Hub Motor & Mid Drive Motor Manufacture China
63 64 65 66 67
Replies
1,664
Views
426,322
QSMOTOR
QSMOTOR
G
High efficiency hub motor
2
Replies
26
Views
3,808
Alan B
Alan B
Back
Top