Using winch motor for weapon drive?

[R9000]

I thought the low startup torque was only because of it being difficult to monitor back EMF until the motor's up to speed - surely sensored brushless should have no problem?

[overkill]

That's correct; I see a big difference in the start up performance of sensored VS sensorless motors in the same conditions. It also depends of the ESC, some brands seem to have smarter firmware to handle sensorless start-up. IMHO, most ground-based ESCs are designed for higher RPM motors that are run flat out. They don't need low speed control and sensors, so asking them to do low speed fine control in bots is not ideal.

[emisnug]

Brushless controllers are very expensive from what I can see. When you say heavy how much are we talking. Motors are heavy at the best of times!

Thanks.

In terms of weight saving, we're getting 4.5ishKw constant power (under load) for a motor weight of 1.25Kg (Slightly less now that I've chopped the excess shaft off of the end).

In comparison, a similar brushed motor (ME0708, which is admittedly the first 4.5ish KW constant motor I found on google) is 12.7Kg.

However, as Gary says, for the first couple of seconds, the raw torque generated by a brushed motor is far superior, but after those first couple of seconds (as low as .5s), it's a dead heat. For anything with a larger moment of inertia, it will be slower to spin up, but not by much - as soon as it breaks that moment of inertia, the brushless can pull more power and dump a bit more torque, so can spin up to speed perhaps 1-3 seconds slower than a brushed. Other roboteers, feel free to correct me.

Brushless motors are also more efficient in general, and don't really require servicing. The PWM curves, as Chris will tell you, are fun to deal with.

[daveimi]

It helps me to think of brushed and brushless and diesel and petrol respectively.

[Ellis]

Sensorless brushless motors are only efficient if your ESC is decent. Otherwise they can be shockingly inefficient at low speeds.

[R9000]

Why does basically nobody use sensored motors? Are they afraid of the hall effect sensors breaking? Even then, surely they'd still run in sensorless mode, assuming you've got the ESC to cope.

[typhoon_driver]

The problem is, those first few seconds, or even second are incredibly valuable. It doesn't take long for a rush across the arena to happen and most spinner impacts occur when the weapon has relatively low RPM. The big impacts are few and far between and generally due to chance more than anything else.

Thermal mass of the motor is another issue. I have no concerns about an etek burning out (my wiring and batteries are more likely to) but I just don't have the same confidence with brushless kit.

[Ianshock]

So team shock dabbled with with brushless for our heavyweight spinner...so learn from our mistakes ! For 1st time users forget brushless...we wasted £400 on HK flux Escs - which fell over just by looking at it and a 6.6kw brushless which should have done the job. But we geared it all wrong...as everyone has said starting torque is rubbish. Get a big old brushed motor and solenoid and it just works..we got lots of advice after our brushless failed. Brushed was heavier setup for sure but for 1st time heavyweight don't mess about is my advice. We also looked at winch motor and we almost went this route. I spoke to a great winch company I got off eBay and they could have supplied a decent looking motor with a usable shaft at about 4hp but Rpm was around 1400rpm at 24v. And weight was ilighter than A LEM 170 which was how we went in the end. they did sell me 600a solenoid for about £35 and that worked faultless. Brushless ESCs at this power required are tricky to get hold of. I think brushless for heavies are still in infancy and only good for advanced teams who have deep pockets !!! Hope this helps.

[shakesc]

No mercy on the control there Ian with the solenoid

I like brushless and it will get there but things can go wrong quickly and its definitely important to think about motor and esc as a system not individual components

[BrentwoodSchoolRoboteers]

Thanks for all the replies - some very interesting discussions. Obviously we are not in a position to risk blowing a chunk of our budget on something that won't work - though this could be true of both brushed and brushless. I will come back to you once we have more info on the drive system. It currently looks like it will be chain driven (ratio of 3:1 assuming a rough rpm of 3000rpm from the motor) with a torque limiting clutch in between. The current design looks to be unsuitable for a belt drive - partly because we can't find belts small enough and also we would want two belts and are limited with the depth of the gearbox.

We have another robot on the go which we plan to use for combat practice. There will be space for a spinning drum so maybe could look into brushless on here as this is more experimental.

I noticed on eBay you can buy a scooter motor and controller (brushless) for not a huge amount. I wondered whether the controller would be a simple case of connecting the throttle wires together to get it to run. If so this might be an easy thing to implement.

Thanks!