It seems to be the next thing with powerful brushed stuff being more awkward to get hold of in the UK, had a look at various ESC's and these seem to fit the bill but wanting other opinions?
http://www.hobbyking.co.uk/hobbyking...Truck_ESC.html
Printable View
It seems to be the next thing with powerful brushed stuff being more awkward to get hold of in the UK, had a look at various ESC's and these seem to fit the bill but wanting other opinions?
http://www.hobbyking.co.uk/hobbyking...Truck_ESC.html
Won't work. You need an esc and motor with closed loop feedback so that the esc knows the exact poisition of the poles in the motor.
Ah okay, makes sense. Still learning about brushless myself, glad to have the pointers :)
Something more like this?
http://www.hobbyking.co.uk/hobbyking...Truck_ESC.html
A non sensored ESC still has feedback through the use of back EMF normally. It is prefectly possible to use sensorless brushless setups and still have good performance. You'll just experience less smooth startups.Quote:
Originally Posted by typhoon_driver
yes as long as the motor is a sensored motor. However I'm not familiar with the ins and outs of using a brushless for drive. As far as I know it's only been done successfully either on the continent or in the states
oh and there's no way that thing is taking 700 amps or whatever it's spec'd to. Those are hobby amps, not genuine. Also bear in mind that these esc are made with relatively lightweight models, not our 30lb machines.
700 amps is laughable, it seems entirely possible to do so may be worth a bit of exploring
You're wrong, back EMF can and is read from a sensorless motor. Back EMF is read from the inputs to the motor. I am actually leaning more towards sensorless setups in fighting robots. Sensored introduces more parts/components so more potential things break and often they are simple optical sensors and under certain conditions can fail (generally not built for robots tumbling all over the place). Part of my lean towards sensorless brushless is the entire motor becomes a single moving part with just windings and magnets. Not much to get knocked about and fail.Quote:
Originally Posted by typhoon_driver
EDIT: The 700 Amps is probably the combined peak rating of the FETS. Wouldn't fully trust that.
Don't try to small a motor. A 28mm won't do.
By all means prove me wrong but there's a good reason that CNC machines have feedback on the axis. It's not that you can't keep good position with a stepper motor, but it's to ensure that if steps are missed or there's backlash, the system can still travel the correct distance. Different application, different motor, same idea.
Challenge accepted.
I've been looking at brushless drive too (Mr Mangle inspired me!) My biggest concern right now though is finding ESCs that will have good Fwd/Rev response, most truck ESCs have braking or a delay.
All brushless esc's have a delay, even the sensored ones. You can't trow a brushless in reverse like a brushed motor. It has to be stopped. The only thing you can have is a minimal delay.
It will take a lot of practise to get the best from the new tech.
You're right different application, different motor. Which is why this doesn't apply. CNC machines keep feedback because they need precision to fractions of a millimeter, a fighting robot does not, no matter how expert a driver your 'precision' does not warrant such control. Back EMF is a perfectly good way of measuring the motor. This isn't some few hundred steps a millimeter stepper motor, this is a brushless that probably has around 10-30 poles. The feedback in a fighting robot is not position tracking but ensuring that the motor RPM and ESC timing match to sustain torque. If you want position tracking in your robot feel free to drive with stepper motors.Quote:
Originally Posted by typhoon_driver
Agree with the practice but I need that regardless but its a slight change in driving style :)Quote:
Originally Posted by maddox10
The availability of brushless motors and esc is improving. The accuracy that sensored brushless gives is not really necessary in our applications where some wheel slip is acceptable. Where sensored does score is the startup of the motor. Sensorless controllers have challenges during motor start-up, as no back EMF is present when the motor is stationary and therefore need to bump the motor initially to determine position on powerup. The motor can become out of sync when running with the position the ESC believes it in. Improvements in ESC s/w should minimise this.
Its going to take some experimentation in building drivetrains for fighting robots and testing them in competition.
If there is a delay on my brushless drives, its too short to detect. I also find that the sensored brushless drives have better torque and control at low speeds than the brushed system I was using. Perhaps a short video will convince the doubters? There are also few problems with the sensors being delicate. The sensors are three small surface-mount hall switches on a well protected PCB; the only thing I worry about is the sensor cable and its tiny connectors.Quote:
Originally Posted by maddox10
BTW: the 2nd Hobbyking ESC mentioned looks ideal to me; I have two of the 150A controllers on the way for testing.
I tried the NTM28-36 motors. Start torque ain't what is needed. It drives, but not what we want.
What gear ratio did you use? With sensorless motors, you need plenty of reduction and/or a fairly large motor. By comparison the 540 size sensored motors I use have very similar performance to drill motors.
There is some testing I did with NTM 42-38 with a 12:1 gearbox (video description is wrong), 15kg of steel dumped on top.
http://www.youtube.com/watch?v=HKg-FZft0Wc
I'm hoping to have some 19:1 gearboxes shortly to test with to see if the start up torque is enough to smoke the tires
6-1 gearratio to 90mm wheels on 3S.
I have 42-35's 750 KV as backups, but implementing those is another matter.
Its hard to smoke a Colson - just breaking traction at low speed is a very good sign the motors are adequate. A 19:1 gearbox and a higher KV motor should do even better. Its good to see us Aussies leading the way on brushless drive!
In Nebelwerfer 2 the Frankensteined Gimson GR02's 24/1 with HK 2838 2850 KV inrunners mated to the Trackstar 25A sensorless speedo's work fine. And that's with large, soft 130mm RC wheels. And that one is standing by for months now. On the other hand, that's a 6 kg raptor.
And yes, it's the starting speed that is a problem with the current Calliope drives. Once it goes, it goes well.
http://youtu.be/2Aa4Qx_iEB0
NTM 42-38's with Trackstar 80 controllers on a 4-cell battery. 10:1 onto 100mm wheels but the total setup is only 6kg and has problems with the chains and can't afford to rebuild it yet.
This is a sensorless setup and as you can see the delay isn't noticable.
I also forgot this older test with the same brushess motors/ESC but with 25:1 onto 250mm wheels. It was a 10kg setup. Video was origionally posted 12 months ago and I still keep reading that sensorless outrunners won't work. Really wish I could afford to build the new robot the drive was origionally brought for.
http://youtu.be/bRkM_OzetfM
I'll just post this link again. Maybe someone is interested in some.
http://vedder.se/2014/01/a-custom-bl...or-controller/
That esc looks bang on Tim- shame there not sold pre-built.
So far our test results are promising, but like Mario said in Caliope we married a too small a motor with too small of a gear ratio. I'm convinced that with a dual stage gearbox the drive would have been fine. But as it is it is not taking off and stopping fast enough. The Brushless drive in de raptor performs really well, similar performance as the Scarifier video from Daniel.
Aren't Botbitz a brushed ESC originally?
No, the TZ85 is originally a brushless plane ESC that Steve reprogrammed to be a bi-directional bot ESC. Its a little ironic that the same people are now leading with brushless drive motors.
In my process for designing a machine for live events I was looking at these as a combination:
http://www.hobbyking.com/hobbyking/s...grammable.html
http://www.hobbyking.com/hobbyking/s...tor_730KV.html
Running on 4S, 73mm wheels with two 3:1 belt reductions (9:1 total) it would be able to do just over 10mph.
Looking at the PDF, the ESC is capable of Forwards/Brake/Reverse and Forward/Reverse:
http://www.hobbyking.com/hobbyking/s...7X53414X36.pdf
Unless the ESC can take the large back currents during braking/soft breaking it may have to be run in simple Forward/Reverse.
The alternative would be to get a large diameter outrunner with lots of torque to help compensate for the lower starting torque. But from the vids and other combinations people have done it may not be necessary.
Thoughts?
Beefy motor for that ESC. I use that ESC for my truggy and it performs well but it is designed for 1/8th scale applications. The settings are adjustable for your needs, the programmer can minimize the forward/reverse delay to as close to 0 as you can get it. The motor itself should have no problem starting up with your reduction ratio.
http://www.youtube.com/watch?v=3puKPFVUfdcQuote:
Originally Posted by Mouldy
Should do nicely in fw-size robots. I'm planning on making a few.
Nick what ratio banebots did you run this year? Id be Kinda tempted if i could find a motor to mate onto my banebot 16:1's to run my 98mm wheels with similar performance to what im currently getting
Yeah I meant brushless ESC lol. Was wondering if TZ85's are already brushless and there are loads of them kicking about to begin with someone could give them a go?
Well I'm going to watch this one from the sidelines with interest.
Unfortunately, the TZ85 in it's original form is not reversible - you need a car brushless ESC for that.Quote:
Originally Posted by daveimi
I am using 26:1 gearboxes and the pinion has a 1/8" bore. The set-up wouldn't have anywhere near the torque of your motors (775's, I guessing?).Quote:
Originally Posted by Mouldy
For your 16:1 gearboxes, I would use a pinion with a 5mm bore and either one of the NTM outrunners like Aaron @ BotBitz is testing, or a larger sensored inrunner like this:
http://www.hobbyking.com/hobbyking/s...g/22409(4).jpg
http://www.hobbyking.com/hobbyking/s...or_1900KV.html
If you want to wait a week or 3, I have a pair of those on the way for testing. You might want a bit different KV motor depending on your battery voltage and the speed you want - luckily there is quite a range to choose from. Will the 70mm can length be a problem for you?
No that would be ok- look forward to seeing your results!
Getting a controller is not the issue, plenty of brushless controllers out there that are proven and reliable in combat robots (the Vedder controller is a big unknown). The big thing we need to do is develop software that will properly for robots and handle a variety of motors (If found some controllers don't like some pole designs in motors ie: H36 motors with a Trackstar controller). BotBitz is working towards hacking the TZ85A to work as a dual purposes controller in both it's current Brushed mode that we sell, but also a Brushless mode. The problem is we need a good amount of time to develop the code, test it with a large range of motors and conditions before we could even consider selling it. At the current moment, I find the Turnigy Trackstar 6S 150A V6 controllers to best off the shelf solution, they are fully programmable, zero delay, have a wide range of voltage (2-6S) which suits most robots and they also have sensor ports (Just need some good sensored outrunners :) ). So while this controller is readily available, it kind of makes the job of hacking the 85A into a dual mode controller a reinvention of the wheel and therefore a slow progressing project.Quote:
Originally Posted by cavecrusher
It's not a case of them not working, it's a case of the conditions they are applied to they don't always work correctly. We have plenty of videos (as posted previously) with robots driving around with outrunners with sensorless controllers, but if you put the robot in a standing start stall condition with full weight and traction, they don't work (Check my video where the weight is distributed level). If your robot gets pushed into a corner and the opponent pins you, the sensorless setups rarely startup and just sit their cogging (as seen in my video) until they can make the first commutation. This is where the sensored controllers really shine, even when stalled, they can begin commutation instantly. This is one reason why I'm trying the bigger NTM 42 series motors is to get better start up torque, not to get 500+ watts of power into the drive.Quote:
Originally Posted by Daniel
Charles Guan (e0Designs/etotheiplusone) has done some epic builds with electric scooters and go karts using big outrunner brushless motors and ended up shifting to external to get reliable standing start performance in go karts, where 300lbs of weight taking off from a stand still with wheel spin is the game...
EDIT: Forgot to comment against Nick's post.
I already have a set of those motors for testing, however this is what I dont like about them:
- Max of 4S means they are only suited to low voltage robots or you have to build split powered robots with two batteries.
- Weight, they are 300+g each, which is nearly double the weight of a comparative outrunner.
- Cooling, they are fully enclosed and rely purely on the small heat sink fins. You could add fans etc etc to them, but this is more weight, space, complexity and it's probably not as good as the cooling an outrunner provides to itself.
- Size, they are pretty long motors, much longer then the outrunners and even Dewalts.
Hi Aaron,
Those HK motors seemed the closest to what I wanted and a logical step up from the 540 sized motors in Mr Mangle. I don't think they will get too hot if geared right - the 540 motors got pretty hot on the bench test but not so much in the arena. The other plus with the sealed inrunners is that the dirt stays on the outside.
Its obvious that both types of motors have their strong points; what we really need is for you and Steve to make an external sensor board for the NTM 35 or 42 series motors :). That would tip the advantage solidly over to outrunners.