Adding hall sensors to outrunners

[Tiberius]

Nick,

These sensors do appear to latch when used with the sort of magnetic field strength we're using. In practice, you find that the sensor changes state when a certain magnetic flux threshold is passed, with a given magnetic polarity. The sensor output then remains "latched" in that state until exposed to a similar magnetic flux with the opposite magnetic polarity.

This is a good characteristic for use as a motor sensor, where the controller is relying on accurate edge-sensing, as it means that the sensors are relatively immune to noise.

My experience, having converted three motors now, is that the output from these sensors, when used in this application, is very reliable. I was surprised at just how accurate the zero-crossing point of the three sensor outputs were, when compared to the trapezoidal waveforms from the motor being driven by a sensorless controller. With the motor delta wired, these were spot on, as close as I could measure on the 'scope (probably better than 5 electrical degrees).

Jeremy

Hi Jeremy,

If you think in terms of a latching relay, then that does sound just like latching. But in terms of circuit operation, its hysteresis. I imagine what's inside is an analogue field sensor followed by a Schmidt trigger. In that particular case its debatable whether there is any practical difference between latching and hysteresis. Normally, though, I'd expect the term latching to be used where the input is digital and hysteresis where the input is analogue, so I still claim its "hysteresis". The latching relay I mentioned at the beginning has digital inputs.

It looks like the "latching" versions just have more hysteresis, so its a bit naughty of Honeywell to call some latching and some not. They would be better off calling some "high sensitivity" and some "high field strength". That would be less misleading.

Anyway, I'm going to be running some of these motors with sensored controllers, which is why my head is full of these details. I'm probably going to try both fitting Halls and electronic synthesis of the Hall signals.

BTW, it looks like most of the prior art reported in this thread is in the UK.

Nick

[gwhy!]

Nick,

These sensors do appear to latch when used with the sort of magnetic field strength we're using. In practice, you find that the sensor changes state when a certain magnetic flux threshold is passed, with a given magnetic polarity. The sensor output then remains "latched" in that state until exposed to a similar magnetic flux with the opposite magnetic polarity.

This is a good characteristic for use as a motor sensor, where the controller is relying on accurate edge-sensing, as it means that the sensors are relatively immune to noise.

My experience, having converted three motors now, is that the output from these sensors, when used in this application, is very reliable. I was surprised at just how accurate the zero-crossing point of the three sensor outputs were, when compared to the trapezoidal waveforms from the motor being driven by a sensorless controller. With the motor delta wired, these were spot on, as close as I could measure on the 'scope (probably better than 5 electrical degrees).

Jeremy

This is exactly what I have found with using these halls, they just work really well. If I had the resources I would like to try some different types/makes just for comparison but to be honest at the moment I can not see how this switching can be improved upon.

I would also like to take this opportunity to thank Jeremy for his help when I first started looking into fitting halls to my outrunner's, it has certainly opened up the true potential of what these little motors can do. Cheers.


Uk rules never settle for second best....

[JEB]

Jeremy, do you have any suggestions as to a sensored controller that will run the Astro (sensors added) motor to its rated 7500rpm? Thanks, Jim

[gwhy!]

Hi Jim,
The couple of e-crazyman/keywin controllers that I have should be able to run the motor upto that speed, I have run my outrunners upto 10k with them without no problems I think Luke has also used one upto that speed.

[Burtie]

You may have already seen this simple demo of the behaviour of the SS411A (from the Astro thread), but I will link it here because it fits.

It seems to demonstrate exactly the latching behaviour that Jeremy describes above.

http://endless-sphere.com/forums/viewto ... 30#p231050


[Edit: Would it be fair to say that latching of output signal is implemented by using hysteresis? ]

Burtie

[JEB]

Thanks gwhy for the info on the e-crazman controllers, all of my bikes/trikes are on a 24 volt system where I just move the batterys over from bike to trike, are the e-crazman controllers potted?.

I would try to get them to work on 24 volts, and not 36v, has any one tried to do this?

Thanks, Jim

[Burtie]

Hi Jim,
The ecrazyman controllers (aka infineon controllers) are not potted. I think a 24v version should work ok at 36.

[ ecrazyman is keywins name on Ebay. ecrazyman@gmail.com ]

There are a number of threads about these controllers here, 6, 12 and 18FET versions. But it may take a bit of digging

Cheers,
Burtie

[TPA]

If it has not been done already, can someone do a summary in the reference area with all the details on how to accomplish this modification. And if you do, can you include the vendors to source all the components. After reading through the thread, it is unclear regarding some of the details. Some details are treated as though they are known, as they should since most of you are quite knowledgeable in this area.

[gwhy!]

bench test was good, And now on my bike its much much better than before there is so much more snap now that its gone and broke my freewheel hub .. ( note to self... roll the throttle on ....SLOWLY.. )..

[Burtie]

Excellent news gwhy! ----errr sorry to hear about your freewheel though .

Do you have any pictures of your revised implementation? I love pictures


It is great to hear that this is working in practise. I have not been able to try It on my bike yet because I am still working on modifying my controller(s).

Burtie.

[gwhy!]

some pics of my modified hall sensor block from 60 degree spacing to 17.14 degree spacing, the timming adjustment is achieved by loosening the 2 bolts this allows the hall block to moved. The gap between halls and motor is approx 1mm I didnt want to have to remake the motor faceplate to allow a bigger gap, as Burtie suggested but the set performs very well with this slightly smaller ( 3.25kw ) motor than Burtie is using.

modded hall block.jpg (51.64 KiB) Viewed 1205 times

flea drive _mount.jpg (50.61 KiB) Viewed 1205 times

[gwhy!]

Just found some pics of my original hall mount.

001_hall housing.jpg (36.5 KiB) Viewed 1192 times

002_hall housing.jpg (32.84 KiB) Viewed 1194 times

bigger_motor_with halls.jpg (97.75 KiB) Viewed 1195 times

[dbaker]

Great pictures, Gwhy! I see your curved block with exit wire hole. Do I understand correctly that there are 3 Hall sensors spaced at 17.5 degrees in the curved block sandwiched between the curved aluminum plates? And you simply feed all 3 Halls with + & - from the Keywin controller and then connect the remaining Hall sensors wires to the 3 Hall inputs to the controller?

Thanks for sharing your work!

Dave

[kfong]

Nice job Gwhy! What kind of performance are you getting with that turnigy motor. I have the exact same one SK6374, I was planning to use in on a push scooter since it seemed small.

[gwhy!]

Great pictures, Gwhy! I see your curved block with exit wire hole. Do I understand correctly that there are 3 Hall sensors spaced at 17.5 degrees in the curved block sandwiched between the curved aluminum plates? And you simply feed all 3 Halls with + & - from the Keywin controller and then connect the remaining Hall sensors wires to the 3 Hall inputs to the controller?

Thanks for sharing your work!

Dave

The sensor is set into the slot on the curved surface, the flat side of the sensor is facing towards the motor and the 3 leads from the sensor run up the small hole to the channel between the ali sandwich ( there is only one slot cut in the picture but there are 3 ) the spacing is 17.14~ degrees between the sensors. All the +'s are joined together and all the -'s are joined then the 3 outputs from the sensors along with + and - go to the controller hall plug.

Nice job Gwhy! What kind of performance are you getting with that turnigy motor. I have the exact same one SK6374, I was planning to use in on a push scooter since it seemed small.

the motor runs very well on 48v, no load its drawing 2.5 - 3A. On my bike It have peaked at about 5kw it also runs quite cool, I have put more glue on the magnets and also have replaced the bearings (didnt need doing but done it anyway ) I also have added a additional bearing on the faceplate of the motor to help support the output shaft under load... All good so far ..

[liveforphysics]

I also have added a additional bearing on the faceplate of the motor to help support the output shaft under load... All good so far ..

That is what makes these motors designed to spin a propeller with no side load, work in a heavy side load chaindrive application. The bearing in the mount is critical IMO.

[dbaker]

Luke,

Where did the discussion on the instant start Infineon controllers that was part of the other Hall sensor on RC motors go? I infer from Methods use of Clyte 18 fet boards that the 18 fet instant start issue went that direction but I have not been able to find a discussion of 12 fet controllers.

Your posts are always an inspiration. I loved the cellman battery test stuff!

Dave

[AussieJester]

Luke,
I have not been able to find a discussion of 12 fet controllers.


Dave

viewtopic.php?f=2&t=4282

KiM

[dbaker]

Thanks, Kim. I recognize the beginning of that thread and I did check the end. Now to the middle of those 29 pages! I was looking for the conclusion or the next step in the discussion of Luke's 12 fet Infineon's instant start problem that sorta' disappeared in the Hall thread (somewhere around Aug/Sept?). I got wrapped up in it and before I knew it the clock showed 3am so I am trying to find the conclusion to that great story!

I always enjoy your spirited posts and the lines on your current build are great. I am planning an HXT build with an Infineon controller so am interested in ES folks' experiences with that combination.

Envy your summer weather!

Dave

[gwhy!]

Thanks, Kim. I recognize the beginning of that thread and I did check the end. Now to the middle of those 29 pages! I was looking for the conclusion or the next step in the discussion of Luke's 12 fet Infineon's instant start problem that sorta' disappeared in the Hall thread (somewhere around Aug/Sept?). I got wrapped up in it and before I knew it the clock showed 3am so I am trying to find the conclusion to that great story!

I always enjoy your spirited posts and the lines on your current build are great. I am planning an HXT build with an Infineon controller so am interested in ES folks' experiences with that combination.

Envy your summer weather!

Dave

I have a 12fet 116 jobbers that I have been playing with and it works perfectly , I haven't encountered any of the problems that luke was having with any of my controllers. I had a slight issue with my latest 6fet (116) but this was down to the 60degree timing I was using then ( it was very odd ) Since I have changed to 17.14 spacing all 4 of my controllers work so much better.

[AussieJester]

I was looking for the conclusion or the next step in the discussion of Luke's 12 fet Infineon's instant start problem that sorta' disappeared in the Hall thread (somewhere around Aug/Sept?).

Can't help you there matey, i know very little about anything electric, i trust Luke has it sorted though he is 'bomb proofing' my 12fet controller for me ....Only thing im worried about is he will make it so it wont start unless it sees at least 100v of Lipos & have ONLY two speed...stop and flat out hehehe

Envy your summer weather!

our weather gets old fast when its over 30c day after day after day i givez you the tip unlike the fellas in the snow keeping them indoors, here its the heat, getz over 35 and even the beach is a horrible place, you get near 3rd degree burns walking across the sand to the surf LoL...

Best of luck with your build i think you will LOVE the Turnigy very 'torquey' motors sound incredible at full throttle too ...

KiM

[TPA]

& have ONLY two speed...stop and flat out hehehe

Will likely only be one speed....Make sure you are on the bike when you hook up the power leads.

our weather gets old fast when its over 30c day after day after day i givez you the tip unlike the fellas in the snow keeping them indoors, here its the heat, getz over 35 and even the beach is a horrible place, you get near 3rd degree burns walking across the sand to the surf LoL...

I used to feel sorry for you when I used to read that you worked in 30-35C weather....until I did the conversion to F.

From mid June to Sept. I do not see less than 35C almost every day. It is not unusual to see 40C (104F) here.

...and all this talk of frock wearing...

[AussieJester]

...and all this talk of frock wearing...

.....has nothing to do with weather.... although, would be nice in a cool breeze ...maybe these frock motor boyz are onto something Hyenaaaaaa

KiM

[nieles]

hi,

i need some help, i am in the proces of adding hall sensors to my outrunner.
i made a test stand with the sensors at 120 mech. degrees. in my final desing i will use 17.14 degrees.

1. is there a big difference between the ss41 and ss411a? i am using the ss41 right now and works fine, but just wondering

2. does it matter what way the halls are facing, they are all facing in the same direction. if i flip them all, does it still work?

Niels

[gwhy!]

hi,

i need some help, i am in the proces of adding hall sensors to my outrunner.
i made a test stand with the sensors at 120 mech. degrees. in my final desing i will use 17.14 degrees.

1. is there a big difference between the ss41 and ss411a? i am using the ss41 right now and works fine, but just wondering

2. does it matter what way the halls are facing, they are all facing in the same direction. if i flip them all, does it still work?

Niels

the 411 has a faster switch time, and is just generally a higher spec hall sensor than the 41 also the 411 latches, the 41 should be ok. As long as they are all facing the same way it will still work.