controller for a MAC

Hi guys,

I am just upgrading my old 500W 10T MAC to the recent model with cassette. I am also looking for a more modern controller compatible with MAC motors. Until now I used an old EB312 controller running fine, but now I checked out a more modern Infineon controller, the KH612 and found out it is not going fine with a MAC.
So guys with a MAC, please let my know with controllers are you using without problems. Infineon compatible controller would be preferred. Thanx!

Mike

Phaserunner. Nearly silent.

The MAC is a high pole, high eRPM design. Many controllers choke while trying to drive it.
Example: the wheel is spinning at 400RPM ( In a 26" wheel, this is 30mph / 50kph ). There is a 5:1 gear reduction, so the motor stator is spinning at 2,000RPM. There are 16 poles in the motor, so that 2,000RPM becomes 32,000 eRPM or electrical RPM.

The EB3 series can handle pretty high eRPM.. maybe 50,000-60,000, but your average controller can only really do 30,000 or less.

In the phaserunner manual, it states that it's capable of handling 60,000 eRPM, which means it is fast enough to handle even some RC motor setups!
There are some kelly controller models that have a high eRPM option available.
http://kellycontroller.com/kls7212s24v-72v120asinusoidal-brushless-motor-controller-p-1348.html
Here is one that can go up to 70,000 eRPM!

I cannot personally vouch for either controller because i have no personal experience with them, but there's some leads for ya.

Phaserunner runs a MAC really sweet, but the true phase current limiting of the PR makes the getaway a lot softer than an EB312 which can be a whole lot more aggressive (wheelie). This is either Good or Bad depending on your application.

I've run both on a MAC10 with a CA3 on the same bike at 52V and either setup gives nice throttle control and can be set up for gentle, smooth getaways if that's your cup of tea. If you want quiet operation, small size, and other PR FOC advantages, then the PR is a good choice. Otherwise, the EB312 + CA3 will give you a really nice setup with temperature monitoring, etc for less coinage than the PR alone at the cost of a bit of trapezoidal growl on getaway...

Frankly, the PR really shines for DD where there is no gear noise and things run truly silent. In particular, DD allows you to use the unique Grinfineon and PR proportional regen feature which is very cool. So, you are starting out with fewer realizable PR advantages since the MAC has a bit of gear noise and can't do regen.

The Grinfineon pseudo sinewave controllers may also be a viable and less expensive choice. These will run eZee/MAC/BMC gear motors in sensored mode - you can check with Grin about sensorless operation, but I believe that's a NOGO. These also do not have automagic phase learning so you will need to fiddle as usual to get that sorted - annoying but a one-time task. No personal experience with these....

teklektik said:

The Grinfineon pseudo sinewave controllers may also be a viable and less expensive choice. These will run eZee/MAC/BMC gear motors in sensored mode - you can check with Grin about sensorless operation, but I believe that's a NOGO. These also do not have automagic phase learning so you will need to fiddle as usual to get that sorted - annoying but a one-time task. No personal experience with these....

Click to expand...

There are only a few true Field Oriented Control (FOC) drivers that IMO are properly implemented. Not sure what Phaserunner uses but the similarity in programming bears an uncanny resemblance to the TI 83zxx family. TI8312 with the 69M Piccolo in particular. I've inquired about which chipset but no one has answers.

FOC takes a feedback loop, in the case of the TI8312 that's every 200 nanoseconds. Fast enough to operate without sensors, but not quite sensitive enough for a clean high torque launch from zero RPM. If motors were equipped with a single AS5211 Hall sensor the Phaserunner would be a perfect match.

Most sensorless controllers have a crude feeback loop that detect the zero-cross point but use a somewhat fixed algorithm to shape the succeeding sinewave. I believe this is the Grinfineon implementation. A true FOC plots the sinewave incremetally based on the feedback loop and can operate without sensors while driving a sinewave as its a dynamic in sinewave propogation. Non FOC drives generally lack the feedback loop precision to be dynamic in Sinewave mode so they need the sensors.

I've been tempted to order a BAC500 to play with.

Hi, thanx for the proposals! The Phaserunner is new to me und sounds interesting. But it lacks of some control pins I use for my own precontroler, like speed/current switching, anti theft, tempomat or EBS. All Infineon compatible controller have these features. So, is there anybody out there using an Infineon compatible controller that works with a MAC? So far I am only aware of one type, the old EB3xx controller family. Is there another controller that works?

If the MAC is one of the older ones, the hall sensor locations are offset slightly which gives it advanced timing. This doesn't play nice with most controllers and you get really high no-load current. Anything that is sensorless should work around this problem. Otherwise there are only a few controllers that were specifically designed to handle this timing shift.

Hi !

Just a +1 to what has been said about the Phaserunner.
Works great with MAC, I had to put phase current limit to near max (96A) to get decent push when starting but then it's a breeze. Also had to tune kp ki parameters (don't forget to put 0 in rad before that) in order to avoid overcurrent protection.

I subscribe to your thread because I am also searching for a suitable controller for my second bike (same motor but I want more phase amps). BAC2000 would probably work but too expensive for me.

I am currently checking whether sabvoton can handle it, otherwise kelly seem an appropriate choice.
Not sure about mobipus, powervelocity, sevcon etc... yet.

I will let you know as soon as I get more information.

Just an update regarding my previous answer, Kelly controller is probably NOT a good option for MAC motors.
First, I was about to buy a sabvoton and realized the max erpm might not be enough (30000 ? to be confirmed) so I read the 20 pages about kelly FOC controllers because they claimed 70000erpm.

Long story short, izeman had a lot of trouble to get his MAC motor running with kelly controllers (not sure if he finally managed to get a stable speed).
Not only that but Kelly controllers seem to offer only speed throttle, no field weakening (some suspect it's not even real FOC), no firmware update possible, etc...

Some of these things might be subjective but I made my opinion and will stay far away from these controllers.

Still need to find the perfect candidate for my next build, I will investigate APT and Mobipus ones and keep you updated

MikeRD03 said:

Hi guys,

I am just upgrading my old 500W 10T MAC to the recent model with cassette. I am also looking for a more modern controller compatible with MAC motors. Until now I used an old EB312 controller running fine, but now I checked out a more modern Infineon controller, the KH612 and found out it is not going fine with a MAC.
So guys with a MAC, please let my know with controllers are you using without problems. Infineon compatible controller would be preferred. Thanx!

Mike

Click to expand...

I've been successfully bench-testing a MAC 8T with a KH612 controller... Haven't run it full throttle yet though, due to my sketchy "test-rig" :-D

https://m.facebook.com/story.php?story_fbid=1662820993738736&id=278230852197764

What problem did you encounter with your testing?

seeing my experience quoted, i chime in
making my motor from hell work with ANY controller is not easy at all. but the main problem is, that i run the MAC at way overrated power settings. 100-150A phase and 2.5kW total power.
if you run your motor near rated settings at about 500-1000W you will have it much easier to find a controller that works.
i now run a lebowski controller and lebowski did a great job refining his firmware to make the controller work with my small mac and andy's leaf/honda crx, not to speak of all those motors in between.
but i also had good results with powervelocities (near)sine wave controllers based on xie-chang boards. they are similar to eb3xx but have a lot more features and make the mac run fine at modest power. at 150A phase and 2.5kW they have their issues as well. but we think this is maybe because of the out of specs operation.
with the kelly sine wave the mac runs quite fine, but i had those strange throttle issues, making the mac not reach the wanted (and expected) erpm and after RELEASING the throttle it revved UP shortly before slowing down shortly after. couldn't solve that.

MikeRD03 said:

... the KH612 and found out it is not going fine with a MAC...

Click to expand...

did you HEAR that OTHERS had issues with it, or is this your own experience? if so please describe your issues here. it may help others.
edit: btw, you don't have your location set. if you are somewhere in europe i could send you a kelly and powervelocity for testing if you wish.

I take advantage of this thread to look for the perfect controller for my next ebike (Mac Motor 8T, 14s 30A 18650 battery).
My first ebike has a phaserunner controller which works perfectly with the same motor, but I would like to get something a bit more punchy when starting (96 phase Amp limit).

One of the main characteristic to look for is the erpm limitation because mac motor has 16pole pairs and a 5:1 gear ratio.
Because the motor can reach 50kmh at 14S, a controller able to do 40000/50000 erpm minimum is required.

I focused my research on FOC controllers and here are candidates I found so far:

BAC2000

pros: Tested up to 90000erpm (Justin), 150A peak phase current (50-75A continuous), 100A max battery current, up to 90V (72v nominal), support from Grin tech, RS-232 PC connection
cons: Price (375$+shipping), no cruise control possible (checked in backdoor) ?, missing detailed manual (available from ASI ?)

Sabvoton SSC 72150

pros: Price (250$), 350A peak phase current (100-180A continuous), 150A max battery current, up to 90V (72v nominal), USB PC connection
cons: unreachable support (Kathy ?), unknown erpm limit (30000 ? [edit] 30000 confirmed [edit])

Mobipus 72200 (not available anymore)

pros: 450A peak phase current (continuous limited by thermal rollback), 200A max battery current, up to 88.8V (72v nominal)
cons: Price (460$+shipping), unknown erpm limit, RS-485 PC connection

VESC 4.12

pros: Price (100-200$), open source, 240A peak phase current (30-50A continuous), ???A max battery current, super small, 100k erpm limit
cons: 60V limit (recommended for 12s max), reported failures, cooling is critical (?), reachable phase amp in real life (?)

Kelly KLS7215 (and other sinus ones)

pros: Price (230$+shipping), High speed versions available (72000erpm and 100k erpm), 150A peak PHASE current (60A continuous), ???A max battery current
cons: Bulky, problems reported with this motor, no field weakening available (real FOC ?), simulated torque throttle ?

APT 96600 & Sevcon

Super overkill for my build...


I also looked at others like Powervelocity and Yuyuang but no FOC is not an option for this bike.
Have you heard feedback about "Lishui" (Nanjing) or "pear control" controllers ?

I tried sending emails to kathy and filled Sabvoton website form in order to get information about this erpm limit but I never had an answer from them. If 50k is doable, I would definitely buy the Sabvoton SSC 72150.

I would be afraid to fry a Vesc with my 14s configuration (57.4v @ 4.1V/cell), however no regen planned with this motor (ie no overvoltage peaks ?). Do you think this is doable ?

The ideal candidate seem to be the ASI BAC2000 so far, but I have to admit I don't like how they hide documentation until you are a customer on their website. And it is quite pricey too...

If I cannot find anything better than the phaserunner I might come back to it and accomodate with the "soft start" 96A phase limit.

What would you do ?

After sending more emails, no answer from Sabvoton but their distributor "QS Motor" kindly answered my question.

He confirmed that Sabvoton SSC 72150 is limited to 30000erpm.
In conclusion the sabvoton will not work for this configuration with a mac motor.

You can (ask someone or do yourself) build a Lebowski controller (FOC, sinewave). Izeman has one with his 8T MAC, I think he has 100A (RMS) phase current, 2.5kW total, If I'm not mistaken he reaches over 50k-erpm (I know he has it set for like 90k-erpm but don't know if he reaches that). He has it mid-drive and reaches 45kmh in his lowest gear,

Hello Lebowski, thank you for the suggestion.

I read threads about your controller and this is definitely something I would like to support and try... Great job BTW
Your controller seems like the ideal one but I will need something more plug and play for this build.
I know some basic electronics and soldering, but assembling everything would take me ages... Do you know people that do it for others on the forum ?

I'll probably try it for a next build. Talking about it, maybe i missed it but did you plan an output data stream from the 'brain' for monitoring (RS-232 or so) ?

100A phase. pahhh. it's 140A
and 55kerpm is EASY to reach. i dropped it to that from higher values as i was afraid that the yoke would fall apart and lose it's magnets (only joking ... i feel eddy current would be too high at those rpms).
but you are right, if you can't build one by yourself than this is the biggest draw back of those controllers. someone should pick up the design and mass produce them.
it's by far the best controller i've seen and tested. the MAC is REALLY difficult to drive. lebowskis controller had NO issues at all to run any of my big DD motors over 1-2 years ago. it took a lot of lebowski's time and passion to bring the controller to where it is right now! kudos for that!
you really should find someone build you one. setup is super easy and straight forward. it runs the MAC with all default (autocomplete) settings.

Thank you for your feedback Izeman, that lebowski idea is making its way. Love the smd version of the PCB also.
As lazy as I am, I wish somebody will start providing kits for it (or assembled ones like it happened for VESC).

For this build I will need something that requires less assembling time. I guess BAC2000 is the best candidate but I can't prevent myself being suspicious about their very 'close-source' policy, even for documentation...

I am surprised not to see any VESC design modified for higher voltage.

Bouteille51 said:

Thank you for your feedback Izeman, that lebowski idea is making its way. Love the smd version of the PCB also.
As lazy as I am, I wish somebody will start providing kits for it (or assembled ones like it happened for VESC).

For this build I will need something that requires less assembling time. I guess BAC2000 is the best candidate but I can't prevent myself being suspicious about their very 'close-source' policy, even for documentation...

I am surprised not to see any VESC design modified for higher voltage.

Click to expand...

The VESC 6. is out there now and so is the schematic. Upgrading for more voltage is pretty easy to do.

Check with Vad at powervolicity.https://endless-sphere.com/forums/viewtopic.php?f=31&t=89673
Ask him how it will turn your mac.

999zip999 said:

Check with Vad at powervolicity.https://endless-sphere.com/forums/viewtopic.php?f=31&t=89673
Ask him how it will turn your mac.

Click to expand...

PV is not FOC and does not have torque/current throttle. This is why it has been excluded from comparison list, but it still looks like a nice and affordable controller.

People asked him so often whether his controller can run a mac motor that I think he would hate me for that :lol:

Final update: I bought another Phaserunner and I'll live with the 96A phase current limit (soft start).
It is compact, versatile, easy to use, and powerful enough for small builds.

Now I'll have some time to think about building my Lebowski's controller or try adapting VESC ones for future updates.

Thank you all for your help !