Compact Field Oriented Controller, ASI + Grin, limited run

I'd like to provide a little constructive criticism on the Phaserunner software.

There is a lot of important info that is provided in the hover-over popups and it's tricky to make it appear, and harder to make it stay visible long enough to read the whole thing without needing to move your mouse off the setting then back on to get the hover-over popup back. I think the best bet would be to move all this into a help menu available in the software. Alternatively, you could make a little ℹ️ icon next to each setting that you can click on to view the popup content and it stays visible until you click away or close it.

I also recommend making the very first step in the manual setting the Feedback Bandwidth Tuning, since if this is too low on a geared motor you'll get phase over current faults just trying to run auto-tune. You also mention this setting almost as an afterthought at the end of the sensorless self start section in the manual. I think that this should really be it's own section and it should be step number 1. Getting this right is critical for geared hubs or else you'll get all sorts of faults, so making this step one and making it it's own section that includes the info that's provided in the hover popup would go a long way in making the setup process much easier.
 
One way to use the limits in the controller is to set the maximum values that you would ever want to use with the battery and motor you have. For example the battery might have a 30 amp limit in the BMS that you don't really want to exceed, or perhaps it has a 50 amp peak rating so you might be willing to set battery current limiting to something in the range of 30 to 50 amps. For the motor current you might want to figure out how much current makes sense to force through the motor at low speeds. Above a point the extra current is mostly making heat (I squared R) and not making much more torque due to saturation. Choose something the motor wiring can actually handle without getting too hot. If that's not enough thrust then consider a smaller wheel, a bigger motor, or second motor, putting more current into the motor above the safe value will mostly just cause it to burn faster... Anyway the idea here is to use the controller for the maximum values you ever want to use, and then use the Cycle Analyst to regulate your actual values down from there. Then you won't accidentally adjust the CA and go higher than you decided was appropriate when you were programming the controller, and hopefully thinking clearly and choosing carefully.

I have a bike with a 9C motor in a 26" wheel, and that just wasn't enough torque for the steeper stuff. Nearly melted it that one day in Marin where we climbed the big hill for the ES gathering. Going from a 12 FET to a PhaseRunner didn't fix that, but the second BMC in the front wheel (in new steel forks) with a second PhaseRunner really did fix that. Now it out accelerates a Cromotor in various real tests that we did, which wasn't really my goal, but it climbs steep stuff without getting hot or bogging down, which was my goal. Two modest hubmotors equals pretty amazing performance. A single PhaseRunner would not be enough to feed the CroMotor (which can handle well over 96A), but using two motors and two PhaseRunners we get a similar amount of performance. If you really want high performance from a single motor then perhaps the PhaseRunner, a 6 FET design, is a bit on the low side. It has a bigger brother that might be a better choice.
 
Thanks. I'll put everything to the max then! 96A for max phase current and regen. 6000w for motor max power.

Hopefully the controller work with this without issue
 
Alan B said:
One way to use the limits in the controller is to set the maximum values that you would ever want to use with the battery and motor you have. For example the battery might have a 30 amp limit in the BMS that you don't really want to exceed, or perhaps it has a 50 amp peak rating so you might be willing to set battery current limiting to something in the range of 30 to 50 amps. For the motor current you might want to figure out how much current makes sense to force through the motor at low speeds. Above a point the extra current is mostly making heat (I squared R) and not making much more torque due to saturation. Choose something the motor wiring can actually handle without getting too hot. If that's not enough thrust then consider a smaller wheel, a bigger motor, or second motor, putting more current into the motor above the safe value will mostly just cause it to burn faster... Anyway the idea here is to use the controller for the maximum values you ever want to use, and then use the Cycle Analyst to regulate your actual values down from there. Then you won't accidentally adjust the CA and go higher than you decided was appropriate when you were programming the controller, and hopefully thinking clearly and choosing carefully.

I have a bike with a 9C motor in a 26" wheel, and that just wasn't enough torque for the steeper stuff. Nearly melted it that one day in Marin where we climbed the big hill for the ES gathering. Going from a 12 FET to a PhaseRunner didn't fix that, but the second BMC in the front wheel (in new steel forks) with a second PhaseRunner really did fix that. Now it out accelerates a Cromotor in various real tests that we did, which wasn't really my goal, but it climbs steep stuff without getting hot or bogging down, which was my goal. Two modest hubmotors equals pretty amazing performance. A single PhaseRunner would not be enough to feed the CroMotor (which can handle well over 96A), but using two motors and two PhaseRunners we get a similar amount of performance. If you really want high performance from a single motor then perhaps the PhaseRunner, a 6 FET design, is a bit on the low side. It has a bigger brother that might be a better choice.
Click to expand...

Dual motor is interesting but too heavy! Do you use 1 throttle for 2 controllers?
 
One throttle feeds both PhaseRunners. This works really well with the PhaseRunner since it it a Torque Request, so both motors adjust to whatever speed and just produce torque based on the throttle position. This is a much better setup than trying to match regular controllers.

In my case the additional weight is not too bad, as it is a geared front motor. Total of both 9C and BMC is less than the Cromotor, and the weight is better distributed across the two ends of the ebike. Two gearmotors would be even lighter. A DD and Gearmotor combination can really offer long life since the gearmotor can be disabled a lot of the time, producing no drag or wear on the gears.

Motor Weights
9C 7kg
BMC V3 4.3kg
Cro V2 12.2kg
 
That's a good idea to have both geared and DD. Maybe something I shall consider! Although I'm dreadful of punctures once I put a motor they tend to gather that!

I also heard that 2wd increase efficiency! So even better
 
zro-1 said:
I'd like to provide a little constructive criticism on the Phaserunner software.
I think the best bet would be to move all this into a help menu available in the software.
Click to expand...

Hi ZrO and thanks for the feedback. Hopefully you found that there is a help menu with all the parameter info Help.jpg

But this:
Alternatively, you could make a little ℹ️ icon next to each setting that you can click on to view the popup content and it stays visible until you click away or close it.
Click to expand...

Is an absolutely great suggestion. I've never been too happy with the tooltip popup behavior that we have available inside QT and a tip that stays persistent after you click on an info icon with an 'x' box to close it would be much better than hovering the mouse over a certain text and waiting.

We're working on a trial implementation of this and if it works as well as it seems it should then that will be part of the next release. And we'll also be flushing out the help menu with more detail and including in the software download the user manual .pdf with a link to that from within the program.

In the meantime we've got an updated download for the MacOS build that now bundles in the necessary FTDI driver and will prompt the user to install them if it detects that the USB->TTL cable is using the wrong one. It turns out that the majority of the issues we've been having recently with Mac users having communication problems during autotune have been related to MacOS updates that are replacing the previously installed FTDI driver with a default driver that doesn't work so well. I'd love to hear if people can try this who'd been having issues getting the MacBook to run autotune. The Mac 1.0.1 is release available here:

www.ebikes.ca/downloads/PhaseRunner_Software_Mac_v1.0.1.zip

Regarding the PLL and Current Regulator bandwidth parameters, what we're really hoping is to get a bit more clarity from ASI and then find a way of having the software provide a suggested set of values for these after the static autotune. At that point, the phaserunner knows the motor winding, inductance, and effective operating eRPM (from the best guess KV value and pole pairs) and should be able to more smartly infer sensible values here.

If we can do that and get it right then it would effectively eliminate the need for most people to touch these values for an operational setup. But if not then I agree that this should warrant a more dedicated help section.
 
Hey Alan, this here:

Alan B said:
One way to use the limits in the controller is to set the maximum values that you would ever want to use with the battery and motor you have. For example the battery might have a 30 amp limit in the BMS that you don't really want to exceed, or perhaps it has a 50 amp peak rating so you might be willing to set battery current limiting to something in the range of 30 to 50 amps. For the motor current you might want to figure out how much current makes sense to force through the motor at low speeds. Above a point the extra current is mostly making heat (I squared R) and not making much more torque due to saturation. Choose something the motor wiring can actually handle without getting too hot.
... Anyway the idea here is to use the controller for the maximum values you ever want to use, and then use the Cycle Analyst to regulate your actual values down from there. Then you won't accidentally adjust the CA and go higher than you decided was appropriate when you were programming the controller,
Click to expand...

Is very well and clearly stated. To the question "what should I set my Phaserunner limits to", above is the answer!
 
justin_le said:
zro-1 said:
I'd like to provide a little constructive criticism on the Phaserunner software.
I think the best bet would be to move all this into a help menu available in the software.
Click to expand...

Hi ZrO and thanks for the feedback. Hopefully you found that there is a help menu with all the parameter info Help.jpg
Click to expand...

Wow, I feel dumb. I totally missed the help menu. It does indeed have everything there. Sorry for not being more thorough in my inspection of the software before I made that criticism.

justin_le said:
We're working on a trial implementation of this and if it works as well as it seems it should then that will be part of the next release. And we'll also be flushing out the help menu with more detail and including in the software download the user manual .pdf with a link to that from within the program.
Click to expand...

That's awesome! I look forward to seeing what you come up with.

justin_le said:
In the meantime we've got an updated download for the MacOS build that now bundles in the necessary FTDI driver and will prompt the user to install them if it detects that the USB->TTL cable is using the wrong one. It turns out that the majority of the issues we've been having recently with Mac users having communication problems during autotune have been related to MacOS updates that are replacing the previously installed FTDI driver with a default driver that doesn't work so well. I'd love to hear if people can try this who'd been having issues getting the MacBook to run autotune. The Mac 1.0.1 is release available here:

www.ebikes.ca/downloads/PhaseRunner_Software_Mac_v1.0.1.zip
Click to expand...

I personally haven't had much (or any) issues with the connection between the Phaserunner and the Mac. The first time I used it I noticed after hitting "save" that the status would briefly show the controller disconnected. But now that I understand more, I assume that this is the controller rebooting with the new parameters so it's normal that it would momentarily disconnect. A little note in the manual or help saying "Hey, when you save settings, you'll see the controller disconnect as it reboots, but don't sweat it unless it doesn't reconnect" would help people using this for the first time with no prior experience in writing parameters to flash memory like that.

justin_le said:
Regarding the PLL and Current Regulator bandwidth parameters, what we're really hoping is to get a bit more clarity from ASI and then find a way of having the software provide a suggested set of values for these after the static autotune. At that point, the phaserunner knows the motor winding, inductance, and effective operating eRPM (from the best guess KV value and pole pairs) and should be able to more smartly infer sensible values here.

If we can do that and get it right then it would effectively eliminate the need for most people to touch these values for an operational setup. But if not then I agree that this should warrant a more dedicated help section.
Click to expand...

[strike]The biggest catch22 here is that autotune can fail with an instantaneous phase error if the PLL and Current bandwidth are too low. SO I don't see how you could rely on autotune to set them unless you have a script that says: if auto-tune fails with this error, bump up the settings, and try again automatically.
Still, if that could happen automatically, that would be awesome. Otherwise, I'd just emphasize it right at the start where you mentioning needing to know your pole pairs and approximate Kv to get started.[/strike]
EDIT: I didn't catch the static test part of that comment. If that could work it would be awesome! In all my trials and tuning, the static test always worked. It was when I tried to run the spinning test that I'd get the error. So if it could offer proposed setting for the PLL and bandwidth that would be super cool.

I will readily admit that 99% of my criticism with the software wasn't due to lack of features or function, but some things not being super intuitive or easy to miss in the documentation. Luckily these things are easier to correct than features or function.

Overall I am super super happy with the Phaserunner. I don't think I'll ever be able to use a different controller on any of my bikes again.
 
I have just received the ASI BAC 800 controller for my Cyclone Motor (from lunacycle). Can anybody tell where I can find the proper cable to communicate with a laptop. Does anybody have a link to the Bacdoor SW or to any compatible SW ?
Thanks Guys !
 
Would adding Bluetooth be in plan? The controller isnt always accessible so when needing to configure I have to unscrew lots of things.

Would be much better with 1 less cable
 
zro-1 said:
Wow, I feel dumb. I totally missed the help menu. It does indeed have everything there. Sorry for not being more thorough in my inspection of the software before I made that criticism.
Click to expand...

No worries, I suppose that's a sign that it's not super obvious there either.

EDIT: I didn't catch the static test part of that comment. If that could work it would be awesome! In all my trials and tuning, the static test always worked. It was when I tried to run the spinning test that I'd get the error. So if it could offer proposed setting for the PLL and bandwidth that would be super cool.
Click to expand...

Exactly that is the hope and I'll be looking forwards to having something in development here. The only other parameters that could then cause the autotune to fail would be the sensorless self startup behavior. That can be trickier for parameter prediction since the controller has no way to know the potential inertia present on the motor, but I've found that the default settings at least for unloaded motors work quite well. A lot of tweaking is needed to have it start sensorless from a standstill on an actual bike through.

Overall I am super super happy with the Phaserunner. I don't think I'll ever be able to use a different controller on any of my bikes again.
Click to expand...

I'm happy to hear that! There is one more major thing we are working on now to finalized the full production standard Phaserunner device, and once that's done we're hoping to then have the resources to expand into a smaller/cheaper lower power version, and then a larger 12 fet model, so that we can better accomodate the broad range of motor power levels used in ebikes.
 
Justin,

You don't know how long I've been waiting for you to mention the 12-fet model :mrgreen: :mrgreen: :mrgreen:
I love the phaserunner - and this is the best controller I have so far. I have yet to find anything else that's potted / waterproof, robust, and reliable. Basically, this is the one controller I never have to worry about.

A couple of humble inputs from me (for your kind consideration) as you're developing this (more like a fine tuning on an already great controller)

1. Please keep the width the same (make it longer versus wider) to accommodate more fets. The current width is perfect for the bike frames (and we can use the frame as a large heat sink for the phaserunners.) The increased length should not be an issue for most bike frames.

2. Integrated display w/ smart phone (Iphone / Android) via wifi / bluetooth would just be perfect!
No more laptop connections. Changing settings on the fly would just be so convenient!
I believe this would also enable the controller to transmit phase amp reading live, versus just the batt amp (which is the limitation when using cycle analyst as it is only able to read from the shunt)

3. More phase amp please (we can keep the peak amp period shorter to ensure reliability - along w/ temp rollback)
The current phase amp limitation is not so much an issue for a mid-drive, but I do find it a bit on the low-side especially w/ a heavy rider or higher KV motor.

4. This is a bit of a stretch - but a 2WD feature would be awesome (where we can adjust the front / rear power.) As you know, I'm running 2WD hubs w/ 2 phaserunners - and they're just so much fun! I just find that the front wheel keep spinning all the time that I have to limit the phase / batt amp for the front hubs.

Many thanks in advance - and really looking forward to this!!!
 
Thanks for that info Justin.

If we’re also writing wishes here, then I’d like to add that I’d love to see the connectors move to higo. The big, non-waterproof 5-pin and 6-pin connectors for the halls and cycleanalyst are a bit of a pain to work with when trying to make a clean build and keep the connections clean and water-safe.
 
district9prawn said:
I recently changed a few sensorless start settings and slightly adjusted current PI loops in Bacdoor. All is well but I've noticed that trying to connect to the phase runner suite running on windows (I tried 0.992 and 1.b3) results in immediate crash. Even with no controller connected, simply importing the xml also crashes. My previous configuration loads up fine but I can't seem to pin down what might have caused this.
Click to expand...

Hey District9 and sorry for overlooking this post earlier. This is a pretty unexpected behavior that we wouldn't mind understanding better. Is it possible for you to email us this .xml settings that you created with bacdoor software but which causes the Phaserunner suite to crash?
 
justin_le said:
Hey District9 and sorry for overlooking this post earlier. This is a pretty unexpected behavior that we wouldn't mind understanding better. Is it possible for you to email us this .xml settings that you created with bacdoor software but which causes the Phaserunner suite to crash?
Click to expand...

Thanks for sending this over. The issue is related to the fast undervoltage fault setting (address 148) and indeed it causes our Phaserunner program to get stuck in a loop. Thanks for finding this bug, and we should have it fixed in the next release build.
 
Glad to see this is still pushing along. How has it been working with ASI so far? My main OEM in the hobby world just changed their terms and margins, making it a very poor partnership. I hope this doesn't turn out similar with ASI going direct to market after you put in so much effort for them.
 
Alan B said:
I have very mixed feelings about the Higo connectors. They look nice but aren't something we can DIY on a piece of cable, as far as I've seen. If they are then where are the DIY kits?
Click to expand...
I also really like the Higo, but find nothing for the DIY fella, and availability and pricing of pigtails is disappointing. Still I like kits with them integrated for ease of build.
 
I own a phase runner and absolutely love it. I recently got a cyclone which was only available with a speed control from Luna. Got the 72v ASI BAC 800 and was sent a controller with zero documentation. The kit is advertised as plug and play as well as pre tuned. The connectors on the controller weren't even compatible with those on the phases of the motor and every wire not used in the 24 pin plug was cut off. I can't even connect e brakes to it. Emailed ASI for a manual and they told me "Sorry to be the bearer of bad news but we don't support products sold by luna"

What a horrible experience and terrible customer support. I can't believe a product manual is not available for the person who buys it.

Justin, your support and manual is priceless!
 
Good thing ebikes.ca sells this!

http://www.ebikes.ca/downloads/BAC500_Controller_Manual_Rev%201.0.pdf

http://www.ebikes.ca/c-bac2000.html
 
Lebowski said:
Making the controller parameters motor independent is no big deal, and it (well, mine) doesnt need to know motor inertia....
Click to expand...

Ah sorry I should have clarified, not motor inertia but the overall total vehicle inertia. At least the way that ASI has written their sensorless self start firmware it runs open loop at a steadily increasing RPM before switching over to closed loop control. So if it tries initially to ramp up the speed faster than the motor torque can actually accelerate the vehicle mass, then it will tend to shudder.

There are surely many more optimal sensorless self starting algorithms possible! But this is what we've got to work with here. Would love it of course if Lebowski was head of ASI's firmware team ;)
 
johnrobholmes said:
Glad to see this is still pushing along. How has it been working with ASI so far?
Click to expand...

Proof will be in the pudding within the coming month. We've still been having a challenging time trying to keep up with demand manufacturing the Phaserunner assemblies in house from the raw PCBA's provided from ASI. Over the last 6-8 months we've worked hard to transition things so that a majority of the sub assemblies and QC testing on the next batch is done by ASI's contract manufacturing partners. We've shipped them all the mechanical parts we make here and the first samples from that are due to be sent for our verification very soon, like next week!?. I'm excited.

If it all pans out it will mean good things for our ability to offer wholesale / oem level partnerships and scale things to accommodate high volume orders and make this controller much more widely available.

johnrobholmes said:
My main OEM in the hobby world just changed their terms and margins, making it a very poor partnership.
Click to expand...

Sorry to hear that. Fortunately there are many more market segments within the ebike world than the hobby / RC world, and support expectations are a lot higher too. That's a pretty large barrier to the less savory business models actually having long term success.
 
You must log in or register to reply here.

Similar threads

G
Phaserunner replace PMAC controller?
Replies
0
Views
472
goatman
G
thorlancaster328
Designing the ultimate (400A+) motor controller - what features should it have?
Replies
11
Views
1,809
SammyMc
S
A
50KW controller with field weakening
Replies
10
Views
1,378
Lebowski
Lebowski
owhite
    • Like
  • Locked
building the BESC - 100V, 400A motor controller [NOT RECOMMENDED]
2 3 4 5
Replies
120
Views
19,960
owhite
owhite
markz
Hacking Controller LVC
Replies
5
Views
2,369
jordanjozsef
J
Back
Top