Back on the original question, for a sensorless controller, there will be 3 forward and 3 reverse combinations for the phase wires. It won't make any difference which one you use.
Sensorless Sync Issues (was Phase Combinations, Aotema)
- Thread starter Beachcruzer
- Start date
Alan B
100 GW
Sorry, thought it was geared.
Regardless, we know there is no problem with the phase wiring. It turns the right direction, done.
So the next issue is the 32 mph noise and loss of power. This indicates the controller has lost synchronization with the motor's position. Discuss this with the controller provider. Sensorless controllers need to be configured to match the motor parameters. There may be adjustments that can be made. If not you may need to get a different controller.
Good luck,
Regardless, we know there is no problem with the phase wiring. It turns the right direction, done.
So the next issue is the 32 mph noise and loss of power. This indicates the controller has lost synchronization with the motor's position. Discuss this with the controller provider. Sensorless controllers need to be configured to match the motor parameters. There may be adjustments that can be made. If not you may need to get a different controller.
Good luck,
dogman dan
1 PW
Probobally the controller losing sych. Personally I'm not sure sensorless is ready for high rpm's. I had a similar problem with a 40 amp 72v sensorless controller. I ended up trading it in for a sensored one. Then I got a sensored motor and kept the aotema one for use with the stock 48v controller.
Beachcruzer
100 W
dogman said:
That may be where I'm going too Dogman. My plan all along has been a rolling upgrade to a higher power rear-drive system. By the time that's all put together, I'll have the original 48v front-drive Aotema back on the basket cruiser. It's a great around-town bike.
I hesitated about the forward compatability of the sensorless controller, but the vendor told me the next-gen 9C and Crystalyte motors will be available in sensorless versions. If I'm going to run into sync issues at ~35 mph though, I may want to reconsider. I'm going to dial down the controller amperage and see if that helps.
Thanks everyone for your advice.
Beachcruzer
100 W
NeilP said:
That did it Neil. Thanks mate. Pole count out of the box was set to 23 for some reason. Reset it to 1, and speedo function is now working. New harness (18s 2p) is all soldered up, so looks like a speed run in the morning.
NeilP
1 GW
Yea, I made that mistake to , I set my pole count to that of the motor..
How it works is this...if you do not have a remote speed sensor, the speed function works from motor rotations, so it needs to know the number of poles/ pole pairs in the motor..and it works the rotation from the pulses it receives from the motor. If you have the remote sensor, then it is just one pulse per rotation
How it works is this...if you do not have a remote speed sensor, the speed function works from motor rotations, so it needs to know the number of poles/ pole pairs in the motor..and it works the rotation from the pulses it receives from the motor. If you have the remote sensor, then it is just one pulse per rotation
Beachcruzer
100 W
Okay, had the bike out again this afternoon, running the Lyen 12 fet sensorless edition, 75 volts lipo, and the Aotema hubby.
I did a no-load test when I plugged in the batteries. It spooled up to 60.9 mph indicated on the CA. Smooth, no troubles.
During the test ride, however, it exhibited the juddering/loss of sync at about 33 mph. With a bit of a downhill and careful throttle control, I managed to record a max speed of 38.4 mph. Hitting the throttle hard on an uphill, I made it lose sync at around 26 mph. So the issue seems to be related to both load and rpm.
I found a thread on this problem from a while back. Symptoms, on a 9C, sound pretty much identical to mine (see below.) There's a trick with reversing the phases and soldering something else that's supposed to improve but not solve the problem. I'm not really interested in half-measures, though. Wondering if anyone has come across a good solution or tuning protocol for this problem? I've PM'd Lyen as well.
From this thread: viewtopic.php?f=28&t=22333&start=75
I did a no-load test when I plugged in the batteries. It spooled up to 60.9 mph indicated on the CA. Smooth, no troubles.
During the test ride, however, it exhibited the juddering/loss of sync at about 33 mph. With a bit of a downhill and careful throttle control, I managed to record a max speed of 38.4 mph. Hitting the throttle hard on an uphill, I made it lose sync at around 26 mph. So the issue seems to be related to both load and rpm.
I found a thread on this problem from a while back. Symptoms, on a 9C, sound pretty much identical to mine (see below.) There's a trick with reversing the phases and soldering something else that's supposed to improve but not solve the problem. I'm not really interested in half-measures, though. Wondering if anyone has come across a good solution or tuning protocol for this problem? I've PM'd Lyen as well.
From this thread: viewtopic.php?f=28&t=22333&start=75
Spacey said:
feyris
100 mW
Hi.
I have the same probleme here.
Same controller than you, my motor is a geared freewheeled sensorless Bafang motor.
I still hope for a Fix but I think I will change my motor to a sensored one.
Any idea on how to turn the sensorless controller into a sensored one? I don't think just solder sensor wires will be enough...
I have the same probleme here.
Same controller than you, my motor is a geared freewheeled sensorless Bafang motor.
I still hope for a Fix but I think I will change my motor to a sensored one.
Any idea on how to turn the sensorless controller into a sensored one? I don't think just solder sensor wires will be enough...
o00scorpion00o
100 kW
Same problem for me with Lyen sensorless and magic pie.
Snellemin is using the same controller and a pie and no issues.
It's not a phase issue for me, Only happens on full throttle and it makes a really loud noise. I've no idea if it's linked to the time I got the motor too hot that burned the halls, it could be. I doubt it's a controller issue.
It makes a surge of power than always at a certain speed it pulls back while making that awful noise and a bad vibration. The whole bike shakes violently, but if I ramp up the throttle slowly it is fine.
Strange thing is that after I ride at full speed for a while and stop, I can full throttle no problem and there is no noise and vibration. But if I then ride a few miles by cycling and no throttle the noise and vibration come back! It's very annoying and I wish I had another pie to test ! It could very well be due to the fact it got too hot and melted the halls.
Snellemin is using the same controller and a pie and no issues.
It's not a phase issue for me, Only happens on full throttle and it makes a really loud noise. I've no idea if it's linked to the time I got the motor too hot that burned the halls, it could be. I doubt it's a controller issue.
It makes a surge of power than always at a certain speed it pulls back while making that awful noise and a bad vibration. The whole bike shakes violently, but if I ramp up the throttle slowly it is fine.
Strange thing is that after I ride at full speed for a while and stop, I can full throttle no problem and there is no noise and vibration. But if I then ride a few miles by cycling and no throttle the noise and vibration come back! It's very annoying and I wish I had another pie to test ! It could very well be due to the fact it got too hot and melted the halls.
feyris
100 mW
o00scorpion00o said:
That's exactly my issue.
It appends during acceleration, or during climbing. It's very annoying, and can be dangerous (no power when you just jump in a traffic circle...)
I don't think it's coming from your motor because of burned sensors, mine is fine and never had sensors.
o00scorpion00o
100 kW
feyris said:o00scorpion00o said:
That's exactly my issue.
It appends during acceleration, or during climbing. It's very annoying, and can be dangerous (no power when you just jump in a traffic circle...)
I don't think it's coming from your motor because of burned sensors, mine is fine and never had sensors.
Yeah it's not because of the halls, but I'm wondering if overheating it did any other damage ?
I wish I had another Pie to test.
Here is a link to a video I posted in a thread I created about the noise.
[youtube]1aqby150u2g[/youtube]
I'm running 24s1p lipo now with the 12fet Lyen controller and Pie2. With the Pie it maxes out at 45mph. With better batteries, I suspect a higher top speed.
50% speed setting 27~28mph, with super smooth throttle. This is the mode I commute with.
75% speed setting mid~high 30's, with ok throttle.
100% speed setting, I need to ease onto the throttle. If I don't, the controller cuts off. I don't hear any noises or experience any vibrations. The power just cuts off.
I thought it was my throttle at first, and swapped in a different one to verify. Same thing happened. I messed around with the settings in the CA and controller and got a better throttle response. Conclusion is that I'm just pulling way to many amps under load and the controller just cuts off. Looks like I would need a 18fet version if Lyen even makes one, but if he did I wouldn't have the space to mount it on my bike.
Second option is to use current controlled throttle with the help of the CA. But I'll just live with the problem for now, until the new CA becomes available.
On 15-16S lipo, I get super smooth throttle. But I wanted more range. Instead of going with 2P lipo configuration, I opted for higher voltage and reduce the throttle setting. I get about the same range as I would with a 2p pack running the same 27mph speed. I thought the controller would overheat at the 50% setting during my 5.5 mile commute to work, but it never happened.
50% speed setting 27~28mph, with super smooth throttle. This is the mode I commute with.
75% speed setting mid~high 30's, with ok throttle.
100% speed setting, I need to ease onto the throttle. If I don't, the controller cuts off. I don't hear any noises or experience any vibrations. The power just cuts off.
I thought it was my throttle at first, and swapped in a different one to verify. Same thing happened. I messed around with the settings in the CA and controller and got a better throttle response. Conclusion is that I'm just pulling way to many amps under load and the controller just cuts off. Looks like I would need a 18fet version if Lyen even makes one, but if he did I wouldn't have the space to mount it on my bike.
Second option is to use current controlled throttle with the help of the CA. But I'll just live with the problem for now, until the new CA becomes available.
On 15-16S lipo, I get super smooth throttle. But I wanted more range. Instead of going with 2P lipo configuration, I opted for higher voltage and reduce the throttle setting. I get about the same range as I would with a 2p pack running the same 27mph speed. I thought the controller would overheat at the 50% setting during my 5.5 mile commute to work, but it never happened.
Lebowski
10 MW
high e-rpm combined with high torque means there's a large voltage across
the inductive part of the motor impedance. In effect you need lots of timing advance
to get the timing correct again. My controller IC has automatic timing advance but
it'll be a while before I can start selling controllers (need to design a PCB for my
setup....)
the inductive part of the motor impedance. In effect you need lots of timing advance
to get the timing correct again. My controller IC has automatic timing advance but
it'll be a while before I can start selling controllers (need to design a PCB for my
setup....)
o00scorpion00o
100 kW
I'm running very high current on Lyens modded 12 fet. But it was the same before I modded it!
I'm dumping 5kw into the pie, just over 80 amps on acceleration, 120 peak. the torque is incredible. 16S LiPo only 30 mph max.
It's a hell of a lot of fun in the woods where the grades are really steep, the torque of the pie really shines there, but otherwise on most normal roads and hills you wouldn't need so much torque.
Snellemin
I was thinking of going to higher voltage, 24S hot is the limit of the controller and I would be afraid of going to limit.
16S with halls took me to 35 mph, why don't you consider using the halls and a sensor controller ?
BTW I never had any issues with the pie using halls, and I'm not bothered to wire them up again to find out if sensorless really is the issue because I hope to have the mac going again in a few weeks, 40 mph on 16S is good and the 8T has plenty of torque.
I need to build another full suspension bike with a pie or cro motor for the woods and for other abusing trails I might be able to find.
I'm dumping 5kw into the pie, just over 80 amps on acceleration, 120 peak. the torque is incredible. 16S LiPo only 30 mph max.
It's a hell of a lot of fun in the woods where the grades are really steep, the torque of the pie really shines there, but otherwise on most normal roads and hills you wouldn't need so much torque.
Snellemin
I was thinking of going to higher voltage, 24S hot is the limit of the controller and I would be afraid of going to limit.
16S with halls took me to 35 mph, why don't you consider using the halls and a sensor controller ?
BTW I never had any issues with the pie using halls, and I'm not bothered to wire them up again to find out if sensorless really is the issue because I hope to have the mac going again in a few weeks, 40 mph on 16S is good and the 8T has plenty of torque.
I need to build another full suspension bike with a pie or cro motor for the woods and for other abusing trails I might be able to find.
If I were to use sensored, timing advance and 120% throttle settings, I would need to upgrade the motor wires coming out of the motor for max performance. Doing that never really interested me, as I would also need to run 2P lipo packs as a minimum. I'm basing all this from my experience running RC toys at max performance.
I prefer High Voltage Low Amperage setups vs the other way around for speed. I can get away with smaller gauge wires and smaller connectors that way.
This is just my preference.
I prefer High Voltage Low Amperage setups vs the other way around for speed. I can get away with smaller gauge wires and smaller connectors that way.
This is just my preference.
o00scorpion00o
100 kW
Snellemin, I'm going to reply on your magic pie thread rather than hijack this thread!
Beachcruzer
100 W
Not a hijack at all. Scorp's symptoms with the sensorless controller and Magic Pie are pretty much identical to what I'm getting with the Aotema. I can feather the throttle a bit and squeeze a little more speed out of it, but basically by the time I'm going 30 mph it's slipping in and out of sync.
This seems like a fairly common problem when pushing a lot of amps through sensorless systems. Would love to hear tuning advice from anyone who has solved this issue, but I'm thinking Lebowski nailed it: It's just too much power for the sensorless motor/controller combo.
I'm going to go back to the stock 35 amp controller at 48 volts with this motor, and save the high-voltage packs for a sensored motor. So looks like I'm in the market for a new motor and controller.
This seems like a fairly common problem when pushing a lot of amps through sensorless systems. Would love to hear tuning advice from anyone who has solved this issue, but I'm thinking Lebowski nailed it: It's just too much power for the sensorless motor/controller combo.
I'm going to go back to the stock 35 amp controller at 48 volts with this motor, and save the high-voltage packs for a sensored motor. So looks like I'm in the market for a new motor and controller.
o00scorpion00o
100 kW
What I don't understand is why snellemin's has none of the symptoms I have running sensorless ?
I'm going to try 24S today, back off the current and see what happens!
I'm going to try 24S today, back off the current and see what happens!
o00scorpion00o
100 kW
Okay, I found out the problem!!!
I spent half the day tuning the controller phase and battery currents and going for test rides.
No more mental vibration or noise.
I configured my 16S to 24S and it flew to 40 mph, the acceleration is good and throttle much smother!
Maybe I could have went faster with higher currents, I'll rise the phase currents tomorrow as I've no more time for testing and my batteries need changing.
I beat the 5ah zippy a little hard and they were fairly warm when I got back.
The acceleration to 40 mph is incredible at much higher currents but I get the noise and vibration back.
So I think what is going on is there is some kind of current limiting going on even with the controller modded, I bet it still knows I'm pulling more than 3kw. And just below 3kw is where the noise and vibration stops.
I thought if there was limiting going on that it would limit by current and not total watts ?
I would need at least 10ah to keep the zippy's happy, but the range was good enough even on 5ah. I got 4.6ah before the controller lvc kicked in. Probably would have got 4.8 by going slower.
I'm glad I tried 24S, with halls I could have probably seen close to 50 mph, too fast for me so 40-41 is good enough. It's amazing how the mac can do all that and have great acceleration all from 16S.
Next week I'll see what happens on 16S if the vibration kicks in over 3kw, we're going away for the weekend.
I spent half the day tuning the controller phase and battery currents and going for test rides.
No more mental vibration or noise.
I configured my 16S to 24S and it flew to 40 mph, the acceleration is good and throttle much smother!
Maybe I could have went faster with higher currents, I'll rise the phase currents tomorrow as I've no more time for testing and my batteries need changing.
I beat the 5ah zippy a little hard and they were fairly warm when I got back.
The acceleration to 40 mph is incredible at much higher currents but I get the noise and vibration back.
So I think what is going on is there is some kind of current limiting going on even with the controller modded, I bet it still knows I'm pulling more than 3kw. And just below 3kw is where the noise and vibration stops.
I thought if there was limiting going on that it would limit by current and not total watts ?
I would need at least 10ah to keep the zippy's happy, but the range was good enough even on 5ah. I got 4.6ah before the controller lvc kicked in. Probably would have got 4.8 by going slower.
I'm glad I tried 24S, with halls I could have probably seen close to 50 mph, too fast for me so 40-41 is good enough. It's amazing how the mac can do all that and have great acceleration all from 16S.
Next week I'll see what happens on 16S if the vibration kicks in over 3kw, we're going away for the weekend.
o00scorpion00o
100 kW
snellemin said:
Cheers, yeah was driving me crazy that's for sure!
The controller has me puzzled now though, after modding it and soldering the shunt you would think that it could no longer limit you to around 3kw ?
I might E-mail lyen and ask him.
But all I can say is the acceleration at about 4kw on 24S is really addictive, until that is, the vibration starts, but when you back off the throttle and increase it again slower, rather than hit it wot from start, you can feel the surge of power.
But what I think would be needed is for starters more ah or nanotech, and an 18fet, would probably do the job.
Only thing is while the sheer torque at 5kw is brutal, on 16S, I need it for the woods, I wouldn't be able to climb 30% grades at with only 35 amps, instead of 80!
a 24 fet controller would be needed for such power levels 90 volts x 80 amps would be 7200 watts, I doubt the pie could take that continuously, however it's only the woods that would take it to the limits, and the hills are not too long so probably wouldn't matter.
I would like to take it up the 16% mountain that killed my halls, maybe the higher voltage would be better and not allow it to bog down as much ?
Maybe I will test it, but would need more ah.
Beachcruzer
100 W
Success! With some input from Lyen I think I've finally got this thing tuned properly.
First, I programmed the controller for 30 amps rated/50 amps phase, as Lyen suggested. That should have reined the power output back to the low-2000 watt range, but it had little discernible effect--the controller was still pumping around 5000 watts at WOT (72 volt lipo pack), efficiency was 107 wh/mile, and it was losing sync at high speed.
Then Lyen told me to change the RShunt value in my Cycle Analyst from 1 mOhm to 2 mOhm, and amp limit from 80 to 100. (It doesn't go up to 100, so I set to 99). I did that, and the controller magically calmed down. Peak wattage is a more reasonable 2,500 or so, efficiency is 45 wh/mile, and no sync issues. Wound up to 40 mph with no problems.
I have zero clue why the controller didn't function as programmed until the CA was set properly, but there you have it.
First, I programmed the controller for 30 amps rated/50 amps phase, as Lyen suggested. That should have reined the power output back to the low-2000 watt range, but it had little discernible effect--the controller was still pumping around 5000 watts at WOT (72 volt lipo pack), efficiency was 107 wh/mile, and it was losing sync at high speed.
Then Lyen told me to change the RShunt value in my Cycle Analyst from 1 mOhm to 2 mOhm, and amp limit from 80 to 100. (It doesn't go up to 100, so I set to 99). I did that, and the controller magically calmed down. Peak wattage is a more reasonable 2,500 or so, efficiency is 45 wh/mile, and no sync issues. Wound up to 40 mph with no problems.
I have zero clue why the controller didn't function as programmed until the CA was set properly, but there you have it.
o00scorpion00o
100 kW
Beachcruzer said:
Glad you got it sorted!
Do you have a program cable for the controller ? It would be very helpful for me if you could program your controller for 44 amps, at 72 volts you should be pulling about 3000 watts. I would be very curious if you have the noise and vibration I get at around 3kw!
If you don't feel comfortable doing that then that's o.k.
You should get a magic pie for 40 mph, it's got super torque and would probably stay a lot cooler than the 9C!
Beachcruzer
100 W
Yes now that it's working by dialing back the current I'm going to slowly increase the current to basically get it to the bleeding edge of power before it de-syncs. I'll try your suggested numbers next and let you know how it works out.
Test rode the 2810 9c yesterday on another ES members bike. Was very impressed with torque and hill climbing. Top speed is a little less but I think I'm sold. Maybe I need to ride a pie with external controller to get a good comparison.
Test rode the 2810 9c yesterday on another ES members bike. Was very impressed with torque and hill climbing. Top speed is a little less but I think I'm sold. Maybe I need to ride a pie with external controller to get a good comparison.
