Q100H Performance with Grin/Infineon Controller Issue

[Kevin Crawford]

Hi all,

I'm finishing up my 2nd ebike conversion. Per information found here, I replaced the original motor cable attached to the controller with the one that came with my Q100H motor, and matched the phase and hall wire colors to what was originally wired in, which was a cable for an eZee motor. The white wire, I left disconnected. The controller is the one that came with an eZee V2 kit, about three years old.

I had tested the motor in the wheel on a bench first. It seems to take quite a while to spin up, maybe 10 seconds for full speed, and makes a varying "whine" sound the whole time it's spinning. I wasn't sure this was a problem, so I continued the build.

It wasn't until I had the wheel in the bike when I realized the rotation is backwards. At least, it's backwards if the side of the hub with disc rotor bolts is on the non-drive side, which I thought was the rule.

Fortunately, it's a front wheel, so I just flipped it over. Yeah, the tread on the tire is backwards for now, but I figured I'd worry about that later.

Trouble is, I think I have a bigger problem. The motor has no power whatsoever. If there's any assist at all, I can't tell. I might as well just ride my road bike. I have a Cycle Analyst, and I'm using a 48V battery. I have the current limited to 5A so my battery pack lasts longer. With my eZee motor and the same parts, 250W was still a pretty decent assist. With me on the bike, this doesn't get anywhere close to the 5A limit. It gets to about 2.5A.

So I have to think something is wired or programmed wrong after all. I'm not expecting the amount of power my eZee had at maximum, but this motor can't possiby suck this much. I've checked the resistances of the windings all the way to the controller - as far as I can tell, they're zero ohms phase to phase, which seems low, but I'm thinking that must be normal. I've also checked the operation of the hall sensors at the controller. All three are working fine. I've checked the output signals to the phases in the controller - at max throttle, I get 42 VAC between every two phases.

So, it seems the controller and halls are working, and that the wiring (including the windings) is all intact. I'm thinking I must have to experiment with the phase and/or hall wiring configuration. Now I wish I had purchased an extra cable, because I could have made a nice testing and troubleshooting jig inline with it, instead of having to unsolder and resolder wires.

So, for those of you with experience, do I seem to be on the right track here? The only other thing I can think of is adjusting settings in the CA, but I skimmed through that and nothing stood out to me.

Any advice would be appreciated. And if any of you know where to get an extra motor cable in the US, or have one to sell, that would also be appreciated.

 

[Jon NCal]

The motor going backwards is a big clue that it's wired incorrectly. Start over and get the phase and halls combinations figured out properly.

 

[Kevin Crawford]

You're right. Flipping the wheel over was a mistake, because it's not made to spin the same in both directions like a regular front wheel.

Thanks. Now, to find that handy flowchart...

 

[leo99]

There is a link to it at the end of my build that you can see on my signature

 

[motomech]

First off,the Q100 has a sprage(ramp and roller)clutch and will not function "backwards". The motor just spins internaly.
Take the CA and the silly 5 Amp limit out of the system and test.
Are we talking about this controller?
http://www.ebikes.ca/shop/ebike-parts/controllers/c3620-nc.html
You don't have the forward/rev. connector connected and running the motor forwards w/ the controller's rev. mode?
All wires should color match, or they do not, the Yel/blu are interchanged(yel w/ blu, blu w/ yel)on both the phases and halls.

 

[Kevin Crawford]

Your Q100 may not run backwards, but my Q100H does run backwards the way it's currently wired up. I'd happily share the video I took of it, if I could upload that large of a file.

No, that isn't the controller I have. Mine does look similar - I think it may be an earlier version. It doesn't have a power switch or a reverse mode.

Why would I remove the "silly" current limit?

 

[d8veh]

The Q100s can be a bit fussy about which controllers they work with. I would advise anybody only to use the ones that are known to work with them.

It's not possible to run a Q100 backwards because it has a one-way clutch in it. The motor will spin, but it can't get any power to the wheel. You might see the wheel rotate in the air, but you'd be able to stop it with one finger.

If it is running backwards, there's something wrong with it, like the clutch is rusted stuck, which is a common problem on those used in the rain then left for a while. Is it a new motor? The clutch is symmetrical, so it can be flipped to reverse the direction. I've never heard of one coming from the factory like that.

I would say that that Infineon controller isn't the best for a Q100. You need to add a Cycle Analyst to get the PAS function. If you don't want PAS, you probably got the wrong motor. With the CA, it's going to work out expensive. I'd normally use one of those sine-wave Kunteng controllers with the nice LCDS. They're more or less plug and play and they have very nice PAS control, which is the main reason I recommend them, plus, they're not too expensive.

 

[Kevin Crawford]

That makes sense. The wheel does spin backwards in the air, but after flipping it over, there's basically no power.

It is a new motor. Now that I've seen how to open it, I'm going to do that once I have the wheel out next time and see if I can put a layer of grease between the cover and the chassis. I did that with my eZee (also did the same for the cover and rotor screw threads), and it solved the problems I'd have when riding in the rain.

I do have a CA already, but I don't use PAS. I use the CA for current limiting and the cruise feature. If I were to ever replace my controller, I'd look for one that has those features built-in. Does something like that exist already?

Thanks once again for all the info.

 

[d8veh]

I do have a CA already, but I don't use PAS. I use the CA for current limiting and the cruise feature. If I were to ever replace my controller, I'd look for one that has those features built-in. Does something like that exist already?

Thanks once again for all the info.

Yes. The Kunteng sinewave controllers are very good. The PAS function is excellent. They have some programmable current limit, but with PAS, they have a primary function of current limit for each PAS level. You can get them from BMSBattery (S06S) and many suppliers on Aliexpress. Search for Sine wave controller. The LCDs show power, speed, distance, battery voltage and give access to the programmable parameters.

 

[Kevin Crawford]

Hey everyone:

After experimenting with swapping wires around, I was able to get the motor running in the correct direction. It is "working", but the trouble is, the power is sorely lacking. There is some assist, but it's so little that it doesn't seem worthwhile to even use vs. just hopping on my road bike. I was expecting more. I only have my eZee to compare to, but even at just 250W (48V / 5A), the eZee had a nice level of assist. With the Q100H, even at 48V / 7A, there seems to be very little torque, and I'm not interested if I have to bump up the current much more to get decent assist. I'm not certain the rotation is "smooth" according to the wiring flowchart I've seen floating around here (it is more noisy than my eZee), but based on some videos I've seen, I don't think it's wired wrong, either. It makes a buzzing sound that I would call moderate. It does seem to spin fine in a work stand, and I'm certain the rotation is correct now, but there's just not much torque on the road. I'm about 180 lbs, and the bike I'm using is not heavy, and if I don't pedal at all the motor *might* get me up to 10 MPH, but it seems even less than that.

It's worth noting that I was using the eZee on a 26" wheel, where the Q100 is in a 700C wheel. I can see that making a small difference.

Am I simply expecting too much, or might there still be an actual problem with wiring? I know at least one user here seems to have had success with this controller (Infineon) and motor combo, so I'm not keen on just trying another controller. So far I'm regretting selling my eZee wheel, and I'm about to the point of selling this wheel and buying another eZee (anyone interested?).

Any ideas would be greatly appreciated. Thank you.

 

[neptronix]

7A is weak. Of course there is barely any assist. Bump it up to 15A at a minimum.

 

[leo99]

Hi Kevin,
I am running the infenion controller on a Q100 cst 201 at 52v and 15 amps. Fully charged I hit 22mph without pedaling with a 26 inch wheel. Either you didn't wire it properly, or you are not running enough amps.
My controller colors were completely mixed up with my motor when I found a good combo.
The ebike simulator says you should hit around 17mph even with 7 amps and a 201 Q100 at 48v
Leo

Edit: there is only 3 combo that work out of the 36, I had a few that were going in the good direction but the sound wasn't terrible when I found the correct combo you could here a big difference in noise.

 

[Kevin Crawford]

neptronix:

LOL, I was kind of expecting a response like that. With the amount of riding I do (100 miles per week), and in the conditions I ride in, I just don't see my battery pack lasting long enough to make using this motor worthwhile. I squeaked out 2 years averaging about 8A before.

At 15A, any battery pack > 24V would be exceeding spec. At 36V, over 10A would be exceeding spec. I just don't see how this motor could be very useful in these conditions.

leo99:

Thanks for the info. Maybe I'll try swapping wires around some more. I've been doing it by sodlering/unsoldering inside the controller, and it's a major hassle. I should have made up a terminal strip inline with the cable from the beginning. Can you describe how the motor is supposed to sound when the combo is correct?

 

[leo99]

It's hard to describe but it's slightly growly in the beginning and at full speed it's very s high pitched but less loud and when you are riding it The bike makes more noise itself then the motor when at max speed on flat. I also so had to swap the phase wire not only the sensors. There is a pic of my wiring on my build, I used quick electronic attaches to find the combo before soldering it.
Do you know your no load speed? You should hit 25mph when the wheel is off the ground.
Leo

 

[amberwolf]

If you look at the ES wiki article on how to determine the wiring for a brushless motor (or the various threads for same) it describes how to tell via the no-load current being used, among other things.



FWIW, since geared hubs (with internal freewheels/clutches) can only pull in one direction, if the rotation is the wrong direction, the motor will spin but it will not drive the hub. This will feel like no detectable assist.

To test for that start the throttle on the bike, wheels on ground, but with you not on it. If you hear the motor spin but it cannot move the bike immediately (or at all) then the motor is still spinning the wrong direction (or the clutch inside the motor is not working right and is allowing it to spin freely in both directions--to test for that try to spin the wheel by hand; it will have resistance in the "wrong" (backwards) direction, but not in the "right" (forwards) direction).

If the motor will move the bike wthout you on it but not with you on it, then something else is wrong, either controller settings or controller itself, or more likely just the order of phase and hall wires.

 

[d8veh]

It should still make quite good power at 48v and 7A. Something is wrong.

 

[Kevin Crawford]

Thanks. I'm going to experiment with wiring combos one more time. I like the quick-disconnect idea. Once I get that all set up running through all of the combinations shouldn't take long at all. I could even test it in real (on the street) conditions easily this way, to really nail down the combinations that work. There's a bit of slack in the cable anyway, so once it's finalized I'll solder and heatshrink the wires and finally the cable neatly.

In hindsight, ordering an extension cable would have been best, as I could have converted it to a test cable with the quick-disconnects and simply remove it when finished.

 

[molybdenum]

Not sure if you saw this in the es wiki, but the chart might help keep track of all the hall/phase combinations as you test systematically:

https://endless-sphere.com/w/index.php/Determining_the_Wiring_for_a_Brushless_Motor

 

[d8veh]

One more thing: Q100 motors are very picky about what controllers they like to work with. If it's an older Infineon, it probably won't work.

 

[Kevin Crawford]

Well, that's more like it! I think I finally have this configuration working properly now. I cut off both halves of the waterproof connector, and set up all of the wires on a terminal strip for testing. I didn't want to have another break in the cable, and I didn't want to go through the tedium of unsoldering and resoldering on the controller again. Following the flowchart I found here, I thought the rotation may have been "noisy", so swapped around all three of the phase wires as indicated. I couldn't tell you what the final overall configuration that worked is now, because I had already swapped around some wires inside the controller, and don't want to go through the hassle of putting them back.

I could immediately tell that the motor was running better. It doesn't sound "sick" like it did before, and I could tell there was more power even before putting the bike on the ground. I just tested it on the road briefly, and it's running just about how I expected. The assist level is pretty good for my taste at 48V / 7A.

I just ordered a new 48V 10Ah pack from Grin - luckily, on sale! It seems they may be phasing out this version of the pack for ones with higher capacity cells, but those are more expensive and 10Ah is fine for me. I'm going to be doing some more testing, but I'll need to get some permanent connectors for the motor wires - I'm thinking something like 5-pin JST for the halls, and I may just do color-coded insulated disconnects for the phase wires. I just cut away the white wires, which I don't use.

Once again, thank you very much for all the advice.

 

[quuxman]

I am running the infenion controller on a Q100 cst 201 at 52v and 15 amps.

This is good to hear. I've been running my Q100H at the recommended 36V and ~20A for a few years. I'm seriously considering upgrading the controllers and batteries on both my ebikes to 52 volts with programmable power limits. It would be really nice to have interchangeable parts, and a little more juice with the Q100H bike. I'm just worried about destroying the motor. How long have you been using yours 52V / 15A? Have any problems come up?