QS 3000w midmotor

Electric Motors and Controllers
User avatar
PITMIX   100 W

100 W
Posts: 175
Joined: Nov 18 2018 3:25pm
Location: France

Re: QS 3000w midmotor

Post by PITMIX » Jan 16 2020 2:24pm

Yes Speed 1 without Fw.
Speed 2 with Fw.
Speed 3 with Fw.
The engine runs empty without any transmission.
Aixam 550 Twin retrophite Qs 138 70H, Thermo-Nucular electronics 24F, LG Chem 20s 60Ah

User avatar
PITMIX   100 W

100 W
Posts: 175
Joined: Nov 18 2018 3:25pm
Location: France

Re: QS 3000w midmotor

Post by PITMIX » Jan 17 2020 12:25pm

1579281981229.jpg
1579281981229.jpg (230.18 KiB) Viewed 1910 times
Hi, I installed a temperature sensor on the motor winding. I disassembled the front flange and passed the probe through the silicone right next to the wire for the hall sensors. Then I fixed the probe with a plastic collar on the winding. There is plenty of space on this side of the engine, it's very easy to do. Don't hesitate to add a probe to your engine.
Aixam 550 Twin retrophite Qs 138 70H, Thermo-Nucular electronics 24F, LG Chem 20s 60Ah

twikis   10 mW

10 mW
Posts: 26
Joined: Nov 24 2015 7:25pm

Re: QS 3000w midmotor

Post by twikis » Jan 17 2020 8:49pm

Interesting. To get faster and more accurate readings (mounted like this, the reading will always be at least a few degrees under the actual copper temp when the motor gets hot), I'd glue it to the windings with a blob of any (high temp) 2 component glue, + maybe remove the rubber cap if possible.

Also, unless you used high temp zip ties, I'd really keep the temperature below 80-100C now...

Thanks for the pic of the ingredients!

edit: The airgap seems to be quite narrow, that's nice
Last edited by twikis on Jan 17 2020 10:18pm, edited 1 time in total.

twikis   10 mW

10 mW
Posts: 26
Joined: Nov 24 2015 7:25pm

Re: QS 3000w midmotor

Post by twikis » Jan 17 2020 10:18pm

Ok here's one more question: Has anybody done a (more or less) direct comparison between running this motor with an IPM optimized (e.g. Votol, ASI, maybe Nucular, haven't found any concrete info) controller, that shifts the field around to utilize the extra torque-per-amp (and thereby efficiency and power) archieved by the reluctance, and non-optimized, standard SPM controller like a Kelly KLS?

Comparing the no load consumption is probably the most accurate value to get, short of putting both setups on a dyno...

larsb   1 MW

1 MW
Posts: 1617
Joined: Dec 10 2014 5:12am
Location: Gothenburg, Sweden

Re: QS 3000w midmotor

Post by larsb » Jan 18 2020 2:01am

IPM optimized (e.g. Votol, ASI, maybe Nucular, haven't found any concrete info) controller,
None of those controllers are IPM capable or have MTPA algorithm for IPM motors, it requires dynamic phase angle adjustment to the specific motor. APT96600 might work but the field weakening setting in my controller did not work so i can’t be sure.

A noload would not answer this question either, has to be multiple dyno runs to get optimised phase angle for each rpm.

The low difference in inductance on rotation angle shows also that this motor will not have large benefit from the IPM design - it’s basicly comparable to a spm motor.
Ride on!

User avatar
Overclocker   10 kW

10 kW
Posts: 638
Joined: Mar 17 2014 5:48am
Location: Philippines

Re: QS 3000w midmotor

Post by Overclocker » Jan 18 2020 2:38am

qs1202kwb.jpg
qs1202kwb.jpg (28.7 KiB) Viewed 1858 times
i'm planning to get the 2000w version (qs 120). i have a Kelly KLS7230 lying around...

now i'm wondering if it would be better to get the Votol also

just wanna confirm if Kelly KLS is compatible w/ these motors. i'm not really interested in getting ultimate optimization. i just wanna know if KLS will do a decent enough job

larsb   1 MW

1 MW
Posts: 1617
Joined: Dec 10 2014 5:12am
Location: Gothenburg, Sweden

Re: QS 3000w midmotor

Post by larsb » Jan 18 2020 3:42am

Hope someone has tried it already and can answer. I don’t regard kellys as good in any regard so therefore the votol (to me) is the obvious choice.

But if you have one lying around then why not try it, should work well enough.
Ride on!

User avatar
PITMIX   100 W

100 W
Posts: 175
Joined: Nov 18 2018 3:25pm
Location: France

Re: QS 3000w midmotor

Post by PITMIX » Jan 18 2020 4:07am

Nucular 24F is the best
Aixam 550 Twin retrophite Qs 138 70H, Thermo-Nucular electronics 24F, LG Chem 20s 60Ah

Bartosh   100 W

100 W
Posts: 121
Joined: May 10 2017 3:32pm

Re: QS 3000w midmotor

Post by Bartosh » Jan 18 2020 4:50am

larsb wrote:
Jan 18 2020 2:01am
The low difference in inductance on rotation angle shows also that this motor will not have large benefit from the IPM design - it’s basicly comparable to a spm motor.
Can you shed some more light on that? Links to sources appreciated.

I've also been trying to make the fw work on the apt. It's the only controller I know where you can set the Id in a rpm-based manner. No success by now, but I will get back to it.

FYI phase to phase inductance of my 138 70h: 56.3, 55.7, 57.0 uH.

larsb   1 MW

1 MW
Posts: 1617
Joined: Dec 10 2014 5:12am
Location: Gothenburg, Sweden

Re: QS 3000w midmotor

Post by larsb » Jan 18 2020 6:46am

Try to find the largest difference max-min on the rotation of the rotor. It’s not much and can be related to the possible reluctance torque (if i remember it correctly.) A link to this is already in this thread i believe.
Ride on!

twikis   10 mW

10 mW
Posts: 26
Joined: Nov 24 2015 7:25pm

Re: QS 3000w midmotor

Post by twikis » Jan 18 2020 12:43pm

larsb wrote:
Jan 18 2020 2:01am
IPM optimized (e.g. Votol, ASI, maybe Nucular, haven't found any concrete info) controller,
None of those controllers are IPM capable or have MTPA algorithm for IPM motors, it requires dynamic phase angle adjustment to the specific motor. APT96600 might work but the field weakening setting in my controller did not work so i can’t be sure.

A noload would not answer this question either, has to be multiple dyno runs to get optimised phase angle for each rpm.

The low difference in inductance on rotation angle shows also that this motor will not have large benefit from the IPM design - it’s basicly comparable to a spm motor.
Oh wow okay, basically I just guessed those controllers would try to find out the optimal angles during initialization, since I thought there wouldn't be much to it (spin it up slowly, while it's spinning up play around with phase angle to find out the lowest consumption at e.g. 10 rpm-points, generate a curve a out of that, then later during normal operation do slight adjustments, easy)
Of course that's just my theory, most definitely likely underestimating it ha

User avatar
madin88   1 GW

1 GW
Posts: 3053
Joined: May 27 2013 2:02am
Location: Austria

Re: QS 3000w midmotor

Post by madin88 » Jan 19 2020 9:45am

twikis wrote:
Jan 17 2020 8:49pm
edit: The airgap seems to be quite narrow, that's nice
Yep thats good for torque, but not so good for iron losses.
larsb wrote:
Jan 16 2020 11:31am
It’s at 12.5A / 72V on my motor, don’t know exact voltage when i tried it.

kv60 without flux
Thats around 900W which i find is a bit on the high side for ~4000RPM on a motor like this.

larsb   1 MW

1 MW
Posts: 1617
Joined: Dec 10 2014 5:12am
Location: Gothenburg, Sweden

Re: QS 3000w midmotor

Post by larsb » Jan 19 2020 9:49am

Yes, i agree, think it might be with field weakening and 6000rpm. Can’t remember :wink:
Ride on!

twikis   10 mW

10 mW
Posts: 26
Joined: Nov 24 2015 7:25pm

Re: QS 3000w midmotor

Post by twikis » Jan 19 2020 10:22am

madin88 wrote:
Jan 19 2020 9:45am
twikis wrote:
Jan 17 2020 8:49pm
edit: The airgap seems to be quite narrow, that's nice
Yep thats good for torque, but not so good for iron losses.
larsb wrote:
Jan 16 2020 11:31am
It’s at 12.5A / 72V on my motor, don’t know exact voltage when i tried it.

kv60 without flux
Thats around 900W which i find is a bit on the high side for ~4000RPM on a motor like this.
If it's high or not, mostly depends on the internal resistance (which the important number I haven't found yet). Judging by the numbers I've seen, it could be as low as 5-7 milliohm, so not to bad actually

larsb   1 MW

1 MW
Posts: 1617
Joined: Dec 10 2014 5:12am
Location: Gothenburg, Sweden

Re: QS 3000w midmotor

Post by larsb » Jan 19 2020 10:45am

It’s 5mohm.
Ride on!

twikis   10 mW

10 mW
Posts: 26
Joined: Nov 24 2015 7:25pm

Re: QS 3000w midmotor

Post by twikis » Jan 19 2020 11:09am

Well when subtracting the 1 mOhm of the 1m 16mm² phase cable off of that, you get 4 mOhm internal resistance.

At e.g. 500A that's only 1000W copper loss, which is great!

Also means though, that this motor simply is wayy to big for my 50 km/h Cityel, and I'll get the 120 2kW for it (hoping that it's got an internal resistance of ~10mOhm)

Edit: forgot factoring in that it's a star connected motor lol.
So actually the copper losses are I² * phase-to-phase-resistance / 2 * root3, which for 500A would be:

500²*0,004/2*1.73=865W

even better

HCx   100 µW

100 µW
Posts: 9
Joined: Jan 24 2020 1:29pm

Re: QS 3000w midmotor

Post by HCx » Jan 24 2020 1:48pm

Looking to purchase one of these for converting my MotoPed.

Planning to pair this motor with a Nucular 24F and a 20S10P pack (limited to ~70A) with the expectation of building a 200A+ capable pack down the line.

Anyway, Where are people in the US finding the best pricing on this motor with a sprocket drive? I've seen it listed on Aliexpress, Alibaba and cnqsmotor and i'm not sure if there is some other trustworthy source I have yet to find.

Also. Are more detailed 2D drawings (or better yet CAD) available for this motor? I've seen some blurry 2D drawings but they're missing critical dimensions like where the bolt holes are located in this photo Image

Or where the sprocket sits in relation to the rest of the motor.

AviatorTrainman   10 mW

10 mW
Posts: 33
Joined: Feb 09 2018 10:21am
Location: Prescott, Arizona

Re: QS 3000w midmotor

Post by AviatorTrainman » Jan 25 2020 11:41am

I’ve made a rough cad model of the motor in fusion if you’d like me to export it, or I could post a screenshot of the dimensioned end sketch. I don’t remember where I found the image i worked from, but it had enough dimensions that I was able to calibrate and work from it. I was able to 3d print a prototype motor mount using what I have, so I know it’s at least accurate enough for that. As for the sprocket, the model doesn’t show the shaft accurately, but I’ll send pictures with dimensions in a bit.
With regards to where to buy it, I found aliexpress to have the same prices as the official QS website, and then you don’t have to go through the hassle of placing an order by email. If you put a comment on the order or message them with your order number right after placing it, they’ll send it with any size sprocket you request.

User avatar
macribs   1.21 GW

1.21 GW
Posts: 3612
Joined: Jul 22 2014 5:59pm

Re: QS 3000w midmotor

Post by macribs » Jan 25 2020 5:41pm

Stevebequik wrote:
Dec 02 2019 5:07pm
Finally got my build up and running.
Full size KTM 250 exc, QS3000, Nucular 24f. 18s 3kw Chevy volt.

Vas needs to build a 36f with 750 phase amps.
First I think that you got a sweet ride. Got a build thread planned?

Second, iirc vas already got plans for a 36f. But that nuc thread has grown big so you got to check it, or maybe just stop by his thread and ask when it will be ready or if he is still planning a 36f release. He is always working hard to improve and develop new cool stuff.
Last edited by macribs on Jan 26 2020 11:47am, edited 1 time in total.

HCx   100 µW

100 µW
Posts: 9
Joined: Jan 24 2020 1:29pm

Re: QS 3000w midmotor

Post by HCx » Jan 25 2020 8:15pm

AviatorTrainman wrote:
Jan 25 2020 11:41am
I’ve made a rough cad model of the motor in fusion if you’d like me to export it, or I could post a screenshot of the dimensioned end sketch. I don’t remember where I found the image i worked from, but it had enough dimensions that I was able to calibrate and work from it. I was able to 3d print a prototype motor mount using what I have, so I know it’s at least accurate enough for that. As for the sprocket, the model doesn’t show the shaft accurately, but I’ll send pictures with dimensions in a bit.
With regards to where to buy it, I found aliexpress to have the same prices as the official QS website, and then you don’t have to go through the hassle of placing an order by email. If you put a comment on the order or message them with your order number right after placing it, they’ll send it with any size sprocket you request.
Whichever is easier for you. I was planning on posting .stp files of the motor once I got mine in or found trustworthy drawings to work from.

thanks boss.

HCx   100 µW

100 µW
Posts: 9
Joined: Jan 24 2020 1:29pm

Re: QS 3000w midmotor

Post by HCx » Jan 25 2020 10:41pm

If anyone is able to provide me with accurate measurements for the sprocket offset from the mounting flange I can complete this model and release it for anyone else trying to figure out how to jam this thing into their project.

Image

AviatorTrainman   10 mW

10 mW
Posts: 33
Joined: Feb 09 2018 10:21am
Location: Prescott, Arizona

Re: QS 3000w midmotor

Post by AviatorTrainman » Jan 26 2020 11:29am

Your model looks significantly more accurate than mine, so I won’t bother to upload mine unless people want me to, but some (non-mounting) features are just flat out wrong on mine.

See the picture attached to see what I mean here: the center line of the 7mm thick sprocket sits 16.5 millimetres from the small circular area around the driveshaft. This small circular area is flush with the mounting area around the outside, where I have that 3d printed mock-up bolted on. The mounting holes are a metric 7 with a 1.0 thread, in case that’s not clear from the material you’re working from. It certainly wasn’t for me. The area where the lock ring/circlip goes is maybe a millimetre out from the sprocket outside face, so it might be possible to space out the sprocket just a little bit.
IMG_3959_resize.JPG
IMG_3959_resize.JPG (255.3 KiB) Viewed 1400 times

Haggan90   1 µW

1 µW
Posts: 3
Joined: Jan 27 2020 2:18am

Re: QS 3000w midmotor

Post by Haggan90 » Jan 27 2020 5:33am

Has anyone tried this motor with the Silixcon SL controller?
I wonder how much you could get out of it with this controller..
https://silixcon.com/product/esc3-sl-controller/

User avatar
VasiliSk   1 kW

1 kW
Posts: 312
Joined: Apr 10 2014 5:11am
Location: Russia
Contact:

Re: QS 3000w midmotor

Post by VasiliSk » Jan 31 2020 3:48am

For those who will try to remove pulley with minimum tools:
Image
Make two holes with thread M6 at some angle so axle will be not touched, tighten screws against axle using some plate, and heat up quickly belt side of pulley (spin it), it should pop out

HCx   100 µW

100 µW
Posts: 9
Joined: Jan 24 2020 1:29pm

Re: QS 3000w midmotor

Post by HCx » Jan 31 2020 1:44pm

A steering wheel puller from autozone or your local parts store will get the job done as well.

most auto parts places do tool rentals for free/cheap with a small part deposit.

Post Reply