Cargobike project. Very long john from chromo tubing

Tomorrow is paintday for the new hubmotor swingarm, the hatch on the cargobox and the eaton chargers "sheetmetal box" ive made.

I also have dialed in the best settings in the adaptto for the revolt 120pro. Will make a separate thread about this aswell incase someone searches.

Basically I have chosen a route on dead calm nights and done it several times with changed settings.
First I did the angle correction.

First session
Shorter trips of close to 2,5km to rough dial it in, speedlimit set and full throttle all the way:

-27,4deg 99Wh
-26,0deg 93Wh
-24,6deg 89Wh <- Winner, but this session I used every odd step.
-23,2deg 93Wh

Nest session I tried the values surrounding -24,6deg that I did not test the previous time.
Also went a longer route, about 5,1km.

-23,9deg 146Wh
-24,6deg 142Wh
-25,3deg 148Wh

At this point I was pretty sure -24,6 was the best angle correction setting for my setup.

The third session I tested ind timing. Same 5,1km trip but colder temperature and higher airpressure.
334us 160Wh
398us 156Wh
462us 151Wh
526us 149Wh
590us 149Wh
654us 151Wh

I think the 526 setting is the best. (Actually had it at 398 for the past two weeks)

I dont know if a higher powerlimit or speedlimit would affect the relationships between the results.
Typical loads in my system are 800w-2000w during these tests. Currentlimit of 30A (very rarely reached due to speedlimit)

From what I have understood, pwr timing affects acceleration but has minor impact on the efficiency.
I have been running 0,17 all along.

Before switching it out, remember it's a cargo bike, so do some runs with cargo aboard. Similar consumption rates with the minimum load could easily lead you to a wrong conclusion.

John in CR said:

Before switching it out, remember it's a cargo bike, so do some runs with cargo aboard. Similar consumption rates with the minimum load could easily lead you to a wrong conclusion.

Click to expand...

These tests doesnt change very much with cargo onboard unless I take a more demanding testroute, too few hills to really make a difference.
Most energy consumed is shuffling air. From the normal 23km commute (also pretty flat) I can see that the consumption only increases a few wh/km at most with cargo.
(Dont think I transported more than 30-40kg those 23km)

The fact that the total weight on the cargoarea seldom is high makes me think the tests are sufficient. Most of the time its a matter of transporting stuff that is bulky. With over 1300km of riding the bike, I think it will be reasonably easy to draw conclusions for the "normal" type of service the bike sees.
I think that the revolt is a better motor if I would make it work more, and I really like the combo of the revolt+adaptto. However, I applied the red paint over the epoxiprimer yesterday and all should be set for a swap later this week. Gotta get the leaftests done before the winter is here.

The hubmotor is mounted, and everything set. Did a short testrun just to see that it worked.
I like that it is silent. Cant hear the motor. The speedo is completely off so cant draw any conclusions other than that it gives less torque for a given current than the revolt (no surprise). Dont even know what topspeed the motor gives yet.

The tempsensor is working. Will post some pictures of the new swing and motor tomorrow night.
If the result is overall lower consumption and the motor handles cargo well enough I will keep it on.
If it ends up being too weak I will switch back to the revolt120.

Wheazel said:

...The fact that the total weight on the cargoarea seldom is high makes me think the tests are sufficient. Most of the time its a matter of transporting stuff that is bulky. With over 1300km of riding the bike, I think it will be reasonably easy to draw conclusions for the "normal" type of service the bike sees.
I think that the revolt is a better motor if I would make it work more, and I really like the combo of the revolt+adaptto...

Click to expand...

The light load comparison will be interesting. I'm not sure anyone has done any thorough same bike comparisons of a hubbie vs mid-drive on the same bike, so sharing your efforts is greatly appreciated.

I try to set up for worst case use, so I never have to worry about heat problems and in regular use the system sees a light load for great efficiency. That's where you are with the Revolt, and it will be interesting to see if the load is so light that the motor is running beyond the point of peak efficiency at cruise and with the added losses from the chain, that the hubbie comes out ahead.

By "load" I don't mean just weight, which as you pointed out really only matters for hills, but wind resistance too, especially if you frequently carry bulky stuff.

If you're getting much chain noise, I've cannibalized a derailleur for tension using a skateboard wheel with a channel cut in the center to make it a chain guide wheel, and rigged it to push down on the bottom (slack side) of the chain, and it made the drive nearly silent with the tire noise drowning out the chain noise. My hubbies have me spoiled with their near silence, but that mid-drive was darn close. I just need silent tires. 8)

Ok so I have started tuning the adaptto for the leafmotor. The top noload speed is pretty much dead on 1km/h per volt. Tested with a drained chargestate to 65km/h at roughly 65V. Went through an autodetect and used those settings.

Did feel that the motor wasn't doing its job properly so I started to do testruns with different settings. Turned out the settings were quite off.
They are better now, but not dialed in yet. Only changed angle correction so far. The conditions were not great for testruns as it was windy and after about the first 10runs the wind started to increase, making it hard to draw conclusions. I did however improve the angle correction setting enough so I was comfortable riding the bike the 23km commute.

I will post all the data when I have dialed in the angle correction and the ind timing. All the testing is done by going with a 48km/h limit (means it will reach 44-45 on a straight, dunno why the adaptto works like this but its apperently normal) and no pedaling, static bodyposition. Full throttle and landmarks where to release throttle, break and turn. Have done about fifteen 2,5km round so far. This is the same procedure I used with the revolt.

The 23km commute today was in pretty windy conditions (strong sidewind and alittle headwind). Average speed in the 37-38km/h range and just under 22wh/km consumed. This already indicates a big gain in range, will see what further tests yield, and how the motor handles some cargo and hills.

And here are the pics of the progress.


















The cargobox is a really nice addition for transporting groceries and stuff that don't want to stay put.

Not done with tuning yet but getting closer. Really like this motor. The topspeed on ~flat road is about 55-60km/h (only tested, very rarely use it).

The last 23km commute only consumed 18,3wh/km with an average speed of 40,2km/h. I did think alittle on the consumption and didnt use max throttle on the steeper hills. The trottle is set to torquemode so I guess partial throttle pretty much means another layer of current limiting? The conditions were also good, calm and nice weather.

This would not be remotely possible with the revolt setup tho. There is a major difference when it comes to light load. This is where the leafmotor shines.
I have gotten the motor to about 65C with the current settings. During the first test after autosetup, the motor reached 78C from just driving a short trip. Wonder why the adaptto autosetup works so bad...

For flat land 43-45km/h riding the motor stabilizes in the high 30Cs. Some mixed terran will make the motor stay in the 40Cs. Hills will warm it up quite quickly if I put boostmode on and pull 50+ battamps. Gotta test some hills with cargo aswell, but I am pretty sure I can live with the lower torque capability when gaining this much in overall efficiency. None of my usual hills pose any problems to climb so far. Might get rougher with heavier cargo.

The bike would probably have a 400km+ range if going around 30km/h and pedaling along.
The 43-45km/h cruisespeed can be sustained for approx 200km if pedaling along.

Have done around 150km with the leafmotor to date. Mostly testing but some commuting aswell.

So to sum up the pros and cons so far:

Pros

Silent
Efficient
Smooth

Cons

Not as much "grunt" as the revolt
More unsprung weight, can be felt on rough gravelroads
Have to bring a clunky 22mm wrench in the cargobox along with the emergency tools incase of a flat on the rear tyre

Waiting for dead calm weather to do the final tuning with the leafmotor, have been windy even at evening/night the past few days.
Did some groceryshopping back and forth trip 25km mostly countryside road with a consumption of 19,7wh/km. Average speed 40,1km/h.
It is harder to see results from changes made to the parameters for the leaf compared to the revolt.
Thats why I really need the dead calm weather and longer trips to find the best settings.

So I think I have come to a conclusion with tuning the adaptto for the leafmotor.
Overall the settings have smaller impact on this combo than what I saw with the adaptto+revolt120pro.

Have used 2 testroutes, the same ones as my previous tests.

First shorter runs to roughly find the right angle correction. ~2,5km roundtrip, speedlimit 45km/h (set to 48 in the adaptto, tops 45 in reality)

Average speed, Watt hours, Angle correction
37,6 86 -4,38
38,2 85 -3,65
38,3 85 -2,95
38,4 83 -2,25
38,5 82 -1,54
38,6 80 -0,84
38,6 81 -0,14
39,0 79 +0,56
39,1 77 +1,26
39,0 78 +1,96
38,9 78 +2,66
39,2 77 +3,37
39,1 78 +4,07
39,2 78 +4,77

Session 2, another day and continuation to see if I could find where it gets worse again.
38,8 75 +3,37
39,0 75 +4,07
38,4 75 +4,77
38,6 78 +5,48
38,5 76 +6,18
38,5 77 +6,88

Hard to draw clear conclusions, but it seems the best range is in the +1-4degrees angle correction.
To see if I could determine the best value in the range, I went the longer 5km route in dead calm weather. Same settings.

Average speed, Watt hours, Angle correction
40,4 135 +1,26
40,1 133 +1,96
40,2 133 +2,66
40,2 135 +3,37
40,3 135 +4,07

Then I tried two settings of the ind timing with the +2,66 angle correction
336us took 136Wh, 494us 133Wh, 622us 133Wh

At this point I decided to scrap further tuning as it doesnt seem to give big enough results to be able to determine THE right setting with the methods I use.
I will run with the +2,66 angle correction and I put the ind timing in the low 500s.

Its very clear the leafmotor saves alot of energy compared to the revolt for this application.
Have done some hills with cargo aswell, (still pretty light cargo at 15-20kg) and the leafmotor handles it well enough.
It will have harmonic vibrations if the speed is ~25km/h or lower and alot of throttle is applied with high resistance from a hill.
The highest temp I have seen is around 70Celcius, the case is barely warm with this reading.

I think the motor is running in the best efficiency range when doing the usual commute.
The tempsensor tends to read 0C until it is around 30C. It can pass several minutes of riding before I get a tempreading, sometimes 5-10km.

The regen braking is also working now after a reedswitch was installed on the rear breakhandle. Very nice feature.

The bike has more than 1600km clocked now. Such a practical and comfortable ride!

Yesterday I discovered something unexpected. I had around 180kms on the tripgauge and it was getting time for a charge.
I wanted to burn some more energy to see how much capacity I had in reality, so I put boostmode on and started to drive around and wasting energy.
Pulling 3-4kw repeatedly.
Motor warmed up from the abuse and reached around 100C at the peak. I was unsure about how high I could warm it up safely so I stoped there.
When I got back the battery was close to drained (had noticed a faster drop in voltage indicating the discharge graph-slope).
Wanted to check noload speed at this point, and was shocked to see ~100km/h noload. I immidiately thought I had overheated the motor and lost some magnet strength.

But then I started thinking and came to the conclusion that my previous noload test was done straight after adaptto autotune.
So I did one of those again, and it ended up on pretty much the same values as the first autotune. Noload was back down to around 75km/h.

So now I am confused about what winding I actually got. The settings I use are obviously the most efficient ones for the type of use the motor sees, but whats up with the 100km/h noload?
Worth mentioning is that noload current is high at 7-8A with the settings I use.
I have the OVS timing set to zero.

Leafbike told me that the motor could handle up to 200C in an email today, so I did not harm the motor unless the tempsensor readings are x2 off. Nothing indicates this to be the case.

So I have been using the bike on and off the past weeks and it works very well. Up to ~2200km now.
Did have a mishap last week when I hit a pothole (maybe 30cm wide) and the hit was hard enough to cause a rear flat.
Tyre pressure was probably 3,5bar at the time.
The pothole was in the middle of a bikepath I seldom use, a hole in the concrete, so the edges were "sharp".
Dont know how it got there and I was defenitely not expecting it. It was dark and I was going maybe 40km/h when hitting it.

After this incident I have a repeating squeeking noise tied to wheel rpm. Also abit inconsistent, because sometimes I can not hear it, and sometimes it can be heard up to 25-30km/h.
Havent been able to pinpoint the cause but I suspect the axle might have bent slightly and bearing misaligments in the motor causes the sound.
Just a hypothesis tho.

Another concern is the approaching winter. Temps are going under freezing now and I have disabled the regen in the adaptto.
To avoid charging the cells at too cold temps.

I'm thinking its probably time to park the bike for the winter soon, unsure of the safety on black ice etc, even with good spiketyres.
Time to switch to the upright trike again, but I would much rather keep riding the much more practical cargobike.

Oh and I almost forgot! I did shoot some video of the bike the other day, will post that when its done.

Something like this might work for your cargobike?
We been riding mx bikes at frozen lakes with studded tires like that. No grip issues. If there is no ice and clear pavement of course you will fill the rough ride a bit - but if most of your winter riding is done on ice and snow I am sure you will be happy with studded front and rear.

That 100kph no-load must be some kind of advance timing thing done temporarily by the Adappto, and nothing to worry about unless you see consumption or temps jump.

Overall it's a great testimonial for DD hubbies in non-stressed use. I love hubbies and the results compared to the Revolt surprised me, though I still have to believe there's an issue with the Revolt's timing since getting quite warm makes no sense to me at that low power. The winding resistance is so much lower that even with the 0.5mm lams and relatively high voltage something is amiss.

If I identify a solution while adding halls to my v1.0 Revolt120, I'll be sure to share. The first thing I'll do is check the no-load current with a sensorless controller, which should give a "true" no-load current with proper timing.

Electric cargo bikes are awesome vehicles every household should have, and yours turned out great. Thanks for sharing so much detail.

macribs said:

Something like this might work for your cargobike?
We been riding mx bikes at frozen lakes with studded tires like that. No grip issues. If there is no ice and clear pavement of course you will fill the rough ride a bit - but if most of your winter riding is done on ice and snow I am sure you will be happy with studded front and rear.

Click to expand...

Those are some nasty studded tyres
Its not very convenient to choose the "best" option for the winter conditions here in Sundsvall Sweden. Costal climate means periods of cold and stable winter weather with a good chunk of snow/ice, but just as much unstable weather with both snow and the occational rain and above 0C. Sometimes the entire winter is low on snow, and sometimes but rarely we get alot. So all in all there are probably just as many days with an ice crust on the tarmac as days with a good layer of snow/ice. I have a pair of suomityres hakka 300 that I will put on the bike if I dont put it away for the winter this weekend. Those are tyres that feels safe on all surfaces on a nonelectric mtb.

Also debating to get a pair of the schwalbe ice-spiker pro tyres with 361 studs and a wintercompound rubber. My hakka 300's have seen 2-3 years of use but are still decently sharp.

Wheazel said:

Also debating to get a pair of the schwalbe ice-spiker pro tyres with 361 studs and a wintercompound rubber. My hakka 300's have seen 2-3 years of use but are still decently sharp.

Click to expand...

Winter tires are more prone to become dull over time. As time goes the rubber compound turns from soft and grippy into harder and less grippy. Even with studs my choice would be to change those 2-3 years old winter tires for some new ones to assure maximum grip when needed.

You are right, safe tyres comes ontop of the list of things "unwise to sneek out on".
Ive had my fair share of happenings over the years to cement this.

All good things comes to an end, at least for this year.
I decided to not ride the bike during winter and put it away today after several weeks of very mild autumn weather. The adaptto said roughly 2820km travelled total since the bike was built.
Cant complain about much. Reliable and comfotable.

I am still not done with the video I took. Am a total noob in that department, but will get it up sooner or later.
The trike feels very odd after all this time on the cargobike, but I figure it is the best choice as it will be much safer when the black ice arrives.

Wow Wheazel, Nice build !!!

Good idea for black ice conditions. How about all the details and pics of your trike. The tiny picture in your avatar is just enough of a teaser for us to know it's interesting.

Stay safe on the ice in the meantime.

John

John in CR said:

How about all the details and pics of your trike.

Click to expand...

teeny bit here
https://endless-sphere.com/forums/viewtopic.php?f=3&t=71540&hilit=+trike
https://endless-sphere.com/forums/viewtopic.php?f=14&t=72939&p=1102075&hilit=+trike#p1102075

but that's all I know of...seems like such a tease. :/

John in CR said:

Good idea for black ice conditions. How about all the details and pics of your trike. The tiny picture in your avatar is just enough of a teaser for us to know it's interesting.

Stay safe on the ice in the meantime.

John

Click to expand...

Problem is I dont have any good documentation of the trike, apart from crude rhinoceros line-drawings to figure out the geometry of the front mechanism.
It was built merely to test some ideas I had. The main purpose was to get a fully suspended bike with added safety on icy surfaces.
I didnt really expect it to be in service still!
The front end has been rebuilt since the pictures that are shown here, but thats about it.

Still it feels like a very rough build compared to the cargobike now that I have that perspective of riding.
I do however have thoughts to refine the trikeconcept and build a new one some day.
Preferably with some added cargocapacity. Not as much as the cargobike, but enough to make it alittle more practical.
That would certainly get a buildthread.

Wheazel said:

It was built merely to test some ideas I had. The main purpose was to get a fully suspended bike with added safety on icy surfaces.
I didnt really expect it to be in service still!

Click to expand...

CrazyBike2 is similar--was supposed to be a testbed to determine what was needed to build what i REALLY wanted, (and to find out what that was), but it's still running 7-8 years later, albeit significantly different from what I started with.

amberwolf said:

Wheazel said:

It was built merely to test some ideas I had. The main purpose was to get a fully suspended bike with added safety on icy surfaces.
I didnt really expect it to be in service still!

Click to expand...

CrazyBike2 is similar--was supposed to be a testbed to determine what was needed to build what i REALLY wanted, (and to find out what that was), but it's still running 7-8 years later, albeit significantly different from what I started with.

Click to expand...

Thats fantastic, such a nice feeling using transportation that you created yourself for your own needs.