MAC motor + Trek 4300 build

[neptronix]

Here ya go.





Also pending the black paint treatment as it's a bit ugly now.
And maybe some lightening as it's 1.5lb. Probably overkill!

[Racer_X]

Thanks for the pics, looks pretty sturdy to me. Is it going to be easy to remove the box(charging purposes)? Is it going to be easy to secure the box (theft proof)? The reason i ask is that my solution may be ugly but has worked really good to secure and remove and that is why i have not changed it's design since it was created November 2009.

[neptronix]

Thanks for the pics, looks pretty sturdy to me. Is it going to be easy to remove the box(charging purposes)? Is it going to be easy to secure the box (theft proof)? The reason i ask is that my solution may be ugly but has worked really good to secure and remove and that is why i have not changed it's design since it was created November 2009.

Yes and yes
Take the two screws off the top and pull to the side..

As for charging, i will have a plug coming out the side that will just need a disconnect/reconnect. No need to remove the box.

I needed a top-opening lid because the lipo needs to be checked from time to time. So that made things more complicated. If i did not have lipo, i would just hang the case off the top tube.

But i will never leave my bike unattended in a place i do not trust. No way! People steal bikes here in Portland all the time. Anything that is worth a dollar will get stolen.

I spent 9 months trying to figure out how to mount this battery. So glad to have a solution now.

Looks like you have a platform also?

[Racer_X]

Looks like you have a platform also?

Nah, just hangs off the top tube double wrapped with one of these luggage belts.

The bottom is being supported by a small screw that is supposed to fasten the water bottle holder.
Believe it or not, I have laid the bike down a couple of times and it has not damaged the battery, the belt managed to hold the battery secure. This project was a complete experiment and it has exceeded all my expectations. I hope that in my next build I can create a really nice and stealthy ride that hauls ass yet hard to tell that it is electric. Yours it looking that way, once you paint the ammo box it will be a little harder to tell what it is.

[Alan B]

Containing the batteries with a strong sealed box is dangerous. If they ever "vent" a strong box will convert the event to an explosion. You cannot contain the venting, and the box will fail most likely with the lid flying upward with a lot of force. An explosion can be prevented by having a large vent for the gasses to come out. A small vent may not be adequate. The vent may be sealed with tape or something that will easily open under pressure.

Be safe!

[neptronix]

Good suggestion about the vent sealed with tape. I will either remove the waterproof seal entirely, or follow your suggestion to a T. thanks Alan.

[neptronix]

Another ride today to test the range, this time on 10S, and on a milder hill:



Top speed: 26mph ( almost none of this terrain was flat )
Average speed: 22mph
Lowest speed: 19mph

Pedaling the entire time of course.

Max amps: 38amps

AH drained: 4.75AH ( according to the iCharger )

The batteries were down to 3.25v per cell average and i was smacking up against the LVC, so i had to turn back
After pedaling home, the batteries rose to 3.39v per cell average and were disbalanced by at most, 0.2v.
Stuck em on the charger w/o a balance charge and they went back into balance

The Infineon 9FET did not get hot this time. Just warm.

It is amazing what abuse these batteries will put up with. I'm going on 26 cycles with no balance charges needed.

Also, i charge to 4.15v per cell, and i only lose ~250mAh by doing so. That's a 5% loss to provide a safe buffer at the top of the charge. Not bad at all!!! that result is better than expected.

I will have to take this route on 15S to see what the efficiency difference is like.

[neptronix]

OK. Same hill test on 15S lipo.

Max speed: 36mph
Average speed: 24mph
Lowest sustained speed ( while hill climbing ): 31mph

Amp hours used: 3.9ah.

This motor, voltage, and controller combination eats hill for breakfast and asks for seconds.

Average speed was not much higher because i had to keep slowing down to go around turns, slowing down on sidewalks, and dodging road hazards. And the stoplights on the hill contributed to a smaller gain in average speed here.

There were a few problems:

+ I was having problems with stalling from a dead stop; I'd hit the throttle and it would literally not do anything. Perhaps the infineon has some kind of internal stall protection. This happened only on inclines. I know my LVC was not kicking in..... very strange. Only happened at higher inclines. Doesn't happen on 10S.

+ The controller was quite hot. The batteries were warm-hot. The 14ga battery wires were even hotter. My bad on that...

+ The motor generates a kind of grinding resonance / ringing type sound at particular speeds at this voltage. Must be the sound of the stock nylon gears crying for mercy. How terrible of me to ignore their screams This doesn't occur on 10S though.

+ When i hit the throttle, i can definitely hear the keyway clanking inside the motor. I can feel the jolt of torque that evens out. This is very pronounced at 15s, and not very noticeable at 10S.

+ The bike is very difficult to maneuver around obstacles above 25mph. Dangerous, is the word i would use to describe hitting a dip at 32mph and bouncing off your seat as a result. I believe i need a dual suspension bike with speeds like this. I think i will stick to 10 or 12S and call it a day.

I am now realizing that the 9FET is not meant for a sustained 2300-2500w peak.

One other note: at this power level, it is pretty much futile to pedal. I noticed that the difference in speed between me pedaling uphill and not is around 1mph.

at lower speeds, pedaling helps a little, at higher speeds, pedal gear ratios and the fact that the motor has 10 times the power you do really work against you.

But dear lord it is stinking fast.

Back to 10S-12S i go. Less abuse on the entire drivetrain, less problems, safer, etc etc.

[neptronix]

OK, hooked up 10AH today, 2x2 5s batteries.

Took the same hill route.

The batteries are thus discharged at 4C each and do not get warm.
Voltage drop is about 1.0- 1.25v when i hit the throttle instead of ~3v.

Actually made the hill run only using 4.8ah as reported by the turnigy watt meter. During the same time period, the turnigy watt mater reported 5.2ah of usage ! It is innacurate a little, but.. that's an 8% difference!

Well, i got to Portland and back with battery to spare. So i met my goal with this build... booyah

[toft]

Hy, one more question: is the (screw) thread for rar freewheel set in this motor is alu? Or steel?
regards, Toft.

[neptronix]

On the one i got... Aluminum

The cover plate is replaceable though, no need to take the spokes off or anything stupid.
Newer ones have steel plates apparently.

I won't be standing on the pedals anytime soon.

[neptronix]

Speaking of which, i went to install the rear disc today and stripped a disc thread hole.
The disc side is also aluminum, and apparently not very strong aluminum at that.

These are the rotor spacers i used:

http://www.chainreactioncycles.com/Models.aspx?ModelID=3288

3mm offset is what i needed, exactly.
Getting the disc brake on these motors is a real tight fit !!

[neptronix]

OK, i got the infineon programming cable from lyen. Works fine with the cell_man controller.
The COM port dongle is a bit weird, likes to be on port 6 only. You have to type the COM port into the programming software, you can't select it from the drop down menu. Also, you have to install the various files that go with the programming software.. on windows 7 64 bit, you have to run the installer as an admin by right clicking on it.

I see where this trips up a lot of people. Lyen's documentation covers all the issues tho.

Anyhow, i got the peak amps down !! yeah!!!



This produces a peak amp value of 34.5 amps on hills. Cruising amps are still in the 15-25 range as before.
The torque kick i've been complaining about is very much reduced!
Hill climbing power is reduced, but.. my pedaling makes more of a difference. I do like that.

No shuddering, no problems! I'm using a phase multiplier value of x2.6 rather than cell_man's x2.6666...

I will try 2.5x later on and see what happens.
I think it is safe to run higher voltage again, though

[gensem]

Neptronix!!!

Please, please, please test the block time set to 0 to see if the max amps respect the rated current.

Thanks

[neptronix]

Gensem, i will do that next.

It is going to draw a ton of current coming from a start anyway though. I measured the amps peaking at a constant 32A-34.5A during a steep 1 mile hill climb. That's the same amount of amps it tries to dump from a dead stop.

So the high amp draw has nothing to do with block time. Hopefully turning off block time will allow smoother starts though. I have noticed that if i can get to 3-5mph by pedaling first, the start is much smoother. But it is hard to pedal from a stoplight and not nail that throttle after the first half revolution of the pedals..

[neptronix]

Gensem.
I tried your version of the keywin software.
ASV file is saving 0.0 to block time, instead of 1.
Transmit worked, no error.

If there is any change, i'm having a difficult time telling.
Seems way eager to eat more amps, but i haven't done a scientific test at all, it's 2am
Engagement at lower throttle levels seems just a bit smoother.

By the way.... at lower amp levels, mid-throttle is less of a problem. Yay! I can go higher voltage and install a speed switch now.

Motor still clunks when engaging the throttle, even lightly.. i think that is a fact of life for this motor; that's just the mechanical clutch / keyway doing it's thing. Thin aluminum covers don't mask the sound.

I will try 15S and report back later; 15S is where i had the problem with a huge jolt of power.
I will also try a 5+ second block time and see if the peak amps go any higher.

-David

[wineboyrider]

I have been using my CA to dial down the amps at higher voltages and it works much better. My motor controller combination get's weird when you go over 1500 peak wattts. Nice posts Neptronix I want to play with the programming eventually and will keep an eye out for what you get. I have the same problems at 15s lipo, but 48v lifepo4 runs great so does 12s lipo. At 12s lipo it dumps 1200 watts continuous at full speed.

[gensem]

Gensem.
I tried your version of the keywin software.
ASV file is saving 0.0 to block time, instead of 1.
Transmit worked, no error.

If there is any change, i'm having a difficult time telling.
Seems way eager to eat more amps, but i haven't done a scientific test at all, it's 2am
Engagement at lower throttle levels seems just a bit smoother.

By the way.... at lower amp levels, mid-throttle is less of a problem. Yay! I can go higher voltage and install a speed switch now.

Motor still clunks when engaging the throttle, even lightly.. i think that is a fact of life for this motor; that's just the mechanical clutch / keyway doing it's thing. Thin aluminum covers don't mask the sound.

I will try 15S and report back later; 15S is where i had the problem with a huge jolt of power.
I will also try a 5+ second block time and see if the peak amps go any higher.

-David

If im not mistaken you got a watt meter, no change on the reading with BT set to 0?
Man I was wanting so much to have absolute control over the motor amps.
Im not so worried with the composite gears they should be able to handle it. But what about the clutch at higher voltage. And heat might be an issue too.
Btw is your phases getting hot?

[gensem]

I have been using my CA to dial down the amps at higher voltages and it works much better. My motor controller combination get's weird when you go over 1500 peak wattts. Nice posts Neptronix I want to play with the programming eventually and will keep an eye out for what you get. I have the same problems at 15s lipo, but 48v lifepo4 runs great so does 12s lipo. At 12s lipo it dumps 1200 watts continuous at full speed.

I have a CA coming my way too. Will have to test whats is better overall BT+ CA, etc.
Im also interested in the Time adjustment tool from Burtie, but no one that has it posted any opinions yet.

[wineboyrider]

I have been using my CA to dial down the amps at higher voltages and it works much better. My motor controller combination get's weird when you go over 1500 peak wattts. Nice posts Neptronix I want to play with the programming eventually and will keep an eye out for what you get. I have the same problems at 15s lipo, but 48v lifepo4 runs great so does 12s lipo. At 12s lipo it dumps 1200 watts continuous at full speed.

I have a CA coming my way too. Will have to test whats is better overall BT+ CA, etc.
Im also interested in the Time adjustment tool from Burtie, but no one that has it posted any optinions yet.

I have the plug in CA, which I usually use on my 9c setup hooked up to my BMC. The speedometer doesn't work right, but I can volt it up. But, no matter what I do with it over 1500 watts it gets really weird so I am sure programming it right would help better.
Can you make the direct plug in Cycle Analyst work with the BMC/Mac motors????????

[gensem]

I have been using my CA to dial down the amps at higher voltages and it works much better. My motor controller combination get's weird when you go over 1500 peak wattts. Nice posts Neptronix I want to play with the programming eventually and will keep an eye out for what you get. I have the same problems at 15s lipo, but 48v lifepo4 runs great so does 12s lipo. At 12s lipo it dumps 1200 watts continuous at full speed.

I have a CA coming my way too. Will have to test whats is better overall BT+ CA, etc.
Im also interested in the Time adjustment tool from Burtie, but no one that has it posted any optinions yet.

I have the plug in CA, which I usually use on my 9c setup hooked up to my BMC. The speedometer doesn't work right, but I can volt it up. But, no matter what I do with it over 1500 watts it gets really weird so I am sure programming it right would help better.
Can you make the direct plug in Cycle Analyst work with the BMC/Mac motors????????

I dont think you can get real speed reading from a geared hub because of its freewheel capabilities; I ordered the CA-DP and it should work with any motor, but the speedmeter has a regular sensor to be tied to the wheel to get actual speed reading. ebikes.ca has all very well explained or their site.
I dont have anything working atm, but I ll be doing a greyborg frame built with either MAC/BMS or Xlyte HT3525 in the coming months.
Cell_man has MACs running over 2,5kw peaks, so im gessing you might have timing problems. What controller what you using?

[neptronix]

Gensem, i have a turnigy watt meter and i watched it like a hawk.

My peak amps from a dead stop are the same as my steep hill climbing amps, whether my block time is set to 0 or 1.
Thus, i don't see how a cycle analyst to limit the amps would help you much. It's a hackish way to do what you should be doing with the controller.

At the mellower 1300 watt peak i am running, the motor does not get over the outside temp, nor do the phase wires.
At 2500 watts constant for 10 minutes, it was getting quite warm on the outside, which means it was much hotter on the outside.

The problem with getting an RPM reading from the motor is that it runs 5 times faster than a DD. I just use a bike speedometer.

So, my solution for the jerky start is to pedal a bit first. I think you should be doing that first anyways, as efficiency at lower RPMs is horrific

Check the ebike.ca simulator graph for the MAC_Shanghai motor here ( 36v ezee limn, 35a controller, 26" wheel ):



If you pedal to 5mph, you double your starting efficiency. Seems like this can help range some in stop and go situations anyway.

[gensem]

There is so much things the controller should do, and there is so much settings that are not available.

I was sure the start amps would be equal to the climbing amps. but was still hoping to kill BT for good.
Im gessing I ll be forced to get premium lifepo4 batteries and better BMS in my new project. And im almost jumping to the HT3525 in the new bike until the jumbo mac (2kw rating) cell_man comented about shows up. And a mild MAC 500w 30a 48v on the old bike.


BTW in your simulation torque is somehow low. Have a look at the BMC torque simulation with the same parameters.
Thats probably why I ate gears so fast on my MAC torque.

[neptronix]

Well, the BMC "torque" and "speed" winds are quite different in how much peak torque they produce.. also, their max speed is quite different too.

[neptronix]

Okay, just did two tests with block time set to 5 seconds.

Test 1: 27 amps batt, 67.5 amps phase ( 2.5x )

Peak amps were 34.5 amps, both climbing hills, and from a dead stall.
Block time had no effect! After 5 seconds, i was still drawing the same peak 34.5 amps as soon as i hit that 5% grade.

Stall test at top of the hill: avoided a stall. peak amp was 34.59.

Test 2: 27 amps batt, 88 amps phase ( 3.26x )
Peak amps were 40.25 amps, both climbing hills and from a dead stall.
Again, no effect from block time.

Stall test at top of the hill failed. Stalled multiple times, had to get the bike up to 1mph.
Motor case was a little warm after the ~1 mile hill climb.
Controller was pretty hot.

So there you go. Block time has no effect, at least on this cell_man controller.

Sorry gensem. Our theory did not pan out, at least in my test.
However, i did rotate the the thumb throttle clockwise a bit. Now i can modulate starts quite a bit easier now.

It's interesting how, if you give this motor a bunch of phase amps, it will happily take them.

I am going to set my amps back to 27a batt, 70a phase for a ~35amp peak. That seems to be where the 9FET's realistic limits are. At this amp limit, the 9FET should be safe during my multi-mile hill climbs.