Cromotor + MethTek Monster 24FET + 20S lipo build

[hjns]

I propose skipping the working part.

-methods

Well, I am almost embarrashed to say this, but I really like my job. Given a couple of millions to start living from the interest, I would probably still be doing what I am doing right now. It is just too much fun to work with nice people who know their stuff and try to create something together that each individual would never be able to do on his own.

So, I guess I am lucky. And thankful for that.

 

[hjns]

Hi all,

Just had some sweaty moments on my way to work. Some of the CA current sensor wires are not well connected. I suddenly experienced a very very sensitive throttle; in other words, when I opened the throttle, it would give me something like full WOT. Not nice. I cant ride with that.

I then looked at the CA settings, because I run a current throttle. Everything was still good, low ramp up numbers, current maxed at 65A, brought it down to 45A, throttle at 0.87V, minimum voltage out 1.20V, max voltage 3.60V, I zero amped. Nothing I did made any difference.

So then I looked again at the current numbers, and the CA was showing -141. Ampere. That did not sound right. I reset the CA, again -141.A. I re-evaluated the settings again, but no difference. I moved my CA wires around, and whoops, the currents change from -141 to +15 back to -141. I disconnected the CA, looked at the connector, but all metal lips are intact. Connect it back, and the CA shows +41A. While I am at full stop!

Ok, after a lot of fiddling with the wires where they enter the connector, I find a way of keeping the current constant. I zero amp the CA again, and voila, my throttle works again.

When (if) I get home, I will rewire all CA wires, remove all connectors, and have a direct connection all the way to the controller.

Lesson learned: connectors are a liability. Minimize where and when you can. With a current throttle, issues with the CA wiring can cause your bike to behave in unexpected ways.

 

[methods]

This has been gnawing at me... and not just in terms of having the CA drive the controller. Back when we were all running 40A and a 5305 it was dangerous enough. I had my share of run-away throttle situations (like when the throttle came off my trike and it took off WOT) but with those setups the mechanical brakes were strong enough to overcome the motor.

This CroMotor and a 150A controller... uhg.
If we loose control of this it could be easy to lose your life - or worse loose your skin and limbs.
I am starting to think that reliability may be a lot more important as we go up in power.

I wonder if there is a module I could assemble that could be installed in the controller to look for "insane" situations and cut power. It could be a latching system - such that it needs to be reset. Usually before a bike goes berzerk it gives you a few warnings (which we of course ignore) but maybe we would be more inclined to investigate if our protection circuit tripped.

I did notice some settings in the XIE Chang software that would cut throttle if the voltage went above what was expected. These things are good....

We also have an advantage in that the motor will not operate if the controller goes completely bonkers - unlike a series wound motor that can just blast off if the controller freaks out.

I am apprehensive about giving over control to the CA.... Justin is thinking about 36V 20A bikes while he develops - not 100V 150A light weight motorcycles. Just a blip could literally land you on your back with this much torque. We are going to need lots of test monkeys to shoot into space before I put too much of my skin on the line

-methods

 

[methods]

Are you ready to be our first Beta Tester???

Just yesterday Matthew sent me a few pictures to drool over. The final piece of the puzzle came in and we have units ready. Fully functional - we just need men who want to lose their skin

"Like a monkey, ready to be shot into space. Space monkey! Ready to sacrifice himself for the greater good."
-Tyler Durden



There is only one more set of tests I want to do before shipping out.
I am going to go into the software and set the over-current time to 5 seconds - then I am going to use a ratcheting strap to tie down my cromotor and I am just going to hammer on it - stalled - with the controller. 5 seconds 5 times in a row.

Then I am taking the bike over to a very steep hill we have here. I am going to crawl up it at 20% throttle until the thermal limiting kicks in.

If I survive these two tests then there will be a controller in the mail for you. I am pretty sure we will.... :wink:

-methods

 

[binlagin]

I have the Cromotor and the balls to test this controller, and I just happend to get paid today

LET ME KNOW!!

 

[gensem]

nice move binlangin... just be carefull doing over 100a main on launchs

 

[methods]

I have been doing 180A launches. Basically at any speed below 20mph it starts burning out. I have a trike tho.... A 2wheel bike would be another story. I am quite sure we are going to see some sore tail bones.

And more than a few blown CroMotors....

-methods

 

[hjns]

Are you ready to be our first Beta Tester???
-methods

Yes Yes !
I will be looking forwards to the beta controller !!

Note, you have some very serious competition here in Switzerland. Lebowsky has his controller ready for testing. Unfortunately, today it has been raining like crazy, so we decided to postpone the testing. Otherwise we would have been able to present some nice pictures.

All the more reason to continue testing. There are two main things I am interested in.

• motor + controller temp after driving uphill a 20% hill with 50kmh continuously
• time to accelerate from 20kmh to 50kmh on the steepest part of that same hill.

The justification for the above is to have some standardized real life tests available and at the same time continue living......
Like Methods said, these combos of high power controllers + cromotor can be very dangerous. If I start pulling wheelies, I will put my 9C 2810 in the front and let it chew that.

Other suggestions for testing are more than welcome.

 

[Alan B]

I am willing to be an early tester. I will test on a Borg frame with a Cromotor and 70V 20AH. I'll set it a bit more conservatively and drive it to work almost every day 26 miles round trip with 1800 feet of climbing up to about 12 percent, so it will be a different test than the daredevils do, but I've already had a 24 FET blow a couple of times in this testing, so it has some stress. I will ramp up the power from where I am running now though.

 

[methods]

I will start a thread soon.
There will be perks to being an early tester - as there is some risk involved

-methods

 

[Alan B]

I will start a thread soon.
There will be perks to being an early tester - as there is some risk involved

-methods

You mean I need to buy a backup motor. And maybe a full body suit.

 

[cal3thousand]

This has been gnawing at me... and not just in terms of having the CA drive the controller. Back when we were all running 40A and a 5305 it was dangerous enough. I had my share of run-away throttle situations (like when the throttle came off my trike and it took off WOT) but with those setups the mechanical brakes were strong enough to overcome the motor.

This CroMotor and a 150A controller... uhg.
If we loose control of this it could be easy to lose your life - or worse loose your skin and limbs.
I am starting to think that reliability may be a lot more important as we go up in power.

I wonder if there is a module I could assemble that could be installed in the controller to look for "insane" situations and cut power. It could be a latching system - such that it needs to be reset. Usually before a bike goes berzerk it gives you a few warnings (which we of course ignore) but maybe we would be more inclined to investigate if our protection circuit tripped.

I did notice some settings in the XIE Chang software that would cut throttle if the voltage went above what was expected. These things are good....

We also have an advantage in that the motor will not operate if the controller goes completely bonkers - unlike a series wound motor that can just blast off if the controller freaks out.

I am apprehensive about giving over control to the CA.... Justin is thinking about 36V 20A bikes while he develops - not 100V 150A light weight motorcycles. Just a blip could literally land you on your back with this much torque. We are going to need lots of test monkeys to shoot into space before I put too much of my skin on the line

-methods

Safety tether that cuts power to the controller wrapped to your wrist would be enough. If you get tossed, bike won't run away. If throttle goes rogue when you are still on it, a tug will cut it.

I think Jeremy has something similar but used as a key instead of a safety

 

[hjns]

Safety tether that cuts power to the controller wrapped to your wrist would be enough. If you get tossed, bike won't run away. If throttle goes rogue when you are still on it, a tug will cut it.
I think Jeremy has something similar but used as a key instead of a safety

Well, for me it was sufficient to activate the ebrakes. With the steering bar moving forwards, they were almost automatically activated by my hands trying to grip on to something in a panic. Also, I think the idea should be to have some safety circuits installed to PREVENT from falling off. If my bike travels for another mile or so AFTER I dropped off with a skull base fracture, I couldnt care less...

 

[hjns]

Oh nice. Another issue.

Came home on friday with my pack 60% discharged. Tried to balance charge it this morning. When I connected one of the balance leads to one of the packs, two balance wires started smoking....

I immediately disconnected it, almost burning my fingers. Then tried to measure voltage. This is 10S 4.5Ah Zippy Lipo from HK. As many know, the main leads are 10S, whereas the two sets of balance wires are 2x 5S. The first 5 cells were all at 3.68V. The next 5 cells were 4.35, 4.38, 0.00, 4.52, and 4.38V. And yes, the pack is bloated, so I have put it aside for surgery.

This is probably the moment where my charge / discharge / balance charge behavious comes into play. Usually, I charge twice a day, with a balance charge twice a week. Since a couple of months I only looked at pack voltage, not cell level. Balanced charging happened while having 6 packs of 10s paralleled together. Usually that takes two hours to complete. Obviously, sometime this week one cell died, leaving the others to keep the pack voltage high enough for me not to notice.

Lesson learned: frequently check cell voltage WITHOUT paralleling, or get Methods cell protection system.

 

[methods]

We put one of my LVC systems on Matthews bike and actually saved his ass on a 30 mile ride. His 18S 15Ah pack had only 7ah burned and we hit cell level LVC. Turns out there was a single weak cell - but since we caught it before taking it below 3V he was able to do surgery and save all the rest.

I swear I am going to put one of those awesome systems on my bike
Just like Justin... when I went to visit him his CA had a broken wire and his battery was dead... so he was just pedaling around like a beaten man.
Lol... I got into this because I wanted all this cool stuff - now I make it and I am just to tired and frustrated with the realities or running a business to even hook up the stuff for myself.

Speaking of which... we are going to very soon start carrying Buss-Bars for paralleling up Hobby City packs. Copper 10mm2 bars with bullets set up for 4P. Going to be one step closer to a perfect plug and play system.
Thinking of including Hyena's 8A balance charger - I think that just could make for the perfect starter system. Bulk charge at 4.1V * #cells and top off with the 8A balance charger once a week or something.

-methods

 

[hjns]

Update:

• Rewired all CAv2 wires to the controller.
• Connected a separate 200A shunt between battery and controller.
• Connected the new CAv3beta to the external shunt.
• Connected the motor thermistor to the CAv3beta. This thermistor lives on the windings.
• Removed the 2x40A car fuses from between battery and controller.

I still need to calibrate the CAv3beta properly. I need to find a solution where the CA is powered, and can actually see some continuous current. At the moment, I just try to have the output (aka reported Ah) of the CAv3 match the output of the CAv2. It looks like I need to set the rdshunt at around 0.390 mOhm. As it is not yet properly calibrated, I do not report any power measurements.

What I can show, is the relation between hill climbing, motor temperature and speed. Nothing new, just nice pics for the cromotor / 18FET / 20S lipo combination with one nice hill. The motor was already a bit warm (30oC) from driving towards the beginning of the hill, at which moment I turned on the CA analogger. In about 4 minutes, I reached the top of the hill with an average speed of 40km/h, and went off on the other side. After that, I continued onwards to my working place, which obviously is relatively flat, and I had to stop several times for traffic lights andsoforth.

CAv2 (well calibrated) reported Amax 79A, Amin 28A, Vmax 73km/h, Regen 11%, Distance 18km, and 22Wh/km.

Next steps will be to have the speed sensor in the CAv3 working, as well as having it calibrated well. I do not yet intend to switch the throttle from CAv2 to CAv3, because I do not want a glitch in the CAv3 beta to cause me problems at 50km/h.

 

[hjns]

Calibrated the CA. This is this morning's ride. Same as previous post, but starting one small hill earlier.

Vmax 68km/h, Amax 79A, 666Wh, 25Wh/km. I ended up with hitting LVC (73V aka 3.65V cell level) during the last km.

Max winding temperature is around 130oC. Looking intensely at the temperature graph, it looks like there is some sort of plateau around 100oC (from time 65801 to 70540), after which the motor continues to heat up to 120-130oC. I am obviously pushing it, and should start looking to cool the motor. Oil bath sounds like the next step. If I can delay the temperature going up to that plateau of 100oC it would be great.

You can also clearly see the effect of stopping. There are two moments I had to stop for a traffic light (aka speed = 0 and power = 0), and you can see the motor temp going down steadily in a linear fashion, with the same rate for the two stops.

Next time, I will continue monitoring temperature after I reached my destination, to see how long it takes to cool down from 100-120oC completely. I can do that when coming home as well, and put the bike on a stand to see if this rate really changes when running the motor at no load. For a closed hub it will probably not make any difference. For the oil bath, I dunno, we'll see. For the vented cooling systems it should make a big difference due to the circulation of fresh air.

 

[Alan B]

Great data. How does the hub feel when you test it with fingers?

 

[hjns]

Great data. How does the hub feel when you test it with fingers?

Are you kidding? If my temp sensor says 125oC, I am NOT putting my fingers on the hub....

 

[Alan B]

A quick tap with a wetted finger should not be a problem. There is so much thermal resistance between the wires and the outside of the hub it may not be all that hot.

One time I got my 9C pretty hot. It was hot for hours after riding up the hill in Marin, but the outside was never hot enough to burn fingers. It took a long time for the heat to percolate out, so there was a lot of heat inside and probably close to the damage level.

Adding oil changes the game and the outside will run much hotter. Which really helps control the temperature inside.

When folks say they stop for a few minutes for the motor to cool they're kidding themselves. The amount of cooling in a few minutes is quite small.

 

[hjns]

I enlarged the resolution of my data.

As you can see, stopping for approximately 1 minute will allow a decrease in temperature of 7-10oC from ~125oC down to ~117oC. Assuming linear extrapolation of data, the motor should cool down to ambient 25oC in 15 minutes.

If the definition of "a few minutes" is 5 minutes, then that would mean a decrease of 35oC, which I think is considerable. So much for theory, I will take some longer readings once I get home tonight to verify.

edit: for clarification, the horizontal axis is actual recorded GPS time, directly taken from the NMEA output. So, "71353" actually means 07 hour 13 minutes 53 seconds in the morning.

 

[Alan B]

Somehow I don't think that rate of cooling will continue. That's the heat propagating from the windings into the laminations, etc, and it would be interesting to watch a full cool down cycle to see how it goes. Judging by the observation of how long heat propagates out of a hot hubmotor.

 

[hjns]

Somehow I don't think that rate of cooling will continue. That's the heat propagating from the windings into the laminations, etc, and it would be interesting to watch a full cool down cycle to see how it goes. Judging by the observation of how long heat propagates out of a hot hubmotor.

You may be right. I don't know yet. I wanted to observe the full cool down cycle coming home, but it was raining very hard, and as the CA analogger is not watertight, I decided to postpone the test. Maybe tomorrow.

 

[hjns]

We are both a little bit right.

 

[Lebowski]

I keep being amazed at the speeds For me 35 kmh would be enough but then again I'm a scary cat ... I think I just crashed on my motorbikes too many times :?