APL's V4 Cruiser Build.

[thundercamel]

Found the image I was looking for. If the cells were bad, then their voltage would sag further the more amps that you try to draw from them. If they aren't bad, and their voltage is staying where you expected, then the system as a whole is either being limited:

At low speeds by the phase amp limit (set inside the controller)
At mid speed by the battery amp limit (set inside the controller)
At high speed by the voltage or back electro-magnetic-field (BEMF, which is generated by the motor spinning) equaling the voltage being sent to the motor. The voltage limited portion is affected by the Kv or winding speed of the motor, and by the voltage of your battery. In your case, you could just change your gearing if this is what's limiting you.

We're confident in your 1000w reading, correct?

 

[APL]

Thanks thundercamel, nice graph and performance/troubleshooting info. 1000w more or less, I don't have the CA shunt dialed in yet, but with the various launches, it's within 100w or so. Voltage sag seems somewhat normal compared to last year.

I just got back from a new test ride with a fresh charge and now she's getting 1200w and 30+mph. Takes 4 seconds to get to 28mph, (counting "one thousand"). Full throttle at take off gets me the high watt reading and that's a pretty solid figure.

So, doing better, and the batteries are still under 60 degrees, so it should improve slightly over time, but not much. I think I just made a bad choice in my selection from the battery pile, and should have tested them before I tore the packs apart.
Could be wrong though.

I can test a few cells, since I have extras from the batch. Not sure how to do that yet or if I have the equipment, but I'll look into it. I have no idea of the history of these cells, so they could easily be just used up.

I don't know what the controller is set at, it's been generic programed for general use, although there is a program port,.. I would need to solder in connector a plug. Rather than do that I had planed to just get the Votol EM 100 and have all the programing and info I need for it here on the ES. Have to send off shore for that though, which I should probably do.

Was hoping to slip by all that for now, it's rated as an 80 amp/3Kw controller, so I assume it's set up to do better than 30A/1500w.

fechter wrote

If your controller could do field weakening, you could get more at higher speeds.

Controller has a three speed plug, and I 'think' it's default to mid., so I wonder if it does have field weakening. (same as EM-100)
Also, "soft start",.. another mystery to me, but I might need that when I get some bigger amps flowing.


Example photo, taken form an ES thread,

 

[thundercamel]

The field weakening should only apply to the voltage limited phase of that graph above. It draws more power to raise your top speed, should you so desire.

 

[APL]

Sounds like it should be a momentary push button on the high setting, for more top speed rpm when wanted, but maybe not something to leave on. Wondering what the low field weakening setting is good for.

 

[APL]

After a couple of charge cycles and a weekend of long rides I realize that the bike is working better than it was last year, and doing
it with less watts. More efficient motor and better gearing no doubt.

Definitely not easy to drain the battery and takes a couple of days, so the cell capacity seems to be good, I'll check the miles on this charge. (Forgot to write down the start miles before.)

The sensitive throttle is both annoying and exciting at the same time, it's nice to feel the power and response when I pull it back, but trying to keep a steady hand all the time isn't. Guess I'll hook up the cruise control to mitigate that a bit.

Pack output is up to 1300w, and the fact that it's climbing tells me that there isn't any hard current restriction from the controller or
system components. However, it occurred to me that the 18" 8awg battery wires might be sucking up some watts. (36" round trip.)
Putting the BMS in the back looks cool, but is a bad idea,.. it should be in front to keep the wires short.

Tried to convert 3' of 8awg & 60v voltage drop into watts but didn't have any luck.

Anyway, this project has moved beyond the ancient Dynabat cells that started it all, and it's time to shelve these things for lesser use, like my electric lawnmower that needs new SLA batteries, or maybe a flashlight or something, but they've served their purpose to get this far.

I just spent a load on the upgrades for this summer, so it looks like it's time to drag out some Craigslist fodder to counterbalance
the new battery cell cost, which looks like just short of $1K. (need 176 cells, so 200 count)

Looking around on the net I'm seeing some of the the best quality deals that are actually available and 'in stock' are:

Samsung 25R max 25A C - 35A pulse, 300 charge cycles at $3.85 per 100 (Battery Store) or...
Molicel P26A max 25A C - 35A pulse, 500 charge cycles at $3.59 per 100 (Battery Store)

Samsung 30Q is a solid $5.00 battery and shows (so far ) 250 charge cycles.

30Q is a sure performer but $$ and less cycles. (?)
25R is found in almost every drill pack and a good bet..
Molicel looks to be the best overall deal,.. but is it?

 

[fechter]

Pack output is up to 1300w, and the fact that it's climbing tells me that there isn't any hard current restriction from the controller or
system components. However, it occurred to me that the 18" 8awg battery wires might be sucking up some watts. (36" round trip.)
Putting the BMS in the back looks cool, but is a bad idea,.. it should be in front to keep the wires short.

Tried to convert 3' of 8awg & 60v voltage drop into watts but didn't have any luck.

3 feet of 8 awg wire will be about 1.9 milliohms. Power loss is I^2*R, so at 80 Amps, would be around 12w. At 40A, loss would be closer to 3w.

I don't think voltage drop in the wire is a significant issue.

 

[nicobie]

Unless it's getting warm, even 8 gauge probably is overkill for your use.

I've always used the 10 G multi strand slinky stuff from hobby king with good results.

 

[thundercamel]

Are you restricted to 18650 cells? Because of the existing battery enclosure, right?

EDIT: I used this site to get 300 panasonic 21700 cells which I'm currently building a pack from. These Panasonic NCR1865J2 3400mAh are unwrapped 180 count for $430.

 

[APL]

I've been eyeballing those Panasonics for a while. :wink: 180 count is as perfect as the price is. I need unwrapped so that's good too.
They say the cells will do 17 amps max. I found another site with more data, that states 2C = 6.8A max constant.
https://flightcomp.com/products/panasonic-ncr-1865j2-3400mah-17a-2s-li-ion-tx-pack

They would work fine with the 11P pack, that would be 75 amps C, and 187 max amps. Fits the system well , and the 3400mah is a nice upgrade too. (Tempting to hit the 'buy it now' button.)
Wasn't sure about the rep of the company or the cells though, so I've been a little leery, and wanted to see what new cells would be.

For a little less than twice as much $$, I could have brand new boutique cells, so I guess that's the dilemma I'm wrestling with.

Really wish I could use 21700 cells, but yea,16S was too tall for the battery box with the system I have, I'm sure there's a way to get them in there though, and I'll be leaning that way in the future for sure. I'm on a mission to test this DIY 18650 dump-box.

 

[APL]

Thanks for figuring the watt draw for 8awg fechter, that eliminates that idea, and puts my mind to rest.

That leaves the cheap $20. 100A shunt, the battery box, and the cells as suspects. Shunt seems in line with past performance watt
readings, but needs to be dialed in a little. Battery box I'm not willing to throw under the buss just yet, it has the watt hours, and surly it's capable of more than 30 A.

Cells were stored at high volt for a few years, which they say is bad, but never say what "bad" means exactly. Dynacrap cells were
rated 2400mah 3C new, or 7.2A,.. and now I'm getting like 3A out of them.

 

[thundercamel]

(Tempting to hit the 'buy it now' button.)

The seller was at least friendly in my communications with them.

Do we know what settings are in the motor controller (phase amp limit and battery amp limit), and what the pack voltage sags to under load?

 

[APL]

So tempting! :lol:

Controller settings are unknown other than it's rated at 80A bat cont., so 50A more than the 30A I'm using, I'm thinking that it's not electronically limited because the watts have climbed a bit. Not suspecting the controller too much yet since it's pretty HD.
But it is a cheapie at $90. shipped, so can't rule it out either.

I did a few voltage sag runs,.. bike has about 15 miles on it at 62v-rest. When I slam the throttle at 1200w I see 56v on the CA pretty consistent.

Battery Clearing House - Panasonic deal is tempting but I'd be starting out 'used' yet again, and don't know how many charge cycles they've been through, plus the power output is a not the greatest. Would get me on the road quick and do the do cheap though.

On the other hand 180 Samsung 25R's are $240. more than the Panasonic deal, and the Molicel 26P's are only $195. more. So that's not really a deal braker, and I'd have brand new. Don't know what the two companies sock you for in shipping though.
Do we have extra charges with lithium?

 

[thundercamel]

The shipping for my order of 300 21700 cells was $35.61. While I'm not sure what makes the cells "like new" instead of new, mine appeared to have never been used. Maybe they sat on a shelf for a while? Also odd that the wrappers have no text printed on them. Wish I had the pack and battery box finished already so I could provide more experience.

Regarding the voltage droop. 1200 watts divided by 56 volts equals only 21.4 amps that the controller is pulling. Either that's what the controller is currently programmed to, or it has a high Low Voltage Cutoff that is temporarily limiting current.

 

[APL]

Thanks for the shipping info thundercamel, definitely cheap enough.

Sorry for the late response,.. I've been chewing the fat over the situation and thinking about what you said. Looks like I still need to do some detective work.

Regarding the voltage droop. 1200 watts divided by 56 volts equals only 21.4 amps that the controller is pulling. Either that's what the controller is currently programmed to, or it has a high Low Voltage Cutoff that is temporarily limiting current.

If the old 80A Yantan controller is going to scam me for 50 amps then it's going into the Yantan trash can!

21.4 amps is pathetic, but even so, it's enough to make this bike fly. 1500w would be perfect, and 2000w would be even more perfect. After that the front wheel is just going to come up and I can find cheaper ways to just brake things. A 50 amp battery draw on a mid drive is probably more than a person needs and while it would make a good Dragbike, that's not really what 'I'm' after.

A high 'low voltage cutoff" makes me wonder if the 60v - 72v option is set wrong, and maybe it's on 72v. That's easy to check.
Then there's the three speed options plug,.. maybe I need to choose one, I thought it was default to Mid. but might need to be grounded to set it. I plan to put a momentary button on the high setting for more top speed, so I'll check that all out.
Waiting on shipping for the handlebar button.

They never get the printed info on these things right, and many times the plugs and connectors are reversed from what they say,
or not even close.

Otherwise, the other limiting factors are the individual cell output, the battery box, and the BMS. The 100 amp BMS is probably OK,
and the battery box I can check by using the welded packs I have and adding three cells for a 60v pack. (A little work, but I can do.)

So, as far as the individual cell output. I've been reading lots of sophisticated drain processes, but all I want to know is "will they give me a high momentary amp draw", or not, so I can eliminate the cell as a suspect and move on to the other things.
(or maybe I shouldn't even bother suspecting them, as most any 18650 should be able to do 5 -10A each)

Can I do it the redneck way and just use a 10 or 20A fuse, or an old 30A analog ammeter and a starter switch on a single cell for a one second test? (Or is that too crude )

 

[amberwolf]

Then there's the three speed options plug,.. maybe I need to choose one, I thought it was default to Mid. but might need to be grounded to set it. I plan to put a momentary button on the high setting for more top speed, so I'll check that all out.

All controllers with a 3-speed switch I've ever worked with so far default ot the "middle" setting with no connection. Grounding one of the two switch wires (to the third, ground, wire in the 3speed connector) changes to either high or low speed, whichever that wire's for.

What each "speed" setting represents varies by specific controller. (some of the ones I've had are the same response under all conditions for all three switch configurations).

They never get the printed info on these things right, and many times the plugs and connectors are reversed from what they say,
or not even close.

Unfortunately...yes. The only consistent thing about them is their inconsistency.


Easiest way to test for high current capability is use a wattmeter on the system as a whole and ride it up a hill, if you have one. If not, try towing something heavy enough you know the bike can barely do it. If voltage sags during the load, the amount of sag vs the battery load current system current will tell you if the battery is the cause of the problem. (knowing what cells you have you can look up the manufacturer (or other testing site) discharge curve for the load you're putting on each cell, and see what voltage sag *should* be; if yours exceeds that it could be a cell or battery interconnect issue).

If you don't have a way to test like this, you could use electric heaters, toasters, toaster ovens, old stove burner elements, etc., to create enough of a static load on the battery, measured via a wattmeter, to emulate the load you would expect during riding, then do the same check(s) as above.

 

[APL]

Thanks for the info on the 3 speed plug amberwolf, I didn't know the mid wire was a ground. I'll check that out, but don't have much hope for any real change. 60v -72v plug is an easy check too. (Still not sure what 'softstart' does.)



Other than that I'm running out of options for waking up this controller, other than going into the brain programming, but probably need software for that and it's not my forte.

I have another even cheaper controller that I could waste more time hooking up,.. but if that one gave me more power then at least I would know that It's a straight up controller issue, and time to buy something decent. Trying to eliminate the simpler things first.
Lots of stuff to try over the next few days.

Most I've ever seen on the watt meter was 1300w but I'll see if I can stress it for more,.. seems to be limiting at that amount, 20 -25ish bat-amps. I hear what everybody has been saying about only using what it needs, and motor winds, but I should have been able to see more than a 1.3kw draw by now, with all the hills and launches I've done.

 

[amberwolf]

(Still not sure what 'softstart' does.)

Softstart usually means throttle response is delayed (ramped) inside the controller. No matter how quickly you push the throttle up, it will always take time (sometimes seconds, for some controllers!) to react.

The effect of this is to prevent hard startups; it limits initial torque and acceleration; prevents high surges of current.

Controllers that don't do this are often called "instant start"...but most controllers don't say one way or the other which they are.

Most I've ever seen on the watt meter was 1300w but I'll see if I can stress it for more,.. seems to be limiting at that amount, 20 -25ish bat-amps. I hear what everybody has been saying about only using what it needs, and motor winds, but I should have been able to see more than a 1.3kw draw by now, with all the hills and launches I've done.

If the load placed on it is high enough, yes. If not, then no, you won't see more power usage.

 

[APL]

Well I gave it a test by applying the brakes and loading it down to a point close to stopping the motor at full throttle, and it's the same 1200 -1300w max watts,.. no matter how much I stress it and load the motor down, it's never shows more than that.

I tried the low and high settings on the speed plug, and the high setting works really nice on this one. Whips it up from 29 to 35 mph! (The low setting was useless.)

Tried the 60v / 72v swap, and that was a fail too.

I tried the old school 'Dial 911' cell tester out and it pegged out past 30A and needs a bigger scale.  Nothing burned up on a one-second test, but don't try this at home folks, you need a "professional" idiot, like me. :wink:



So unless I'm missing something these cells should have enough power, at least momentarily, at 11P and 330A, and the controller is looking much more suspicious.

Well, since I might not need the new batteries right away, the Votol EM-100  is looking more affordable and necessary it seems.
So I guess I'll start working on that angle.

I have another even cheaper 42A current limited controller that I want to jury rig as a finial test. If it will pull the 42 amps, then it will show that the Yantan is limited for sure, and it will be the final death null.

 

[fechter]

Either the controller isn't allowing more current, or the motor winding resistance is doing it. At near zero RPM, the motor resistance will be very low, so shouldn't be the limiting factor, but hard to say. If you are at the limit over a fairly wide RPM range, it would indicate the controller. Trying a different controller would be about the easiest test.

 

[thundercamel]

I tried the old school 'Dial 911' cell tester out and it pegged out past 30A and needs a bigger scale.  Nothing burned up on a one-second test, but don't try this at home folks, you need a "professional" idiot, like me. :wink:

911 Tester.jpg

So unless I'm missing something these cells should have enough power, at least momentarily, at 11P and 330A, and the controller is looking much more suspicious.

:lol: This made my day, and it's only 9am here!

 

[APL]

Well my precious battery box is still a white elephant in the room, but I refuse to believe that it won't give me more than 30A.
Only 1P is needed to do that. Plus, I road over 50 miles on the last charge without a single pedal stroke and I didn't baby the throttle, so if a large number of cells were not making contact then the watt hours wouldn't be there.

I have a test for the battery pack too if it comes down to it, but the controller swap is probably the easiest, so I'll try that next.
Another crap controller, so it may be more quick-sand,.. but we shall see.

 

[spinningmagnets]

Many builders bypass the BMS on the output. Do so at your own risk, but...it does work.

As long as the controller/motor doesn't draw more amps than the battery can provide, the battery shouldn't get hot.

If the battery can provide more amps than needed, then the next major issue is the LVC. Low Voltage Cutoff. Bypassing the BMS means you can drain the battery to below 3.0V per cell.

 

[thundercamel]

I was going to say this earlier, but I thought that your amp meter test made it clear. The batteries don't really have a limit on how many amps they can provide. The voltage will sag down more with the more current (amps) you're trying to draw. The batteries will heat up, or something like the current pathway to the positive terminal or a built in safety will melt eventually, at a very high amount of amperage.

Testing the other controller is the right path

 

[mxlemming]

100% this is a controller problem. 12s11p of cells in pretty much any condition is going to give you more than 1.2kw. If it was BMS, it would completely cut out not just be a bit shit and if it was wiring causing the limit you'd have a fire on your hands.

I would get a proper controller. There's a chasm of difference between good and bad.

Unfortunately you might be about to learn that controllers are the bain of ebikes existence. You've got a choice of locked down ASI, 20 types of Chinese garbage, 18 month lead time on Russian nucular ones, exploding VESCs (very good once you've tamed them but that can take your life expectancy down from frustration...) Ones for rc planes that are 5x over rated and explode... It's not a happy place in controller land.

 

[APL]

I was really losing faith in the cells, but now it's been restored a little, (for now), and need to look elsewhere first.

Yes I've been learning about the controller bain, like a rocky road, rough and harsh. I was hoping that this controller might save me from the 'search', but it appears to be just an illusion.

Controller specs, price, country, delivery time, payment type, availability, trustworthiness, etc. Not as easy to get as I thought either, so now I'm in the soup looking though controller world again. Life is rough when your bike is down!

If things get desperate, I might even go so low as to try hacking the Yantan with a USB port. But the jury isn't in yet, so still some detective work to do.

Bypassing the BMS is something I should do, since it's in the loop, it should be eliminated. I thought it unlikely that a 100 amp monster BMS could falt to 30A, but I shouldn't,.. one of these components is messed up.

Cant be my fault, I know that!