First build was running well, then I let the smoke out!

[slofunk]

Hi All,

Long time lurker, first time poster. Was actually going to try to get some pics up and share the success of my somewhat low budget setup but alas it seems as though I may have been too generous with the voltage too soon!

Ok here is what I am running:
http://tncscooters.com/product.php?sku=101217
24 - 60 V Controller (Model YK43B)

http://tncscooters.com/product.php?sku=106130
450W GEARED Motor - 36 Volts (Style: MY1018Z)

Mounted RHD using this plate & and reversing the polarity of the motor
http://tncscooters.com/product.php?sku=106030

Driving a normal thread-type hub with dual/independent 16 tooth freewheels, as I have seen mentioned a few times on this forums and other places (using a LH BB bearing cup as a means to allow the mounting of them both on the same side of the freewheel.

Original Plan Was:
Get all of the above, and 6v12ah sla's from Gruber wired in series for a 48v setup. Since I had read a few forum postings about folks running the 24v my1018z at 36, I thought ok, then I can surely run the 36v my1018z at 48 right? Well, I never actually got that far, being my usual impatient self I ran to a local shop and got 4 12v10ah sla batteries and wired them in series to try out. As luck would have it, I ended up creating a nice shower of sparks and sounds when that battery pack slipped out of my hands and it ended up completely nuking one of the batteries (no continuity at all/zilch across the terms).

I ended up removing that cell and ran things as a 36volt setup. It seemed to me to function as expected just with some reduced range and such with I thought was most likely due to the accidental havoc the batteries had been exposed to.

What happened today:
So I got my batteries from Gruber in (qty 8, 12ah 6v) and got them all arranged and lashed together into a pack wired in series for 48v.
Got them on the bike, no sparks, arcs or accidents. Proceeded to get the best egrin yet, with much stop and go, pedaling when starting, but not as much as when I was running the 36v because it felt like the bike no longer needed it, but this was maybe was created the soon to be smelly situation.

After a few shakedown runs around the block, I came back to strap a fork to the rack to take to a local bike shop a few miles away. When I was about 2 blocks away I noticed the power had cut out and I thought the chain had jumped off (that was the annoying then I was routinely fighting at first when trying out different ways to mount the motor with the plate on the rhd side.

When I was safely off the road, I looked down at the chain/motor, observed it was on then started to get this sinking feeling in my gut as I noticed that there was some pretty warm looking/smoking insulation from the general vicinity of where the power wires enter the motor, and it almost looked like heat was literally boiling the grease out of the geared portion of the motor, as well was black crap coming out of the seals around the motor itself (melted enamel from windings??).

So frustrated and distraught I pedaled it back to the house (at which time my pos left pedal sheared off it's spindle) and park it for the day.

So, I am pretty sure the voltage was not the whole issue, but perhaps, my weight and using things in a bit more of a spirited fashion when the newfound zoom zoom that I was initially presented with after dropping the new 48volt pack into use. So when from a stop, or slow start, if much throttle was applied and it was more juice.. with no where to go, so to speak (akin to trying to start a manual trans car in a higher gear/etc) that was my main problem?

So now I am basically trying to decide which way to go here, do I cut my losses with this project (more to come later I hope) and replace the motor with the same one (as the integrated reduction and the way it mounts makes it very attractive) and run it on this same controller and just not go for 48v and drop it back to 36 and enjoy that as it is (and perhaps focus on getting a lighter battery pack next) or change something else to continue running at 48v, like using a lower gear/larger freewheel, but if the end result is a less strained motor running at 48v that is spinning faster so it doesnt bog, but it ends up being the same as the 36v on the 14 tooth which gave me no issues then perhaps it isnt worth it.

Feel free to flame away here, or point me to something I may be missing out here. I see that while my currently controller doesnt have a low voltage cutoff due to the range of voltage it tries to work with, it does seem to have some current limiting, however I am thinking that perhaps by the time this controller hits its current limiter that my1018z I have is already toast. Or its just plain and simple super heat and I could maybe get this to work next time around by utilizing some of the cooling techniques Ive seen on this forum that others have done on the 24v my1018z motors.

Thanks!

Jamie

 

[amberwolf]

I may have missed it, but I don't see what gearing ratio you are using from the motor to the wheel. What I suspect is that you are not letting the motor run fast enough, so it is overheating due to too much current flowing thru it, because it's back-EMF (BEMF) is not high enough to prevent that.

I doubt the controller has any current limiting, so it'll happily supply whatever amps the motor asks for until something burns up between batteries and wheel.

Even if we assume that motor can do 450W continuous (which I am hesitant about, at best), I doubt it is only taking that much power to do what you're doing with it.

Lets see: 450W / ~36V (really more like 39-40 at full charge plus some sag) would be about 12.5A max the motor could draw to reach 450W. I'd venture that if you don't have it geared right, you'll probably pull 15-30A instead, and much more during startup from stops and hard accelerations at WOT.

At the likely higher voltage of full charge, it's probably more like 11-11.5A max, depending on how much voltage sag there is during these moments, to reach 450W of power consumption in the motor.

At 48V (more like 50-53V fully charged, depending on sag), the allowable current to maintain 450W is much lower: less than 10A. (at the "real" voltage, it's as little as 8.5A!).

Many motors will do a lot more than their rating for a moment or two, but if you either keep it up longer than that, or keep hitting it with those bursts frequently, it'll overheat, sometimes with disastrous consequences.

Brushed motors have it harder than brushless, in that a brushed motor will get even hotter (from arcing at the brushes) at higher voltages than at lower ones. The harder arcing will even damage the commutator eventually, if done at high currents a lot it'll do it sooner.

I hope you can revive the project ok, but I think you should get a watt meter like the Watts Up, Turnigy Watt Meter, Cycle Analyst, or something, so you can monitor both current draw and voltage (and power usage in Watts) at the same time, while riding, and also see the peak currents, minimum voltage (sag), and peak watts you're using. These will help you figure out what's going on in your setup and hopefully help you tone down the throttle usage enough to save the motor, controller, and batteries from destruction.

Calculating out the gearing needed to let the motor stay near it's top RPM is a good idea, too, so you can make sure it isn't going to be run too far out of it's efficient zone, and waste less power as heat.



BTW, the reduced range vs expectations is probably due to the Peukert effect, which is really hard on lead-acid, as you're probably drawing a lot more current than you think you are. LA batteries probably don't deliver more than half their rated Ah in the uses we put them to, often not even that, when pushed heavily. And that is assuming you run them down to dead, rather than stopping at about 1/2 nominal capacity or so (which will extend their life a lot, as will immediate recharging after use).

 

[spinningmagnets]

Because the motor is brushed, limit yourself to 36V. It will work at 48V (obviously it has), but others who have tread this path have noted the brushes wear out very rapidly, as they were made for 24V.

You are using a 16T freewheel on the wheel, driven by a 9T sprocket on the motor shaft (1.7:1 ratio). That is the biggest problem as AW mentioned. You are bogging the motor down with too small of a reduction, and it is drawing massive amps. It would have drawn more than it got, but SLA batteries of that type sag when you try to draw too much.

Figure out how big of a chainring/sprocket you could adapt onto the 16T freewheel, or onto the wheel. The 9T will be noisy, 12T or more will be significantly quieter, but I wouldn't use anything smaller than the 9T on the motor as you will have a hard enough time improving the ratio as it is.

Best of luck, and you might learn a lot by reading McDesigns early posts. http://endless-sphere.com/forums/search.php?author_id=16686&sr=posts

 

[slofunk]

Thanks so much for the replies so far all! Very helpful information for me.

And to make sure I am sharing the correct info here ratio wise..

I have the stock 9T on the motor with a 16T on the driven wheel which is 26" with a cruiser type tire of 2.125".

I know there is 7.18:1 reduction that is built in to the motor correct? Based on this if its right..
http://tncscooters.com/product.php?sku=106130
9 tooth sprocket for 1/2'' x 1/8'' pitch chain.
12 gauge power leads
Gear Ratio: 79:11 (7.182)

So I guess that no load 3000rpm motor speed they mention would be reduced to 417 almost 418 rpm. (At 36 volts)
If I take that and my 9:16 ratio of the motor cog to the freewheel cog which is .5625:1 that means that if it could reach that speed or close to it ideally, if thats where it needs to run it would have my 26" wheel/tire spinning at 235.125 RPM correct give or take.
Pumping those numbers into this url http://www.machars.net/bikecalc.htm show 18.1 mph/29 kph.
From looking around here, it seems that the izip/ezips maybe had a 20T sprocket driven by these types of motors. Which gives me 14.5mph when I toss those numbers into the bikecalc link above which seems to go along with the stock 15mph speed mentioned by all of the various reading material for any of their conversion kits you can buy with this motor. If that is kind of the range I need to be looking at to not burn things up (especially with my weight, and if my current controller doesn't have a means to limit the duty cycle until the current draw is under X (unless I can build a circuit of my own to work along with my controller). I do have a 25amp DC circuit breaker, perhaps I can put that inline (or another suggested value) as something id want to really grab my attention and let me know if I am getting into the zone to do this little motor harm. It seems like its a decent little unit, I was just obviously asking too much of it! Without some fancier electronics at least to take care of it if I was going to bog it down. Maybe based on my weight, and understanding that with that, and this motors abilities I should look at going even to a 22T freewheel perhaps, to go for more torque than top end. I could put my 14T on my crank driven wheel and use my ole legs to stretch things on the flat/assisted and the 22T would give some pretty decent acceleration and better hillclimbing assist for what this motor may be capable of.

Thoughts there? Definitely sounds like I should be looking at gearing down and volting down, or looking for some type of fancier brushed controller with a tweakable current limit that I can set myself to allow me to protect the motor, which could allow me to reach some higher speeds on flats maybe but wouldn't do much at all for any hill assist as I would be taking the motor out of its sweet spot anyways and the controller would be cutting back.

So again.. to summarize, I should pretty much banish any kind of ideas of 20+ mph on this setup it sounds like and I should be looking at keeping the stock 36 volts on the 36volt motor if I am going to use this controller which doesn't limit the current (even if it claims to, or perhaps its limit is past the point of preventing any damage on this motor).

I think I will go ahead and order another motor and look for a larger toothcount freewheel to experiment with!

Thanks for the info guys. I really appreciate all of you. I continue to glean a great deal from all of those who participate here.

-Jamie







So all of these various Unite MY1018Z motors like we see on the tnc parts page. Is there anything different about them internally at all? Winding/brushes wise or are they more or less the same thing just being sold at different prices? If that is the case seems like Id just buy the cheapest one and run it at the 36 volts then.

 

[neptronix]

That looks like the motor that comes on eZip Trailz.
Let me tell you something about that motor... at 36v, it does have a limited lifespan.

Read up on 36v eZip mods; you will see many people that burned their motors up by increasing their voltage.

 

[amberwolf]

Dunno about differences in the motors; I've only ever used a 250W Unite scooter motor, and it didnt' survive the experiments, which kinda like M1 thru M4 were "not entirely successful". :lol:

Regarding the circuit breaker, 25A isn't going to help much to preserve the motor. It will probably not pop until more than 25A has been going thru it for some time, several seconds to over a minute, depending on how it's made. (depends on how much more than 25A)

@25A and 48V, you'd be putting almost 3x the rated power into the motor. It'll survive short bursts of that, infrequently. But do it very long and it'll get pretty hot pretty fast, and it may not be good for the gearbox either, if it's suddenly applied from a stop (I don't remmeber what that gearbox's power rating is, for the teeth and so on before they shear off under load).

For 48V, I recommend a 10A or at most 15A breaker. If you keep popping a 15A, you know you're pushing teh motor too hard. The 10A might pop just from a startup, if you're not pedalling.

There's nothing wrong with abusing the motor, as long as you appreciate that it will damage it, and doing too much damage will eventually break it. :lol: Lots of people here abuse their motors (and other components) with few or no issues, as long as they don't push them past their limits.

That's one reason I recommend the watt meter, as it will help you know when you're pushing too hard.

Spitting on the motor case can tell you too--when it sizzles, you better stop.


You could still do higher speeds and keep the gearing you have, but you're going to have to go way easy on the throttle at lower speeds, and help it with pedalling, until you have an automatic way of limiting current.

One issue with automatic current limits is that they don't give you the option of high power when you really want (or need) it, like some emergency situation that requires sudden high acceleration to get out of the way. Without a limit, then you could at least *try* to do this, even though it has a good chance of damaging the motor or other components in the process. Kinda like adrenaline--your body can tear itself apart trying to do something when you're in emergency-mode, like lifting a fallen bookshelf off somebody, only to find a minute later that you have torn muscles and ligaments and dislocated fingers. (or stopping a dog fight only to later notice all the dripping holes in yourself... )



Anyway, if you're getting more stuff for it, get a watt meter too.

 

[slofunk]

As far as an automatic way of limiting the current in proper way such as cutting back the duty cycle am I looking at a different controller or possibly modifying this one or some kind of hack with something like a CA where it would interface itself with the existing controller and it could affect throttle backoffs during too high off an amp load.

This makes me wonder how hard it would be to create a circuit that could just be inline with the throttle line and also inline via a shunt and upon x amps just open/close the throttle to simulate a complete back off of the throttle until the load was down. Could also integrate a temp sensor perhaps that could also open the throttle circuit when unsafe temps are hit or when they are maintained for x number of samples intervals indicating more than just a quick run that created high heat for a predetermined non fatal amount of time.

This actually has me interested in trying this out with one of my arduinos I have sitting around. I could definitely see using a relay that could hold closed when energized for the throttle to operate as normal and I could sample my other io pins for temp sensor and such. I am just not sure what type of device it is that let's me measure current with a shunt and could present that as may e an analog voltage that I could poll for to determine given amp. I shall be googling for what would allow that.

I guess one could really make their own controller with something like an arduino for brains if they had the proper other bits to handle the high amp/voltage parts that could be triggered by the micro controller.

Definitely has me thinking. I just have to find that balance of stuff I can beg borrow or steal versus what I can buy because of the ole wife calling me crazy factor for messing around with ebikes.

I sold my old BMW k75 before we got married and I quickly learned it will be a miracle if I will ever be able to replace it!
Of course with gas prices, oh I guess I can says that but with all of the stuff I end breaking and buying I could have bought enough gas to go a long long ways in my little Crx but that wouldn't be near as fun as this tinkering right!!??

 

[spinningmagnets]

create a circuit that could just be inline with the throttle line and also inline via a shunt and upon x amps just open/close the throttle to simulate a complete back off of the throttle until the load was down. Could also integrate a temp sensor perhaps that could also open the throttle circuit when unsafe temps are hit or when they are maintained for x number of samples intervals indicating more than just a quick run that created high heat for a predetermined non fatal amount of time.

If you make one of these and get it to work, a lot of builders here would copy this. Its a very good idea. It would also be nice if it had a setting where when its getting warm, it cuts the throttle signal in half instead of just dying suddenly when it reaches the hot set-point...a halfway point could be a warning to back off E-power (and perhaps pedal more) while still providing some assist.

 

[slofunk]

As you can see, looks like it got pretty toasty in there for sure!
That is the understatement of the day!

Thanks!
Jamie

 

[amberwolf]

Wow...that's way worse than the one I destroyed. :lol:

As far as an automatic way of limiting the current in proper way such as cutting back the duty cycle am I looking at a different controller or possibly modifying this one or some kind of hack with something like a CA where it would interface itself with the existing controller and it could affect throttle backoffs during too high off an amp load.

That's relatively easy even with just an analog circuit. If it already has a shunt you can use a tap off of that, or you can tap across the FETs' drain-source junction, to measure the voltage drop there, as a kind of shunt, and use an op-amp to amplify that voltage and another op-amp to alter the throttle output downward, based on the first op-amp's voltage. Just combine the real throttle and the second op-amp's outputs another op-amp, to allow normal operation if the op-amp doesn't detect a problem. I had in mind a way to do this so the detection circuit would always have priority, but I forgot what I was going to type.

This makes me wonder how hard it would be to create a circuit that could just be inline with the throttle line and also inline via a shunt and upon x amps just open/close the throttle to simulate a complete back off of the throttle until the load was down. Could also integrate a temp sensor perhaps that could also open the throttle circuit when unsafe temps are hit or when they are maintained for x number of samples intervals indicating more than just a quick run that created high heat for a predetermined non fatal amount of time.

Would be useful if you do make something like this. Temp sensor is difficult to "modulate" because different motors have different soak times before they reach thermal capacity. Where you put the sensor makes a whole lot of difference, too. If you put it right at the windings, it's probably best for the burst of heat measurement, becuase that's where much of the heat is. In a brushed motor that is hard to do, since they're spinning, so best you can do is mount it so it is very close to teh windings as they spin, and rigidly mounted so it cannot bounce into the windings during bumpy rides. Commutator and brushes would be good to monitor temp of too, but again it's difficult to directly connect to those.

Brushless motors don't spin their windings, so they're a lot easier to measure temp of windings directly, and for the most part that's the only part whose temperature is important. Magnets, too, but usually if you got hot enough to kill those, the windings are already fried.

It might take some doing to determine timing on how long you have after a certain temperature is reached before it's just too much, because if the motor is shedding heat well, like in a cold breeze or humid or rainy air, it might take a long time. But if it's hot and dry, and/or the motor doesn't get any airflow, or is simply a design that doesnt' move heat out of the motor very well, it could happen really fast, under the same power usage conditions as the other.

Also depends on the power usage at the time, since going uphill you'll heat up much faster than on the flats, usually.

The actual circuit to do what you want is pretty easy, even just using comparators and op-amps. Probably all of it already exists in schematics here on ES, for various throttle and power-control addons, as well as bits of reverse-engineered controllers, as well as custom-designed ones. It's just not all in one place, in one unit.

You need:
--an op-amp that isolates and measures the shunt voltage drop
--a comparator with adjustable trigger point that compares the actual drop to a preset limit
--an op-amp(s) that isolates and measures the thermal sensor(s)
--a comparator with adjustable trigger point that compares the actual sensor voltage to a preset limit
--a second comparator with adjustable trigger point that compares the actual sensor voltage to a higher preset limit
--logic (gates or diodes or transistors or whatever) to combine the comparator outputs so that throttle is cut down or cut off based on desired operation. Possibly adjustable.

And that's pretty much it; just figuring out component values and working out other "corrective" circuitry (to fix any issues that crop up, or allow adjustments, etc.) is "all" that is needed.

I am just not sure what type of device it is that let's me measure current with a shunt and could present that as may e an analog voltage that I could poll for to determine given amp. I shall be googling for what would allow that.

Basically you're measuring voltage drop across the shunt resistance. So, an op-amp to amplify that drop to something close to the full-scale input voltage of the MCU's ADC. Without the op-amp, the drop may be pretty small (200mV or less) and thus low-resolution in the ADC scale. Plus the op-amp helps isolate pack voltage from MCU; the op-amp might blow but hopefully not the MCU, in case of problems like hooking up battery backwards.

I guess one could really make their own controller with something like an arduino for brains if they had the proper other bits to handle the high amp/voltage parts that could be triggered by the micro controller.

[/quote]
THere are a few controller projects ongoing here on ES right now.

EDIT: fixed quotes so it isn't just one gigantic quote.

 

[slofunk]

Thanks for the good info there Amberwolf.

I was looking around again this morning and noticed this handy little item from sparkfun.

AttoPilot Voltage and Current Sense Breakout - 90A
http://www.sparkfun.com/products/9028
Features:

51.8V Max
89.4A Max
Very low zero current offset
Analog output scaled for 3.3V ADC
Self Powered

That wouldn't work for some of the super voltage setups but could be something to look at for 48 and under I guess (even though a hot off the charger 48volt pack would be above that voltage max depending.. hrmm.

Something like this, I could take the output into some analog pins on the arduino and measure the voltage to determine the amps and at least make some basic decisions like to cut the throttle or not I guess. Not fancy, but could be fun for tinkering with the concept.

 

[slofunk]

New motor should arrive today!
I hope to swiss cheese it a bit as I have seen on some of the threads here.

I also have a CA coming from the nice folks @ ebikes.ca so I hope to be able to use it to limit the current and also do some tinkering with some of the folks at the local hackerspace to try some non-CA ways to do some current/temp limiting with a microcontroller/arduino.

I see mouser has some dc fans rated at 48volts, if I were to rig up a fan/screen_filter to get some air into the newly swiss cheesed my1018z would it be better to run the fan just on a toggle switch so that it just stays on when I am riding (or maybe even parked during an errand) or just connected so that its speed is controller by the controller output so that it increases with the throttle?

Although, I think perhaps I just answered my own question, it would seem that the most "heat sink" or where it really would soak in would be at a stop.. or maybe after stopping after a long high load situation when the motor itself would be idle, this would make me think I'd want that fan on just all the time, or if I got fancy and had it rigged to a temp sensor to cycle.

Any thoughts there? And sorry if I am rambling on in the wrong section here! Let me know if this needs to be moved to another section.

-Jamie

 

[digitalmahem]

slofunk,
Is it possible to post a pic of the motor and bracket setup? I've been researching this for 2 months and would like to get the same motor and bracket but not sure as to the final drive.
Is the motor driving the rear sprocket or front chainring? I'm trying to find a way to modify the rear hub to run the motor without having the pedals move.

Thanks.

 

[slofunk]

slofunk,
Is it possible to post a pic of the motor and bracket setup? I've been researching this for 2 months and would like to get the same motor and bracket but not sure as to the final drive.
Is the motor driving the rear sprocket or front chainring? I'm trying to find a way to modify the rear hub to run the motor without having the pedals move.

Thanks.

You bet.
I will try to take a picture tomorrow when I get the new motor mounted up.

I have mine mounted on the right hand side for one, which is different than the usual currie setup.

I have a normal 10-speed-etc thread-on freewheel hub.. and i have two independent 16T bmw style freewheels on it.

To be able to mount both youll want to do something like mentioned here.
http://www.youtube.com/watch?v=dKT97jOceXE
Watch at about 50 second or so when he begins showing how to stack two freewheels..

Once you see this I think youll get it. This was the biggest thing I tried to figure out before making mine.. ps ignore the retro-direct chain wrapping stuff.. unless you want to do that.. but what i have is my crank spinning one freewheel.. and my motor attached to the other freewheel..

Perfect for single speed.. but nothing else I guess.. Depends on your needs!

Thx
Jamie

 

[digitalmahem]

Jamie,

THANK YOU!!!! Thats what I've been trying to find. Next week I'll try to get someone at work to thread some pipe for the extension.
I see freewheel sprockets are pretty inexpensive and I don't mind single speed.
Looks like I can still keep this conversion well under $300. Sweeeeeeet.

Dave

 

[amberwolf]

I see mouser has some dc fans rated at 48volts, if I were to rig up a fan/screen_filter to get some air into the newly swiss cheesed my1018z would it be better to run the fan just on a toggle switch so that it just stays on when I am riding (or maybe even parked during an errand) or just connected so that its speed is controller by the controller output so that it increases with the throttle?

I'd do it via temperature sensor inside the motor case, if you're going to do anything other than off/on.

But if you get a fan, get a squirrel-cage type. Regular old case-fan style fans don't move enough air to do much cooling, especially in situations where the airflow is restricted (like with filters and screens and such).

Also, set it up to suck the air thru the motor, with the fan as an exhaust, rather than intake, and put it at the non-commutator end. Then do the ventilation holes at the commutator end, so you are pulling fresh air onto what is probably the hottest part of the motor under load, and that air will then pull thru the case, across the spinning windings *and* the magnets, and help cool the whole motor.

You would probably have to run a duct from the motor's output hole to the fan, as it probably won't fit at that end against the casing, with your chain drive and mounting brackets there.

If space doesnt' allow either fan or duct in that spot, then set it up to blow into the commutator end, at least, and out the other end.

Either way, definitely use the squirrel-cage type fans to move more air.

 

[slofunk]

Finally got around to swiss-cheesing the motor and giving it a spin this past Saturday.

http://www.flickr.com/photos/lvl1/5798173356/in/pool-1298317@N20/

http://www.flickr.com/photos/lvl1/5798173718/in/pool-1298317@N20/

No forced air yet, and I was much more conscious of my throttle lever and didn't seem to have any issues.

Made sure to pedal up a bit before SLOWLY rolling on the throttle. (Still running at 48V).