BMC V3 -- Infineon Build

It lives! I got my new controller hooked up tonight: Crystalyte 48v48amp 18fet from Maxwell. He still has more of those for sale too. The wiring that ended up working:

Phase wires
Motor- Controller-
yellow--> blue
blue----->green
green--->yellow

I used a premade adapter from ElectroRide for the hall wires, so I'm not sure what the wiring combinations are for those.

This controller was way smoother and more powerful than the Ecrazyman/Infineon one that burned. Now I have a fuse in-line, and the biggest fuses I have at the moment are 40amp. I blew one, but being easy on the throttle, the second one made it most of the way home--at least until the battery was drained. Our one remaining problem is that the phase wires got very hot very quickly. After only 2 minutes of some uphill and flats, the wires were very hot. The controller was warm, and the motor was cool. I think the wires are getting so hot because of being too small gauge for that much power: they are 16 gauge I think. They tell me that it is because that is all they could get out through the axel. I wonder if there is any way to get bigger wires in there. As I have it routed now, I could feel the hot wires on my pants--not a good thing. Tomorrow I'll be having more fun, looking for some 50amp fuses, and seeing about measuring the speed. It was fast with this new controller though, I was cruising right along with minimal throttle. :D
 
Good stuff Muad'dib. Glad to see you finally got it up and running. Yeah these Crystalyte controllers are really nice. I just got back from riding my ebike with the new controller installed and it's a lot more accelleration and uphill torque than I'm used to (but I'm sure I'll adjust quickly) :mrgreen: . Do you have a Cycle Analyst hooked up to your rig? It is still pulling a lot of amps when it's sitting idle like Dirtdad's? Oh regarding increasing the phase wire thickness on the motor... someone gave me the idea to bore thru the solid side of the axle so you can bring wires out the axle in both directions. Might need a lathe to bore the axle out properly and then be very carefull that the wires coming out the drive side don't get caught in the chain or sprokets.
 
You could also try to find some higher current rated cable of the same awg and size. Look for Ultra High Strand Count wire. Trossen Robotics used to carry it (it was silicone jacketed, too) and gave its Ampacity rating but looks like they only have some yellow 12 and 14 awg left in their garage sale area and it doesn't give its Ampacity. I remember it was rated at least 1.5 and maybe 2 times that of high strand count wire found elsewhere.

I just remembered Dean's Wet Noodle and Ultra wires with even more strands per size:

http://www.robotmarketplace.com/products/wire_and_accessories.html
 
BVH,

Can I ask a dumb question? Would a solid wire be better for passing higher amps? My vague understanding of electricity is that the higher the cross section of a conductor, the more amps it can pass without heating up. I.E., in data centers that use DC power they use a huge copper bar to provide power because it has the most cross section. therefore, for a given diameter, would a solid copper wire handle more amps without heating up?

I tried to look this up, but I didn't find much helpful info, and It's been a long time since I studied this stuff.
 
I'm certainly no expert. What I've been taught is that current flows along the outside of a strand of wire. The more strands in a given sized cross section, the more paths for the current to follow. In a given wire size and given use, I typically see higher strand count wire spec'd by manufacturers as having more current carrying capability.
 
Stranded vs. Solid Wire
This one is a bit of a mind-boggler, but it's important. When electricity flows through a wire, it mostly flows on the surface of the wire, not through the middle. This effect is more pronounced on high frequency AC than it is on DC or low frequency AC. This means that a "wire" of a given size that made up of many smaller strands can carry more power than a solid wire - simply because the stranded wire has more surface area. This is one reason why battery cables in your car and welding cables are made up of many very fine strands of smaller wire - it allows them to safely carry more power with less of that power being dissipated as heat. However, this "skin" effect is not as pronounced in a typical 12V DC automotive application, and the wire and cable used there is stranded for flexibility reasons.

When looking at a chart or description of wire capacity, take note of whether it is referring to stranded or solid wire - some charts may not specify but instead assume a default based on the typical wiring used in a given application. For example, almost all automotive wiring is stranded while almost all home wiring is solid. For most applications, flexibility or the lack thereof will be more important, but for very high frequency AC applications, stranded wire might be a requirement.

http://www.rowand.net/Shop/Tech/WireCapacityChart.htm

also see ... http://en.wikipedia.org/wiki/Skin_effect

OTOH ...
At high frequencies, current travels near the surface of the wire because of the skin effect, resulting in increased power loss in the wire. Stranded wire might seem to reduce this effect, since the total surface area of the strands is greater than the surface area of the equivalent solid wire, but in fact a simple stranded wire will have worse skin effect than a solid wire, because of its increased average resistivity due to inclusion of air gaps within the wire.

http://en.wikipedia.org/wiki/Wire#Solid_versus_stranded
 
Great info Knuckles. You may have posted something similar recently in reply to a post I made on the same subject. I somewhat understood back then but this is more clear.
 
http://endless-sphere.com/forums/viewtopic.php?f=2&t=10878&start=15#p167109
 
Sacman said:
Good stuff Muad'dib. Glad to see you finally got it up and running. Yeah these Crystalyte controllers are really nice. I just got back from riding my ebike with the new controller installed and it's a lot more accelleration and uphill torque than I'm used to (but I'm sure I'll adjust quickly) :mrgreen: . Do you have a Cycle Analyst hooked up to your rig? It is still pulling a lot of amps when it's sitting idle like Dirtdad's? Oh regarding increasing the phase wire thickness on the motor... someone gave me the idea to bore thru the solid side of the axle so you can bring wires out the axle in both directions. Might need a lathe to bore the axle out properly and then be very carefull that the wires coming out the drive side don't get caught in the chain or sprokets.
Click to expand...

I just hooked up the Cycle Analyst, but haven't calibrated it yet. I tried lifting up the rear wheel to see what kind of amps it drew and it was around 5 or 10. I have blown quite a few fuses this weekend––all at slow speeds--but figured out that if I am easy on the throttle until I get going faster it won't blow (40amp fuse). I'll be ordering some 50amp fuses, but the problem is still these phase wires heating up. I took it out for a 10 minute ride to push and see what it would do, and I took it up some hills and a mile or two around. Back up to my place is a rather steep hill and it made it up fine, but when I stopped the controller was pretty hot, I could hold my hand there, but it was definitely hot (motor was cool). I checked the wires and saw that the plastic covering/sheathing for the motor wires had melted away, so this is going to need something changed. I'll tell the dealer about it for sure, because I am only running it at his recommended wattage--2500. He says the lower power controllers won't work with it anyway, so sounds like a major manufacturing design flaw passed on to the consumer.

Thanks for the idea about drilling the axel, that will be an option on the table.
 
Muad'dib said:
I just hooked up the Cycle Analyst, but haven't calibrated it yet. I tried lifting up the rear wheel to see what kind of amps it drew and it was around 5 or 10. I have blown quite a few fuses this weekend––all at slow speeds--but figured out that if I am easy on the throttle until I get going faster it won't blow (40amp fuse). I'll be ordering some 50amp fuses, but the problem is still these phase wires heating up. I took it out for a 10 minute ride to push and see what it would do, and I took it up some hills and a mile or two around. Back up to my place is a rather steep hill and it made it up fine, but when I stopped the controller was pretty hot, I could hold my hand there, but it was definitely hot (motor was cool). I checked the wires and saw that the plastic covering/sheathing for the motor wires had melted away, so this is going to need something changed. I'll tell the dealer about it for sure, because I am only running it at his recommended wattage--2500. He says the lower power controllers won't work with it anyway, so sounds like a major manufacturing design flaw passed on to the consumer.
Click to expand...

Very nice ... thanks for all that good info.

Oh yeah the Cycle Analyst I finally hooked up was a Stand Alone version with it's own known shunt resistor so I didn't need to calibrate. I did have the CA Direct Plug in version at first and was trying to get Max Kogan to get the correct RShunt value from Crystalyte so that it can be properly calibrated but we couldn't get that info from Crystalyte. You wouldn't believe the number of emails back and forth between me and Max and to Kenny at Crystalyte ... and Crystallyte still ended up not able to provide that info. Good luck on getting your CA calibrated.

BTW Dirtdad, E-Ticket, and I normally get together once a month to do an e-bike ride together (it's become our tradition), and yesterday we all just finished a 40-mile ride. So the nice long ride gave us a a good chance to try out our equipment and so we ran lots of experiments.

Dirtdad used his 1000W BMC motor and a 36-72V 20A Crystalyte controller. Why only 20A controller?... I guess for extended range and because that 20A controller seems to be the most stable controller he's tested so far with that 1000W BMC motor. We swapped bikes part way so I got a chance to ride his bike for a while. His setup ran fairly smooth as expected but I think his 20A controller was really limiting his motor from coming to life. Cycle Analyst showed that his max current was only 17A. Also, it was a hot day to ride and so his little 20A controller kept overheating on long, steep hills and kept shutting down until it cooled off... and it would work again.

Dirtdad did make an interresting observation with the !000W BMC motor and switching between 36V and 72V. The voltage didn't make that much a difference in top speed. At 36V he could get to about 27mph. Doubling the voltage to 72V only got him to 33mph. Twice the voltage only gave a 6mph difference. So it looks like the 1000W BMC motor is made to run more efficient at 36V or 48V.

Now I'm running the same Crystalyte controller as yours but in 72V 50A version with my 600W BMC (Torque) motor. This setup turned out to be "very torquey" (as Dirtdad put it) and it made hill climbing and accellerating a breeze. And at 72V it was able to hit 34-35mph. Controller had no problems with heat and was fairly cool to the touch the whole time. The motor felt warm but not hot and no signs of wires overheating. Cycle Analyst showed my max current at only 36A (so controller still had a lot more headroom before maxing out). I've got more work and more tests to run but for now I'm pretty happy with this motor-controller combination.

-Sacman
 
Sacman said:
Oh yeah the Cycle Analyst I finally hooked up was a Stand Alone version with it's own known shunt resistor so I didn't need to calibrate. I did have the CA Direct Plug in version at first and was trying to get Max Kogan to get the correct RShunt value from Crystalyte so that it can be properly calibrated but we couldn't get that info from Crystalyte. You wouldn't believe the number of emails back and forth between me and Max and to Kenny at Crystalyte ... and Crystallyte still ended up not able to provide that info. Good luck on getting your CA calibrated.
Click to expand...

How off could it be by default? I could live with 10 or 15% error.

Sacman said:
Now I'm running the same Crystalyte controller as yours but in 72V 50A version with my 600W BMC (Torque) motor. This setup turned out to be "very torquey" (as Dirtdad put it) and it made hill climbing and accellerating a breeze. And at 72V it was able to hit 34-35mph. Controller had no problems with heat and was fairly cool to the touch the whole time. The motor felt warm but not hot and no signs of wires overheating. Cycle Analyst showed my max current at only 36A (so controller still had a lot more headroom before maxing out). I've got more work and more tests to run but for now I'm pretty happy with this motor-controller combination.
-Sacman
Click to expand...

I learned that increasing voltage does not increase wire heating, so one thing that would help is for me to go to a higher voltage in order to limit the current used.

I set the CA to 30 amps max, so no more blowing fuses. The motor wires are still unfortunately very hot at this level of current--oh, and I shortened them as much as possible--so the problem is not solved. At least no more burning through the sheathing. Also, while I am happy with the speed and torque on level ground, I am in a hilly area and I bought this expensive motor for that reason, but it doesn't have great power for real hills. At the 30amp limit, it starts drawing that much whenever I hit any kind of hill, and on the steeper ones it bogs down a lot. I push it up the really steep hills, but if I am careful in routing my ride I can avoid most of those.
 
I stripped back the outer sheathing on my 1000 V3 and the phase wires are MAYBE 16 awg and very inflexible so a very low strand count. I'm going to cut off most of the factory wiring and leave about 4 inches out of the axle and then solder in 12 AWG, ultra high strand count wires up to the controller plug. Nothing I can do about the wire passing thru the axle. Too bad. 'Even if the factory wires are 12 awg, they're too small according to Ampacity charts. I think wire size is part of what's holding back this motor. I've also heard that limiting current to 30 Amps give the best performance given the restraints.

Once I try mine out, if It doesn't perform well, I may just buy the 600 Watt Torquer and mount it to a second wheel as an option depending on where and how I'm going to ride.
 
i was able to get 3 solid 12G wires into a goldenmotor hub, along with all 5 hall wires. you can google up the wire sizes and the cross sectional area is almost 5 times larger for 12G solid over the 18G stranded that was stock.

http://endless-sphere.com/forums/viewtopic.php?f=2&t=10667#p163404

the skin effect is of no significance for conduction at the frequency (100hz) of the phase wire comuttation.

if you can install the controller close to the motor you can reduce the power losses that way as well. i intend to install this goldenmotor controller up front above the fork to cut the phase wire lengths in half so conductivity is 10X better than stock 18G stranded.

i would expect if you are able to push that much power through the geared motor, then it will eventually get hot, but is delayed since the heat cannot escape directly since the motor is inside and separate from the hub.

i can't remember who now, but i think it was geoff, who was able to increase the conductivity by using flat braid for the phase wire conductors through the axle on his BMC when he rebuilt it. that may be in the archives since it was a classic.

i think knuckles ran 12G solid into his gruber too. or another motor he had.
 
Good info dnmun, Thank you. I wanna ride this thing! Had it for 4 days now and have not even ridden it 2 inches! So I'll start by doing the splicing as described above and see how much of a problem I have/don't have. I tend to "baby" my equipment so it may be an issue I'm fretting over for nothing. If I do have a problem, then I might try the 12 ga solid trick you did.
 
i had the opportunity to push in new phase wires since i had to open it up and replace the hall sensors and pull phase wires anyway (i will never leave the wire inside the axle of a twistoff after seeing the damage inside the axle).

the solid is a disadvantage because it has to be kept rigid, so the flex at the fork will be in the wires coming out of the controller and i can strap the 12G solid onto the fork so it doesn't flex. also have to avoid damaging the solid by flexing it a lot while working and then it has to be large radius leaving the axle, which is part of the downside.

i actually think the 12G was overkill, and have kept my eye out for 14G stranded to use if i do that again. it really was a bitch getting the wire to turn out of the hole at the bottom, but i have the picks to hook the end and pull it over gradually as i pushed it through. binding up on the hall wires was the other big problem, they tear easily too so i had shorts initially where the hall wires had torn but i got them all inside the hub.
 
I just thought of something. Is each Phase wire carrying full controller allotted current 100% of the time or is the current divided between them? I'm definitely not a motor-knowledgeable person yet.
 
Hmmmm..... i might be looking into also upgrading the phase wires on my BMC motor as well since I just melted some wires I place temporarily on the phase wire connectors climbing mountains earlier today. I just made a thread on it here http://endless-sphere.com/forums/viewtopic.php?f=4&t=11767

These original phase wires are just teflon coated 16 gauge silver stranded wires right? So dnmun (depending on the wire insulation) you think we can fit 14 or even 12 gauge wires in that axle?
 
i don't know the diameter of the hole through the BMC axle and how close it is to the goldenmotor.

the 12G solid i used was covered with an extra 'slick sleeve' on top of the insulation so that made it even larger. also i left the twist intact on the 12G 3 wire bundle and that may have also been a mistake. maybe if it ran straight in with the hall wires positioned carefully in between the 3 phase wires it would fit easier. but once i started, i wasn't gonna change my mind when i finally pushed the phase wires to the opening at the bottom inside the hub. taurus is my sign.

like i said, i thought using solid would make it easier to push, rather than pushing stranded, which is why i thought of using the 12G solid. plus it was free from dumpster diving, had the slick sleeve and it had the 3 colors green black white so i could not get too stupid to remember which color was which, an important requirement at my age. BYG=BWG.

now i think the stiffness WAS the problem when getting it out at the inside, and i don't think the extra conductivity over 14G stranded is worth the extra hassle, so that's why i am thinking i will use 14G stranded if i do it again.

i was surprised reading your comments about how they melted under only about 250-350W continuous, but that was a climb. hope it did not damage the controller. i think patching in 14-12G right down to the axle is worth doing too, but i am surprised the motor is not getting hot. maybe because the power is going into heating the phase wires outside. maybe even solder up the connectors and the place where the wire is crimped in the connector and then also run another 16 or 14G wire in parallel tied on the outside of the bundle and soldered on top of the phase wires close to the axle. you would have to open the sleeve to do it so stagger the joints so it can be resealed easier after.
 
dnmun said:
i was surprised reading your comments about how they melted under only about 250-350W continuous, but that was a climb. hope it did not damage the controller. i think patching in 14-12G right down to the axle is worth doing too, but i am surprised the motor is not getting hot. maybe because the power is going into heating the phase wires outside. maybe even solder up the connectors and the place where the wire is crimped in the connector and then also run another 16 or 14G wire in parallel tied on the outside of the bundle and soldered on top of the phase wires close to the axle. you would have to open the sleeve to do it so stagger the joints so it can be resealed easier after.
Click to expand...

Not just 250-350W continuous... I was actually doing WOT burning 1600W continuous for 4 minutes trying to catch up to my friends on the trail ahead of me when the wires melted and i lost power. I posted a pic of the melted wires on that thread but that's it... I done and tired to worry any more about it for today. I'll look at it more tomorrow but I'm hoping the controller is okay. :cry:

I'm definitely thinking about upgrading the motor's phase wires tho.
 
Please report back with your results. I hate doing a job halfway which is what splicing outside of the motor is doing.
 
it's more than halfway, most people end up pushing in bigger phase wires when they have to rebuild because of a twistoff, which was my case.

doubling up the phase wire to the axle and soldering all the connections to get maximum conductivity to that point is more than 1/2 way. maybe 90%.

at max, maybe only an extra 25-40W more power or cooler motor is the result, and of course the wires don't melt the insulation, but that was really only in his jumper.

the geared motor may have a smaller axle hole than my direct drive hub too.
 
OK, you got me! It was a figure of speech! I should have said half a$$. :oops: Yes, it will go a long way in helping. I just don't like the thought of a few inches of those darn small wires left in the "chain" - the weak link, so to speak.
 
You must log in or register to reply here.

Similar threads

Q
Broken display EV-meter s890
Replies
8
Views
380
qvalff8
Q
V
Laws of Three Phase Wiring?
Replies
10
Views
366
Vector
V
Z
Ecotric Swallow 14 inch information needed
2
Replies
28
Views
633
zort
Z
M
60v 33.8Ah Battery with 70A BMS, 60v controller 48v 1000w motor Advice/Will it work.
Replies
4
Views
479
MrMoneymitch
M
adamsavage79
Adapter Cable for KT LCD3
Replies
4
Views
258
adamsavage79
adamsavage79
Back
Top