10S custom skate ESC: testers wanted!

[chaka]

Long motor wires are fine. 2000uf of capacitance is usually safe to go 12 inches in battery wire length, you could probably push this a little.

 

[psychotiller]

Lot's of people use Lipo backpacks. https://www.facebook.com/RedRockBoardShop/

Problem is, Having a couple of long battery leads dangling between a backpack and a board pushing speeds in excess of sometimes 30mph isn't cool. Actually it's quite dangerous. Even if you stood on a board in a completely static position while riding (which you can't) it still sucks.

 

[Error]

excited to see the performance of the esc for 10s!

 

[hexakopter]

Ok, thanks for the answers. I will maybe think a bit longer if I will risk to be tied to the board. I think on an bike it would be more useful, because you are sitting all the time and the backpack don't move "a lot".

@chaka I was more thinking about 1-1.5m wire and not only 12 inches.You say motor wire can also be longer. But it is an bad idea to have the VESC in the backpack also, right? So the "longer" is maybe also just 12 inches and not more i think.

 

[chaka]

High output VESC!!!

 

[jacobbloy]

Introducing the dual VESC made for simplicity.
This is a twin ESC designed and based off of the VESC 4.10.
Benefit:
compact design when compared to wiring 2 vesc to gether
Less soldering and wiring
Batteries input is shared between both ESC
Capacitors are pcb mountable at 45 or 90deg
Both ESC connected via can bus
1 USB input
Built in nrf24l01 footprint ready for nunchuk control
Reduced componants by
Both ESC share 3v and 5v power rail
Both share the 1 set of LEDs
And again 1 USB and reduces the need for installing jst plugs meaning a small componant reduction.

Foreseen problems:
3v and 5v Power rails not supplying enough amp
Batterie input traces not handling required current
Added cost of 6 layer pcb
(Please add more)

Ways we have tried to fix:
Made pcb 6 layer to allow 3layer thick power traces

 

[torqueboards]

Well done Bloy! Looks great!

Heatsink looks good as well Chaka!

 

[akiraEC]

High output VESC!!!

Nice one !!
Is that a second VESC below the first or a heat transfer metal plate ?
I guess you have the mosfet below pressed against the plate for cooling right ?

 

[Vanarian]

Elkick Thank You again I got your PM right! That's awesome to get VESC straight from Europe

Chaka Nice unit but apart from heatsinks I fail to see the difference, is there thermal past applied to transfer heat?

Jacobbloy This is what Jason was talking about! Just a question, why put it the square way instead of longitudinal way? I thought it would be slimmer. Nice work anyway.

 

[exco]

Introducing the dual VESC made for simplicity.
Foreseen problems:
3v and 5v Power rails not supplying enough amp
Batterie input traces not handling required current
Added cost of 6 layer pcb

Do you supply each "side" individually from the DRV8302 buck converter?
or are these paralleled (how?)? How much current is anticipated / needed on the 5V / 3.3V rails?

Could it make sense to have one DRV8302 do the 3.3V rail, the other the 5v one?

Regarding thermal design - it could make sense to move the Mosfets to one side because
now (if pushed beyond the power the longboards need at the moment) the fets heat each
other.
If moved to one side - the other side could be used for cooling.

Another suggestion: high current pcb mounted screw terminals

 

[Dr_T]

High output VESC!!!

Weren't the shunts the bottleneck of output? Or did you change those too?

 

[chaka]

Chaka Nice unit but apart from heatsinks I fail to see the difference, is there thermal past applied to transfer heat?

Yes, thermal adhesive is used on the fets. I have found the VESC to overheat when riding at higher speeds, 30mph. Heat sinking the vesc in this manner has alleviated those issues.

Weren't the shunts the bottleneck of output? Or did you change those too?

This unit is what I ride on. I usually ride with a constant output around 800 watts. Without heat sinks I was getting thermal shutdown after extended high output sessions.

Introducing the dual VESC made for simplicity.
This is a twin ESC designed and based off of the VESC 4.10.
Benefit:
compact design when compared to wiring 2 vesc to gether
Less soldering and wiring
Batteries input is shared between both ESC
Capacitors are pcb mountable at 45 or 90deg
Both ESC connected via can bus
1 USB input
Built in nrf24l01 footprint ready for nunchuk control
Reduced componants by
Both ESC share 3v and 5v power rail
Both share the 1 set of LEDs
And again 1 USB and reduces the need for installing jst plugs meaning a small componant reduction.

Foreseen problems:
3v and 5v Power rails not supplying enough amp
Batterie input traces not handling required current
Added cost of 6 layer pcb
(Please add more)

Ways we have tried to fix:
Made pcb 6 layer to allow 3layer thick power traces

Thanks Jacobbloy! Great addition to the open source community! Did you upload this to the github yet?

 

[chaka]

We have a shipment of VESC compatible vertical usb ports on order so we can make this happen. Looks to like we can save about 10mm in width by sticking with Vedder's original design!

 

[okp]

Hey,

I got my two new VESCs installed using CANBUS on my upcoming e-MTB. Everything is fine except that I got an error code : OVER_VOLTAGE

I am using four 3S 5AH in series to get a 12S 5AH with SK6374 168kv - dual VESC with CANBUS. PPM set on VESC ID 0.

I got my settings on BLDC set to Voltage limits
- Minimum Input : 8
- Maximum Input : 50
- Battery Cut Off start: 40,80
- Battery cut off end : 38,40

My currents limits should not affect this error code
- Motor Max : 60
- Motor Min : -60
- Batt Max : 60
- Batt Min (regen) : -20
- Absolute Max : 130

EDIT : I found this piece of information and will try to move it to 57V.

This is weird, really as I am charging my lipos to 95% and I don't really get why with the 4.10 I got this error code...

Will keep you posted folks

 

[boesi]

Hey,

I got my two new VESCs installed using CANBUS on my upcoming e-MTB. Everything is fine except that I got an error code : OVER_VOLTAGE

I am using four 3S 5AH in series to get a 12S 5AH with SK6374 168kv - dual VESC with CANBUS. PPM set on VESC ID 0.

I got my settings on BLDC set to Voltage limits
- Minimum Input : 8
- Maximum Input : 50
- Battery Cut Off start: 40,80
- Battery cut off end : 38,40

My currents limits should not affect this error code
- Motor Max : 60
- Motor Min : -60
- Batt Max : 60
- Batt Min (regen) : -20
- Absolute Max : 130

Anyone got a clue about that?

I guess your maximun input voltage is to low. If you have a fully loaded 12s battery the voltage could reach 50.4V (12 * 4.2V). So increase the Maximum Input Voltage to 51V and everything will be fine.

Hopefully this woks for you.. have a nice ride

 

[okp]

Thanks ! You were definitely right. I checked on my Vanguard and it's set to 57V. Works perfectly. I will get to the store tomorrow to get some screws and will post my E-MTB build on dual VESC 4.10 CANBUS FW1.14 on 12S. This thing is going to be amazing.

Take care everyone.

By the way. I got a hall sensor detection error on the motor detection. Never noticed this error, but I assume it is because I don't have any sensor on these motors. The other two values are well detected. 1200 & 120.

 

[elkick]

By the way. I got a hall sensor detection error on the motor detection. Never noticed this error, but I assume it is because I don't have any sensor on these motors. The other two values are well detected. 1200 & 120.

That's quite normal if you're using them without sensors. Enjoy your ride!

 

[vedder]

FOC SUPPORT!

I just pushed the first FOC implementation to github with a quite massive commit:
https://github.com/vedderb/bldc/commit/ff090fd4e0028f6a8ceb28f5441c1d0cca40af7e

This is a screenshot of the FOC configuration page in BLDC Tool:

R and L can be autodetected without starting the motor, and lambda (flux linkage) can be detected by spinning up the motor in bldc delay commutation mode. These are the important parameters to get the sensorless observer running. There is also automatic PI current control parameter calculation as described here (the time constant has to be chosen, but this is not critical for eboards):
https://e2e.ti.com/blogs_/b/motordrivecontrol/archive/2015/07/20/teaching-your-pi-controller-to-behave-part-ii

I have tested this with a bunch of different motors, and it seems to work well so for. For my ebike I had to set the detected resistance a few % lower than measured for startup from a complete stop to work. I will make a video tutorial about FOC configuration soon.

 

[Stielz]

Awesome work Vedder! Cant wait to try this

 

[Hummina Shadeeba]

!!! Been waiting for this

 

[okp]

awesome ! means a new firmware with a compatible BLDC right?

 

[elkick]

Thanks Benjamin, that's amazing!

Stupid question maybe: do we have to put commutation mode back to "integrate" after detection and before "write config" in the BLDC tab? Or should it stay in "delay" mode?

@okp, yes!

 

[jacobbloy]

We have a shipment of VESC compatible vertical usb ports on order so we can make this happen. Looks to like we can save about 10mm in width by sticking with Vedder's original design!

why don't you flip the second esc so that the the 2 can bus are next to each other the same as what i have done, the you only need 1 usb.

 

[jacobbloy]

here are a the software packages for windows and OS X if you need them and want to test FOC

OS X BLDC TOOL 2.3 FOC
https://drive.google.com/file/d/0Bym9XrdeViekVG9vQUU1S2ljT2s/view?usp=sharing

WINDOWS BLDC TOOL 2.3 FOC
https://drive.google.com/file/d/0Bym9XrdeViekNVdRVUN0M2VTMkE/view?usp=sharing

Motor configuration and Firmware
https://drive.google.com/file/d/0Bym9XrdeViekb3lzcGNDdjk4QUU/view?usp=sharing

 

[elkick]

Sorry, my question some posts before was stupid indeed since it doesn't matter when FOC is selected in the Motor tab. So please forget about it.

It's just awesome, I don't hear the motor (6354) anymore while accelerating and driving.

There is only a little noise while braking sharp, but that's almost nothing compared with the former BLDC setup. I'm really excited!