10S custom skate ESC: testers wanted!

Lightweight / Folding / Portable EVs - seats optional
User avatar
sl33py   10 kW

10 kW
Posts: 686
Joined: Apr 24 2015 5:21pm
Location: Seattle, WA

Re: 10S custom skate ESC: testers wanted!

Post by sl33py » Jul 20 2015 5:27pm

@fasser - what motor kv? I saw something here about >200kv on 12s could be a problem.

User avatar
onloop   10 kW

10 kW
Posts: 600
Joined: Jun 24 2013 5:24am
Location: Australia
Contact:

Re: 10S custom skate ESC: testers wanted!

Post by onloop » Jul 20 2015 7:19pm

fasser wrote:
Another thing is that when I was testing the VESC even if I set the motor current limit to 120A The realtime graph never showed more than 80A and it did not feel like it was more than 80A. Is there some type of limit I did not recognize yet? The limit was exactly 80A.

Best regards,
fasser
The vesc also has temp limiting function.. I think the temp sensor is positioned near fets. What do you have it set at in BLDC

fasser   100 µW

100 µW
Posts: 7
Joined: Feb 11 2015 12:15pm
Location: Zurich

Re: 10S custom skate ESC: testers wanted!

Post by fasser » Jul 21 2015 6:12am

Hi everybody

Thanks for your answers.
sl33py wrote:@fasser - what motor kv? I saw something here about >200kv on 12s could be a problem.
The motor is an EMAX GT5345/09 with 170KV. The error occured at around 9000 ERPM in sensorless mode, so I don't think it's the reported KV limit.
onloop wrote:
The vesc also has temp limiting function.. I think the temp sensor is positioned near fets. What do you have it set at in BLDC
I left the temp limit as default at 80°C start and 100°C end for the current limiting. But as mentioned the temperature was not reported near this limits when the VESC died.


I'm asking myself if maybe the batteries are limiting the current, but it would be strange that it is exactly at 80A. I'm using two Turnigy 5000mAh 6S 20C Lipos in series.
http://hobbyking.com/hobbyking/store/uh ... duct=35828
They should have 150 Amps peak discharge. Does anybody have some experience with these? I think they're quite common.

Best regards
fasser

pf26   100 mW

100 mW
Posts: 39
Joined: Nov 17 2014 1:37pm

Re: 10S custom skate ESC: testers wanted!

Post by pf26 » Jul 21 2015 10:10am

Hi,
I'm also using two Turnigy 5000mAh 6S 20C Lipos in series and 168kV motor. I had no issue so far (after some 100 kilometers now). The VESC works great for me.
But my current limit is set at 60A (discharge) and 40Amp charge (for mechanical reason, otherwise my timing belt might slip then). I also use a low amp limit on battery currents (10Amps, that is 2C for regen - but also against overvoltage).
I have a 1500uF/63V Nichicon capacitor (60mOhmESR / 3Amp ripple) + a transil1.5kW56V (1.5KE56A) very near to the VESC.

In your case, I suspect an overvoltage killing DRV8302 chip.
Were the batteries fully charged ? Was the motor loaded ? Did the VESC fail upon deceleration ?
What capacitor did you use ? How near is the capacitor from the PCB ?

I wish there was a DRV with a higher max voltage, also because we could then use a more common 13s battery (possibly using other FETs).

silviasol   10 kW

10 kW
Posts: 878
Joined: Dec 30 2012 8:13pm

Re: 10S custom skate ESC: testers wanted!

Post by silviasol » Jul 21 2015 10:18am

My drv8302 died after one ride with my 270kv motor and 12s. After talking with vedder I found that you can not use more then 200kv with 12s. If you gave it alot of throttle at low speed would the vesc cut off then back on? That is what mine did, otherwise after about 15mph no issue's going WOT. I got home the next day and tested sensored and it cut off at high speeds then shortly after I got drv8302 errors. Might still be kinks that need to be worked out with 12s even with 200kv or lower. I am working on my 10s pack now, have my fingers crossed it will work well. No issue's with 6s right now. If you want me to replace your drv8302 message me.
Current project: 10s 20ah sensored goped hoverboard
http://endless-sphere.com/forums/viewto ... 35&t=62155

User avatar
Vanarian   10 kW

10 kW
Posts: 556
Joined: Nov 26 2014 12:24pm
Location: France!

Re: 10S custom skate ESC: testers wanted!

Post by Vanarian » Jul 21 2015 11:15am

fasser wrote:Hi everybody

Thanks for your answers.

The motor is an EMAX GT5345/09 with 170KV. The error occured at around 9000 ERPM in sensorless mode, so I don't think it's the reported KV limit.

I'm asking myself if maybe the batteries are limiting the current, but it would be strange that it is exactly at 80A. I'm using two Turnigy 5000mAh 6S 20C Lipos in series.
http://hobbyking.com/hobbyking/store/uh ... duct=35828
They should have 150 Amps peak discharge. Does anybody have some experience with these? I think they're quite common.

Best regards
fasser
Hi fasser, that's quite the beast you have to drive there! Maybe I'm misunderstanding if it is the setting or real RPM but why do you want to spin the motor at 9k RPM? The VESC can handle >200KV motors but overspinning them is useless (you lose power) and frenquency of erpm can damage the DRV.

BTW if you want killer LiPos to get huge 150 discharge peaks, you can't go with Turnigy 6S 20C Lipo you are asking too much of only one pack of 5Ah. Either you parallel more packs or you get something with less internal resistance, but less internal resistance also equals more expensive. Though I wouldn't burst 150A on a RC sized motor, it is a LOT of heat to bear for the motor.

May I suggest that you look at a123 LiFepo4 cells? 26650 cells are rated for 70A continuous at 2,5Ah each for 3.3 nominal per cell. Stack 20 cells in 10S2p and you can have 33/36V 140A capable continuous current from your battery pack, but hey it costs more than Hobbyking lipos! Your choice is defined by how much you want to pay for your battery.
-+AT-One project : electric inline skates - no remote, hub motors, one frame for various wheel sizes & quasi-universal boot mount

Rollo Ergo Sum!

pf26   100 mW

100 mW
Posts: 39
Joined: Nov 17 2014 1:37pm

Re: 10S custom skate ESC: testers wanted!

Post by pf26 » Jul 21 2015 11:33am

I think fasser said 9k ERPM. So that's maybe just 1285RPM for a motor with 7 pairs of poles (not sure this one actually is).
The motor voltage should then be around 8 volts (=1285/170), so 100Amps in the motor yields around 18Amps for the battery (assuming 44Volts), which is Okay for those Lipos. Or am I wrong somewhere ?

User avatar
Vanarian   10 kW

10 kW
Posts: 556
Joined: Nov 26 2014 12:24pm
Location: France!

Re: 10S custom skate ESC: testers wanted!

Post by Vanarian » Jul 21 2015 12:09pm

pf26 wrote:I think fasser said 9k ERPM. So that's maybe just 1285RPM for a motor with 7 pairs of poles (not sure this one actually is).
The motor voltage should then be around 8 volts (=1285/170), so 100Amps in the motor yields around 18Amps for the battery (assuming 44Volts), which is Okay for those Lipos. Or am I wrong somewhere ?
Ok I understand better for the ERPM :) But if I'm not wrong, 100 amps in the motor means 100 amps from the battery, because even when dividing all amps between each poles you must have the total amps from the start... so 100 amps from a pack of 5Ah 20C is asking too much of it IMHO ?
-+AT-One project : electric inline skates - no remote, hub motors, one frame for various wheel sizes & quasi-universal boot mount

Rollo Ergo Sum!

lizard   100 mW

100 mW
Posts: 43
Joined: Oct 29 2014 6:38am
Location: Germany

Re: 10S custom skate ESC: testers wanted!

Post by lizard » Jul 22 2015 1:28am

Here are some load tests with different "R/C-grade" lipos:
http://www.elektromodellflug.de/hobby-king-40-65c.html (german, but the load diagrams on the right are in english).
http://www.rcgroups.com/forums/showthread.php?t=1767093

But apart from all those numbers, as far as I understand, it's pretty easy to find out if your lipos are up to the task:
- Check if single cells drop below 3.3 or 3.2v under load ( can be checked with a cheap lipo warner buzzer connected to the balancer leads)
- Check how warm they get. Opinions differ on this, but I'd say if they go over 50 degrees Celsius regularly it's too much for them.

fasser   100 µW

100 µW
Posts: 7
Joined: Feb 11 2015 12:15pm
Location: Zurich

Re: 10S custom skate ESC: testers wanted!

Post by fasser » Jul 22 2015 9:25am

Thanks everybody!!
pf26 wrote:Hi,
I'm also using two Turnigy 5000mAh 6S 20C Lipos in series and 168kV motor. I had no issue so far (after some 100 kilometers now). The VESC works great for me.
But my current limit is set at 60A (discharge) and 40Amp charge (for mechanical reason, otherwise my timing belt might slip then). I also use a low amp limit on battery currents (10Amps, that is 2C for regen - but also against overvoltage).
I have a 1500uF/63V Nichicon capacitor (60mOhmESR / 3Amp ripple) + a transil1.5kW56V ( transil1.5kW56V) very near to the VESC.

In your case, I suspect an overvoltage killing DRV8302 chip.
Were the batteries fully charged ? Was the motor loaded ? Did the VESC fail upon deceleration ?
What capacitor did you use ? How near is the capacitor from the PCB ?

I wish there was a DRV with a higher max voltage, also because we could then use a more common 13s battery (possibly using other FETs).
Yes, maybe overvoltage is the problem. I had a 2000uF/63V Nichicon capacitor placed in the same way as on the Onloop BETA batch. So approx 2-3cm away from pcb connecting points.
The VESC failed at full throttle in current mode and was still accelerating. It felt a bit like cogging in the last moment, but it's hard to say if this was the reason for the VESC to fail or if this was the result of the failed VESC.
Maybe I should add a second capacitor or a transil 1.5kW56V.
Concerning the transil1.5kW56V: I am right that this is a diode that shorts the circuit if the voltage rises over 56V?
silviasol wrote:My drv8302 died after one ride with my 270kv motor and 12s. After talking with vedder I found that you can not use more then 200kv with 12s. If you gave it alot of throttle at low speed would the vesc cut off then back on? That is what mine did, otherwise after about 15mph no issue's going WOT. I got home the next day and tested sensored and it cut off at high speeds then shortly after I got drv8302 errors. Might still be kinks that need to be worked out with 12s even with 200kv or lower. I am working on my 10s pack now, have my fingers crossed it will work well. No issue's with 6s right now. If you want me to replace your drv8302 message me.
Yes, in current control mode the VESC cut off a few times if I did not apply ramping. What surprised me was that even if I set the current limit to 50A and the absolute current limit to 150A, I got Absolut_current_limit fault codes. I'm using an arduino to convert the signal of an original wired wii nunchuk to ppm. I did not encounter this problems after integrating ramping in my arduoino code. If somebody is interested in this Arduino code I can post it here, but it's quite crappy.
Vanarian wrote: Hi fasser, that's quite the beast you have to drive there! Maybe I'm misunderstanding if it is the setting or real RPM but why do you want to spin the motor at 9k RPM? The VESC can handle >200KV motors but overspinning them is useless (you lose power) and frenquency of erpm can damage the DRV.

BTW if you want killer LiPos to get huge 150 discharge peaks, you can't go with Turnigy 6S 20C Lipo you are asking too much of only one pack of 5Ah. Either you parallel more packs or you get something with less internal resistance, but less internal resistance also equals more expensive. Though I wouldn't burst 150A on a RC sized motor, it is a LOT of heat to bear for the motor.

May I suggest that you look at a123 LiFepo4 cells? 26650 cells are rated for 70A continuous at 2,5Ah each for 3.3 nominal per cell. Stack 20 cells in 10S2p and you can have 33/36V 140A capable continuous current from your battery pack, but hey it costs more than Hobbyking lipos! Your choice is defined by how much you want to pay for your battery.
Thanks Vanarian. The reason I use these Lipo's is that I just needed a cheap and easy way to test my system and I underestimated occuring power peaks. So I think for the moment I will just buy two more Lipo's for $80. I will look for a nice battery pack with BMS as soon as I really know my needs.
Since these are only short power peaks the motor temperatur was never really a problem.
Vanarian wrote:
pf26 wrote:I think fasser said 9k ERPM. So that's maybe just 1285RPM for a motor with 7 pairs of poles (not sure this one actually is).
The motor voltage should then be around 8 volts (=1285/170), so 100Amps in the motor yields around 18Amps for the battery (assuming 44Volts), which is Okay for those Lipos. Or am I wrong somewhere ?
Ok I understand better for the ERPM :) But if I'm not wrong, 100 amps in the motor means 100 amps from the battery, because even when dividing all amps between each poles you must have the total amps from the start... so 100 amps from a pack of 5Ah 20C is asking too much of it IMHO ?
I'm not an expert but I think Vanarian is right, it's not the voltage that goes down if the motor isn't running at full speed, it's the efficiency, meaning more energy is converted into heat (in the motor).
Can somebody approve this?
lizard wrote:Here are some load tests with different "R/C-grade" lipos:
http://www.elektromodellflug.de/hobby-king-40-65c.html (german, but the load diagrams on the right are in english).
http://www.rcgroups.com/forums/showthread.php?t=1767093

But apart from all those numbers, as far as I understand, it's pretty easy to find out if your lipos are up to the task:
- Check if single cells drop below 3.3 or 3.2v under load ( can be checked with a cheap lipo warner buzzer connected to the balancer leads)
- Check how warm they get. Opinions differ on this, but I'd say if they go over 50 degrees Celsius regularly it's too much for them.
Thanks for that, since I'm from Switzerland german is perfect for me. I always had such a buzzer connected to my lipos.
They never got hot, so if this is true I don't think that I have a problem with my lipos.


My plan now: I think I'll fix my VESCs, maybe add an additional capacitor and maybe the overvoltage diode mentioned by pf26 and get another set of Lipos to connect in parallel.
Then I can test again and even compare Vedders and Onloops VESC to see if there's any difference.

Besides that I thought about adding cooling by just glueing copper heat sinks to the mosfets using heat conductive glue. Any doubts?

Best regards
fasser
Last edited by fasser on Jul 22 2015 4:42pm, edited 1 time in total.

lizard   100 mW

100 mW
Posts: 43
Joined: Oct 29 2014 6:38am
Location: Germany

Re: 10S custom skate ESC: testers wanted!

Post by lizard » Jul 22 2015 10:10am

Oh cool, didn't realize you were from Switzerland :)
Yes, maybe overvoltage is the problem. I had a 2000uF/63V Nichicon capacitor placed in the same way as on the Onloop BETA batch. So approx 2-3cm away from pcb connecting points.
The VESC failed at full throttle in current mode and was still accelerating. It felt a bit like cogging in the last moment, but it's hard to say if this was the reason for the VESC to fail or if this was the result of the failed VESC.
Maybe I should add a second capacitor or a transil 1.5kW56V.
Concerning the transil1.5kW56V: I am right that this is a diode that shorts the circuit if the voltage rises over 56V?
Isn't that "voltage rises too much problem" only occuring during braking?
I mean this one that MGM Compro had observed and analyzed:
http://rc10.fi/index.php?action=dlattac ... ttach=7378

They also found that adding a diode won't help because it's too much energy.
But I am in no way an electrical engineer or something like that, maybe I got it all wrong :)

pf26   100 mW

100 mW
Posts: 39
Joined: Nov 17 2014 1:37pm

Re: 10S custom skate ESC: testers wanted!

Post by pf26 » Jul 22 2015 1:29pm

I think you get the 100Amps from the battery during the ON times of the controller PWM, and around 18 Amps on average; (If the controller was dissipating the extra voltage, that's more than a kilowatt, and it would just melt quikly - I am positive the VESC has above 95% efficiency in most cases). Also, the capacitor should take some of these 100Amps, if not most (it's difficult to calculate, it also depends on the battery internal resistance and wiring inductance).
The little Transil I use is just meant to deal with inductance based transients: The wiring acts as an inductor. Much current means storing energy like force in a spring, and this can be suddenly released causing very sudden and brief overvoltages. I agree the Transil will not help against regen if the battery voltage raises to much.

I have less than 10mm between my capacitor and PCB inputs, and these two thick wires are very near one to another, in order to reduce the inductor (At 100 Amps, even a few nH will cause non negligeable voltage rise). The transil is connected there in parallel, also that's only a few mm from the board.

I measured my low cost 6s 5Ah Lipo at around 10mOhms internal resistance, so these currents should not create that much voltage drops/rise. I doubt it was main issue. But of course it does not help.

You may want to change your controller parameters, because I think it should not be able to raise 150Amp overcurrent limit, even with fast changing setpoint.
If the controller is subject to reach much larger currents than set (charging wire based inductor), and then suddenly cancels current due to the alarm being raised, then the released energy from the inductor may damage something. It is alway better to keep the controller running, thus controlling the FETs (and in synchronous mode, so it is better than freewheeling diodes). To reduce the overvoltage risks I would recommand :
- make sure the VESC is running all the time (adjust parameters so as it never raises alarms, and also allow up to 60 volts operation.
- Short connection thick wires, between low ESR capacitor and PCB, wires very close one to another (smallest possible loop area)
- Short connection between battery and capacitor (no large loop)
- Transil
- low battery internal resistor.

vedder   100 W

100 W
Posts: 249
Joined: Nov 05 2011 9:37pm

Re: 10S custom skate ESC: testers wanted!

Post by vedder » Jul 23 2015 4:02am

fasser wrote:Hi everybody

Thanks for your answers.
sl33py wrote:@fasser - what motor kv? I saw something here about >200kv on 12s could be a problem.
The motor is an EMAX GT5345/09 with 170KV. The error occured at around 9000 ERPM in sensorless mode, so I don't think it's the reported KV limit.
onloop wrote:
The vesc also has temp limiting function.. I think the temp sensor is positioned near fets. What do you have it set at in BLDC
I left the temp limit as default at 80°C start and 100°C end for the current limiting. But as mentioned the temperature was not reported near this limits when the VESC died.


I'm asking myself if maybe the batteries are limiting the current, but it would be strange that it is exactly at 80A. I'm using two Turnigy 5000mAh 6S 20C Lipos in series.
http://hobbyking.com/hobbyking/store/uh ... duct=35828
They should have 150 Amps peak discharge. Does anybody have some experience with these? I think they're quite common.

Best regards
fasser
Hi fasser,

Is it this motor?
http://www.agmhobby.com/item/outrunner- ... p-648.html

You should not set the current limit way higher than in the specification. If you do that, the back-emf gets distorted and tracking the motor won't work properly. I don't recommend more than 80A at most for that motor. With a too high current limit, the protection features will also not work properly and components can fail easily. What happens when you the realltime graph (which shows a filtered version of the current) shows much lower current than you set and you run far from maximum speed is that the current rises above the set limit and the hard current limiting feature kick in all the time, which is very bad condition to operate in (notice that this can also can happen if you have the wrong motor parameters).

If you'd like, you can send me your VESC and I can try to repair it.
Some of my projects: http://vedder.se
Support my open source ESC development with a small donation

lizard   100 mW

100 mW
Posts: 43
Joined: Oct 29 2014 6:38am
Location: Germany

Re: 10S custom skate ESC: testers wanted!

Post by lizard » Jul 26 2015 6:13am

I have run the VESC now for the first time. Startup is very smooth and silent compared to my Mamba Monster. Throttle and brake feeling is also very nice. But unfortunately, it didn't last long :(

Drove around slowly for a minute, max. 20% throttle. After giving a little more throttle (maybe 40%) it just silently died. No popping sounds, no heat, no sparks or whatever.

Now the red led lights up when trying to give throttle. Haven't connected it to a computer yet to see what the software says. Will do that later.

Image


Pedrodemio   100 W

100 W
Posts: 112
Joined: Jun 30 2015 8:38pm
Location: Brazil

Re: 10S custom skate ESC: testers wanted!

Post by Pedrodemio » Jul 26 2015 2:58pm

Hello vedder,

In the near future i plan to get a VESC for my board, but for now i'm working with a standard ESC and a external arduino current control

I wonder if you could explain how you did the battery and phase current limitations, i'm using a current control, but in the low duty cycle i have to limit phase or i will fry my motor, but i can't find a efficient way to implement this

In my current code the max battery current in dependent on the duty cycle ( Ibat = Ibatmax/dutycycle), but for startup i have to ignore the o duty cycle, or the motor will never start, but doing this results in a very un-smooth transition

Thanks

lizard   100 mW

100 mW
Posts: 43
Joined: Oct 29 2014 6:38am
Location: Germany

Re: 10S custom skate ESC: testers wanted!

Post by lizard » Jul 27 2015 2:49am

Connected the VESC to a PC now. Looks like the DRV is toast again.

vedder   100 W

100 W
Posts: 249
Joined: Nov 05 2011 9:37pm

Re: 10S custom skate ESC: testers wanted!

Post by vedder » Jul 29 2015 2:21am

lizard wrote:Connected the VESC to a PC now. Looks like the DRV is toast again.
I'm sorry about that. Maybe I damaged the DRV when soldering it. If you send the VESC again I can fix it again. This time I can put your motor in my RC car and test it myself. I'm sure that it should work well since your motor seemed to behave nicely when I tested it.

Can you send me the settings you used on the VESC as an XML file from BLDC Tool?

Do you have a separate BEC/UBEC or do you use the 5V from the VESC to power the receiver and servo and how did you connect it? If you have a separate BEC, there can be issues with ground loops. On my RC car, I used 5V from the VESC to power everything, but the servo draws quite a lot.
Some of my projects: http://vedder.se
Support my open source ESC development with a small donation

lizard   100 mW

100 mW
Posts: 43
Joined: Oct 29 2014 6:38am
Location: Germany

Re: 10S custom skate ESC: testers wanted!

Post by lizard » Jul 29 2015 6:55am

No worries, thanks for your help Benjamin, I really appreciate it :)

Hmm, why are the DRVs such divas?


I am using an external BEC that is wired like this:
Image
The steering servo is plugged into the receiver also.

Can this cause ground loops? I'm not that electrically savvy, but in R/C car forums and in Castle's manuals it is explicitly stated that the ground wire from the ESC to the receiver is needed because otherwise the ESC may not be able to read the throttle signal correctly and strange things and glitches may happen (If I understand correctly, because it doesn't have a ground reference then). I run it like that in my other cars too (including the twin 1717 car with 2 ESCs) and never had problems.

The settings I used are the ones that it came back with from you. Only settings changed are related to PPM pulse duration. I'll pull the XML file from it when I get home later.


Regarding putting the 1717 in your car: That would be really cool, but I'm a little afraid that would be asking too much from you:
It has M4 mounting threads and is 50mm in diameter, it will probably not fit in an 1/8th scale buggy motor mount, most of them have M3 holes and allow for a maximum diameter of around 45mm. Gearing will also be a problem, with most 1/8th scale buggies you can't gear to 160kph without modifications to the motor mount, you'd need to mount a 30t or bigger pinion with common 1/8 buggy spurs and diffs (the Traxxas Slash is nice in that regard because it can use spurs down to 34t and has a 2.85:1 diff ratio ...).

I have ordered some TPPower motors from TPPower in the netherlands that haven't arrived yet, those I could redirect to you. It's a 40x106mm and a 42x82mm can with M3 threads, that would atleast make the mounting hassle-free. But I don't know how comparable they are, the 106mm TPPower weighs almost exactly the same as the 1717 (580g), but has a long and "thin" rotor compared to the 1717 which has a shorter "fat" rotor. Apart from that I think they are pretty much the same design and look like almost any 1/8th scale inrunner with 4-Poles and 12 Slots.

I also have a TPPower 56x92mm 6-pole sensored inrunner for testing if you want, but I'm not sure if the VESC can do enough ERPM for it, it's 1130KV and was originally intended to be run with 8s.

erwincoumans   100 W

100 W
Posts: 141
Joined: Apr 02 2015 11:25pm

Re: 10S custom skate ESC: testers wanted!

Post by erwincoumans » Jul 29 2015 3:52pm

Does anyone know how can you configure a Linux serial terminal program, such as 'screen' or 'minicom' to communicate with VESC, similar to BLDC_Tool? The BLDC_Tool connects fine to /dev/ttyACM0

I tried running screen /dev/ttyACM0 115200 but there is no feedback, when typing 'help'.

chuttney1   100 W

100 W
Posts: 243
Joined: Apr 03 2013 2:44pm
Location: Los Angeles, CA

Re: 10S custom skate ESC: testers wanted!

Post by chuttney1 » Jul 29 2015 9:28pm

Who here has issues uploading firmware to the VESC. I'm stuck since I arrive an openocd init failed problem.

fasser   100 µW

100 µW
Posts: 7
Joined: Feb 11 2015 12:15pm
Location: Zurich

Re: 10S custom skate ESC: testers wanted!

Post by fasser » Jul 30 2015 3:25am

Hi everybody!

I was quite busy the last week so I had no time to write here.
I hope I don't scramble up the thread since it already continued with other topics.
vedder wrote: Hi fasser,

Is it this motor?
http://www.agmhobby.com/item/outrunner- ... p-648.html

You should not set the current limit way higher than in the specification. If you do that, the back-emf gets distorted and tracking the motor won't work properly. I don't recommend more than 80A at most for that motor. With a too high current limit, the protection features will also not work properly and components can fail easily. What happens when you the realltime graph (which shows a filtered version of the current) shows much lower current than you set and you run far from maximum speed is that the current rises above the set limit and the hard current limiting feature kick in all the time, which is very bad condition to operate in (notice that this can also can happen if you have the wrong motor parameters).

If you'd like, you can send me your VESC and I can try to repair it.
Yes, it's exactly this motor.
OK, thanks for this explanation. I thought as long as the motor does not overheat it's no problem to apply higher currents.

I decided now to have a dual motor setup to prevent this.

Thank's for the offer to try to repair my VESC, I'll check the post costs to sweden and send you a PM about that.

Best regards,
fasser

vedder   100 W

100 W
Posts: 249
Joined: Nov 05 2011 9:37pm

Re: 10S custom skate ESC: testers wanted!

Post by vedder » Jul 30 2015 3:43am

lizard wrote:No worries, thanks for your help Benjamin, I really appreciate it :)

Hmm, why are the DRVs such divas?


I am using an external BEC that is wired like this:
Image
The steering servo is plugged into the receiver also.

Can this cause ground loops? I'm not that electrically savvy, but in R/C car forums and in Castle's manuals it is explicitly stated that the ground wire from the ESC to the receiver is needed because otherwise the ESC may not be able to read the throttle signal correctly and strange things and glitches may happen (If I understand correctly, because it doesn't have a ground reference then). I run it like that in my other cars too (including the twin 1717 car with 2 ESCs) and never had problems.
This is what can cause ground loops. ESCs that are for high power often have optical isolation on the input to avoid this issue when using external BECs. What happens is that when the motor pulls a lot of current, the voltage between the battery connection and the VESC drops a bit because of cable resistance and cable inductance. This voltage difference will cause a current in the signal ground through the minus connection, which can cause issues. You can avoid this problem in some different ways:

* Power ell electronics from the 5V output of the VESC. It can supply a bit over 1A, which maybe is too low for a servo though. I have been using my RC car like this and it works well most of the time even when the servo saver kicks in, but if I block the servo with my hand the VESC will reset because of voltage drop.

* If you use an external BEC, connect the minus wire of the BEC to the point where minus is soldered to he VESC, because this is the point where the signal ground and power ground are connected.

* Control the VESC using CAN-bus. The CAN transceiver has a differential input and can handle some voltage difference even below ground.

* Buy or build an optical insulation device for the servo input. Maybe something like this works:
http://www.ebay.com/itm/WRL-OPTO-Glitch ... 19e3d74902
lizard wrote: The settings I used are the ones that it came back with from you. Only settings changed are related to PPM pulse duration. I'll pull the XML file from it when I get home later.
It is also possible that I messed up the settings when I had the VESC because I was testing some things and had my development version of BLDC Tool. I'll see that when I get the XML.
lizard wrote: Regarding putting the 1717 in your car: That would be really cool, but I'm a little afraid that would be asking too much from you:
It has M4 mounting threads and is 50mm in diameter, it will probably not fit in an 1/8th scale buggy motor mount, most of them have M3 holes and allow for a maximum diameter of around 45mm. Gearing will also be a problem, with most 1/8th scale buggies you can't gear to 160kph without modifications to the motor mount, you'd need to mount a 30t or bigger pinion with common 1/8 buggy spurs and diffs (the Traxxas Slash is nice in that regard because it can use spurs down to 34t and has a 2.85:1 diff ratio ...).
I noticed that it would not fit easily which is why I didn't test it the last time. My 1:8 car was originally running on gas though, so I have modified it to electric and it has a non-standard motor mount.

It is also interesting for me to make sure that the VESC works well with this kind of motor, so I would be willing to spend a bit of time on it. What I can do is to make a new motor mount for my RC car in a CNC mill that I have access to so that the 1717 motor fits on my RC car. I need a bigger pinion than the one I have for that since the motor would come too close otherwise, so if you have some pinions and want to send the things to me again I can give it a try. I don't think I can run the motor on full power though since my RC car will probably end up broken in tiny pieces then.
lizard wrote: I also have a TPPower 56x92mm 6-pole sensored inrunner for testing if you want, but I'm not sure if the VESC can do enough ERPM for it, it's 1130KV and was originally intended to be run with 8s.
6-pole inrunners are usually quite difficult, but I think this one should work fine with up to 6s. I was planning to experiment with the switching frequency to handle higher ERPM, and this motor would be perfect for testing that. I will also implement a check that detects if the motor kv and voltage are so high that tracking the motor becomes an issue and make a fault code for that. Some people who have tested the VESC fried it by connecting motors with high KV to too high voltage, so detecting that before it is too late would be great.
Some of my projects: http://vedder.se
Support my open source ESC development with a small donation

vedder   100 W

100 W
Posts: 249
Joined: Nov 05 2011 9:37pm

Re: 10S custom skate ESC: testers wanted!

Post by vedder » Jul 30 2015 3:50am

erwincoumans wrote:Does anyone know how can you configure a Linux serial terminal program, such as 'screen' or 'minicom' to communicate with VESC, similar to BLDC_Tool? The BLDC_Tool connects fine to /dev/ttyACM0

I tried running screen /dev/ttyACM0 115200 but there is no feedback, when typing 'help'.
A serial terminal won't work since the communication is raw data and split into packets. Each packet also has a checksum that you need to calculate for the VESC to accept it, so if you just type commands to it, it won't reply. Is there a reason why you would like a serial terminal to connect to it? Would you like to use it without an X server?
Some of my projects: http://vedder.se
Support my open source ESC development with a small donation

vedder   100 W

100 W
Posts: 249
Joined: Nov 05 2011 9:37pm

Re: 10S custom skate ESC: testers wanted!

Post by vedder » Jul 30 2015 3:55am

chuttney1 wrote:Who here has issues uploading firmware to the VESC. I'm stuck since I arrive an openocd init failed problem.
I'm writing a reply where you posted the question on my blog now.
Some of my projects: http://vedder.se
Support my open source ESC development with a small donation

jamesonotc   10 W

10 W
Posts: 70
Joined: Jun 20 2015 12:49am

Re: 10S custom skate ESC: testers wanted!

Post by jamesonotc » Jul 30 2015 2:04pm

posted in wrong thread

*deleted*
Last edited by jamesonotc on Jul 30 2015 4:41pm, edited 1 time in total.

Post Reply