Castle HV Controllers - Use Extra Caps!

My experience has been similar in that lugging the controller quickly increases the esc temp. Were you measuring temp of the esc when it blew? The highest temp my esc has reached was 164 deg F. This was when I did a very slow grind up a very steep 120 ft driveway. It went from about 104 to 164 deg F in maybe 10 sec or so.

Bubba
 
I have a temp sensor on my controller and it's protected me twice already. I can almost guarantee that temperature was your problem, and the reason why was it was probably at such a high temp is that low-RPM = high motor current and the heat generated by the esc is determined by the motor current. Might want temp protection next time or consider better ESC cooling.

(At too low RPM on a current-hungry enough of a motor, though, you might blow the ESC no matter how effective the cooling is. But, that hasn't been determined yet by our experiences.)

Another possible issue is motor synchronization issues and subsequent motor current surges with low RPM, but I haven't seen any problems like that with my motor and castle creations controller.
 
Sounds to me like the PWM rate is too low. High pole count motors need higher PWM rates. For instance, Bob Boucher (from AstroFlight...) told me his 32-series motors, which are 8-poled with 24 slots, need a PWM rate of 20kHz+, or they have startup sync problems. this generates a lot of heat, which sounds similar to what you saw.

-- Gary
 
Gary,

Actually the opposite is true. The higher the pole count, the lower the PWM setting needs to be. My Pletti is a perfect example. It is a 20 pole motor and will not run at higher than 8khz. Also, the inrunner versus outrunner is another huge issue. Bob said outrunners ALL outrunners need to be run at low PWM. He said much of that has to do with higher winding count as well as other factors. His motors are best run right at 20khz.

Matt
 
Not a bad idea. :D

He did say the 16khz setting Gary is using is very close to ideal. The 24khz works, but generates more motor heat than needed, so the efficiency drops.

Bob also said the reason outrunners need lower PWM is related to the high inductance created by the higher winding count. He also said thinner laminations make for higher PWM handling (outrunners still need low PWM even with thin laminations). Astro uses thin laminations. This reduces saturation and inductance (according to Bob).

At any rate, I think 16khz should be fine.

Matt
 
You are right, I forgot the big Pletti is an outrunner. The old Pletti's I used to use all were inrunners.

All we really need, to dial in the best PWM rate is a "Custom" option, so we can enter in a frequency.
 
GGoodrum said:
You are right, I forgot the big Pletti is an outrunner. The old Pletti's I used to use all were inrunners...
Click to expand...
In- or outrunner does not matter here. PWM rate depends on motors inductance. The lower the inductance, the higher the PWM rate. This will keep current ripple within bounds (ripples contain lots of power). Only ironless/slotless (e.g. Kontronik Tango) motors need high PWM, they have low inductance.
However, I high PWM rate on a high inductance motor is no problem. Controller switching losses are accordingly higher, that's all.

Disclaimer:
The above goes for plane/propellor use i.e. low torque at low rpm, contrary to bikes. No knowledge of or experience with e-bikes (yet! :))

Prettig weekend ;) Ron
 
Thanks for the info Ron.

I know my Terminator made a heck of a racket with anything higher than 8khz pwm and got super hot without any load. However, at 8khz, it runs fine. :D

Matt
 
thanks for all the input guys, each step helps us get it right, i'll run 16khz and hopefully with my esc having it's own fan i should be set, looking forward to posting my findings soon to help too.

D
 
Popped another cap, again I was sitting idle when it happened. Right after blasting through downtown and stopping for a moment. I talked to Patrick for a bit about it, and he said it was most likely the startup stresses affecting the caps. His suggestion was to double the capacitance. I asked about using 100v caps and he stated that it wouldn't make a difference.

Just swapped out more caps to 330uf 100v ones. We shall see if the remaining 3 original caps hold up.
 
johnrobholmes said:
Popped another cap, again I was sitting idle when it happened. Right after blasting through downtown and stopping for a moment. I talked to Patrick for a bit about it, and he said it was most likely the startup stresses affecting the caps. His suggestion was to double the capacitance. I asked about using 100v caps and he stated that it wouldn't make a difference.

Just swapped out more caps to 330uf 100v ones. We shall see if the remaining 3 original caps hold up.
Click to expand...

just out of interest what voltage caps are you using and what voltage are you using for your bike ?.
 
i haven't done anything with mine Y ? i thought richards board negated the need?


D
 
Im using 12s lipo. I bought some 330uf 100v caps from digikey a while back, some low ESR ones listed on this site.


Took 200 miles between blown caps.
 
How many did you use? Are you replacing the four stock ones, or did you add some? I'm using Richard's throttle board on my setups, and it has four of the same 330uF low ESR caps you have, I think. This is in addition to the existing ones, but they are at the end of the HV110's main battery leads.

This HV140 I just got from you is quite the beast, with double sets of 10-gauge wires. Are you breaking the soldered ends and putting separate 6.5mm bullet connectors on them?

-- Gary
 
I haven't used the 140's yet. I am doing well with my 110 and don't want to steal more from inventory than I absolutely have to.

I would probably cram the ends into one 6.5mm plug, or use two plugs per power wire.


There are 6 fets stock on the HV110. I have replaced three with the 330uf ones and added another on top. So that is 1860uf capacitance vs the stock 1080. Not quite the double that Pat recommended, but close enough.
 
A graph might be nice.

Depending on the sampling speed of your measurement device, you might be getting aliasing of the PWM frequency. In other words, the oscillation you're seeing may be a measurement artifact.
 
fechter said:
A graph might be nice.

Depending on the sampling speed of your measurement device, you might be getting aliasing of the PWM frequency. In other words, the oscillation you're seeing may be a measurement artifact.
Click to expand...

That's probably it. I should put a low pass filter on my multimeter.
 
Popped another stock cap. Extra capacity isn't enough, the stock caps can't take the use IMO. Rich (one of castle's engineers) stated that the ripple voltage was likely exceeding the cap's rating.

The next firmware update will also allow better control of high inductance motors. It has made a WORLD of difference on my Puma/ BMC v1 motor. Maybe I should shorten those wires too....
 
johnrobholmes said:
Popped another stock cap. Extra capacity isn't enough, the stock caps can't take the use IMO. Rich (one of castle's engineers) stated that the ripple voltage was likely exceeding the cap's rating.

The next firmware update will also allow better control of high inductance motors. It has made a WORLD of difference on my Puma/ BMC v1 motor. Maybe I should shorten those wires too....
Click to expand...

What kind of voltages are you running at?
 
You must log in or register to reply here.

Similar threads

mxlemming
    • Like
    • Brainy
    • Spicy
  • Sticky
VESC Tuning tips
Replies
17
Views
9,833
chuyskywalker
chuyskywalker
dozentrio
Mid-Drive Folding Ebike Build
Replies
3
Views
1,802
Chambers
Chambers
M
Gazelle Heavy Duty NL BBSHD
Replies
7
Views
4,587
Chalo
Chalo
A
R/C EV 'groundskeeper' project
Replies
6
Views
5,170
sleep4life
S
methods
Azure Dynamics Transport Connect - I have one
Replies
17
Views
8,405
methods
methods
Back
Top