Have you ever burned a Castle Creations ICE 160 HV ?

Beuhji

Beuhji

100 mW
Joined
Dec 6, 2009
Messages
38
Location
Bordeaux - FRANCE
Hi guys !

During testing of my e-KMX, my Castle Creations ICE 160HV exploded ...

My motor is a Turnigy 80-100-B (with 7,2 reduction) should not create difficulties for this controller (I have never reached more than 120 A).

Do you think that using this controller designed for aircrafts in an electric vehicle can create such damages ?

I had already traveled for 30 minutes and the temperature of controller was reasonable (by touching).

One of the 3 motor wire from the controller was detached, do you think a bad weld may be causing the problem (two wire for few seconds) ?

Do you think that warranty can apply ?

Thanks all, soon, the post to show you my e-KMX

p1030098t.jpg


p1030100.jpg
 
I have been running Castle controllers in bikes for years now. The HV160 is the best yet. YEs they work in E-bikes.

Here is my analysis of your situation;

Your controller blew up from HUGE current ripple. I know this because of the exploded capacitors. Caps typically explode from over-voltage. That happens from current ripple.

Here is what you need to do;

#1 Make the wires between the controller and the battery as short as you possibly can. I cannot stress this enough.

#2 Add at least four 100mf low ESR caps to the input leads of the new controller as close to the controller as you possibly can. I try to solder them within an inch of the controller.

#3 set the throttle response to #1 in the programming of the controller. That will soften the throttle hit.

#4 Make sure the controller is set for "Outrunner" mode.

Matt
 
Yes, add capacitors as close to the ESC as possible. If you're running with 12S Lipo, then 7 volts of ripple at 50.4 V will easily put you past the "53 - 54 volts" absolute voltage ratings. What kind of battery were you running with? Chemistry and series count would be helpful to know.

Hmmm... that's also interesting that the motor wire was detached. What kind of system are you running it in? (What kind of bike and with what size wheel?) It sounds like the 7.2:1 reduction may not have been high enough causing huge phase currents with the motor which may have also exacerbated the current draw on the batteries (And thus the stock caps). I'm imagining a reduction of 15:1 would be more appropriate on a 26 inch bike (Or maybe 12.5:1 on a 20 inch), since that's the equivalent ratio on my electric scooter with 8 inch wheels.
 
swbluto said:
Yes, add capacitors as close to the ESC as possible. If you're running with 12S Lipo, then 7 volts of ripple at 50.4 V will easily put you past the "53 - 54 volts" absolute voltage ratings. What kind of battery were you running with? Chemistry and series count would be helpful to know.

Hmmm... that's also interesting that the motor wire was detached. What kind of system are you running it in? (What kind of bike and with what size wheel?) It sounds like the 7.2:1 reduction may not have been high enough causing huge phase currents with the motor which may have also exasperated the current draw on the batteries (And thus the stock caps). I'm imagining a reduction of 15:1 would be more appropriate on a 26 inch bike (Or maybe 12.5:1 on a 20 inch), since that's the equivalent ratio on my electric scooter with 8 inch wheels.
Click to expand...

I absolutely agree.

I have had a motor wire disconnect. The controller tends to short one battery lead to one motor lead thus blowing that wire off the board (at least on the 160).

Matt
 
Thanks for your advises ! Unfortunately data logged by the ESC are lost ...

More informations about my project :

=> KMX Tornado
=> 12S4P LiPo 20Ah 20C (from Turnigy)
=> 20" rear wheel
=> 1/7.2 reduction
=> max input speed of 6000 rpm

Do you ever try to send a burned ESC to Castle Creations for warranty ?

Matt, do you want to say 100mF or 100µF ??? :shock: Have you an example of reference for a low ESR capacitor ?
 
Yes, Castle will repair [replace] it for you. I think it is a flat $150.

I use part number 9D29MF Panasonic caps.

Lastly, I would gear the trike for 45mph max with one motor and controller and you should always pedal first. It is possible to start without pedalling. But, it is much easier on the controller if you give a couple pedal strokes before hammering it. :D

Matt
 
I just checked the simulation and it appears it's geared for around 50-55 mph which isn't as bad as I thought it was. Lower is better, but it seems that's likely not the problem.

Adding capacitors would help.

You could also, probably, reduce the series count to 10s or 11s. May not be practical since you already have the batteries, but it seems far less likely you'd exceed the capacitors' voltage rating given the ripple wasn't too bad (The highest seen so far seems to have been 7 volts?).
 
When I blew my 110 controler one of the things mentioned here was I was using too much low throttle. I was just just using %20 over a long period of time and its better to use high thottle, something about the controler has to work to hard at low power.
 
I was looking at a data capture done by AussieJester for his bike.

file.php


It seems that he has ripple BUT it doesn't seem like the voltage ever goes up, it only goes down from the "base voltage". How does the cap explode by over-voltage if this is the case? Is there an overshoot on the capacitor's voltage during the rise back to the "base voltage" that isn't recorded by the data logger?
 
The voltage level and the ripple level are two different things. Ripple is a "Bouce-Back" affect of pulling milisecond high current pulses through a length of wire. That bouce back causes a voltage rise (again, in milisecond bits) between milisecond amperage pulls. Those voltage rises will not be caught by the voltage level indication, but only by the ripple reading. Actually, you will see an overall voltage drop on the average because the voltage level looks for an average voltage over a number of samples and the voltage drop caused by cell resistance overshadows the very short duration ripple (voltage increases). The ripple software in the 160 does not average anything out, it simply takes extremely quick voltage samples and displays those as individual ripple spikes. When the ripple indication appears in the data logger, it is listed not by what the actual voltage is, but by what the increase in voltage is over the pack voltage itself and displays that as a ripple number. So, a 2 volt ripple on a 48 volt pack means an even 50 volts is being seem by the FETs in tiny doses.

Matt
 
Since swbluto posted this data log with the high ripple i would just like to add that this was before
i added the extra cap on the input lines, that data was recoreded with the 3x 1000uf caps mounted on pcb and directly soldered to the esc board.


This below graph is the result of fitting the caps as Matt.S does, i.e ~1inch from the board on the input wires-->



Just so people know ...

I can't give this hv160 a big enough rap, puting up with the hammering i give it...yes i blew it but due to user error i wasn't even
on the bike it was sitting stationary ...

KiM
 
Also, perhaps most importantly, the PWM rate must be set at 24kHz. That will essentially cut your voltage ripple in half compared to the factory setting of 12kHz.

Like Matt stated, you should definitely fit 4 of the 1000 microfarad caps.

If possible, I'd try to get it geared down to about 45 mph at full motor rpm.

I would also recommend charging to a max voltage of 50.0V, that gives you a little headroom for the stressful inrush of current when arming the controller.


I'm running a single 6-turn Delta 3220 geared for 45 mph and I've blown a couple HV110's, a couple HV140's, and one HV160 after 30 miles, and another after 15 miles, and these are the things I've learned. I've now been running mine for over 300 miles, and this is how I have it setup. My ripple never goes much beyond 3V.

Here's my ripple data for comparison's sake:
TrikeData_6_10_10.jpg
 
Hello, I have an ICE HV160, brand new, I believe castle will still trade them for the ICE2 HV160. My question is do I want to trade it in or just use it? It seems I read in one thread that the older one is better. Thanks!
 
Just for a reference, I recall a link to an RC discussion group where the operator had a model blimp that was used for advertising and was used during a group meet. The owner, battery, and controller were on the ground walking (to save weight on the blimp). Previously the owner had the controller next to the motor, so he would only have two wires ascending to the blimp, and he was blowing controllers.

After he added input capacitors, and mounted the controller on the battery pack, everything worked fine. I don't recall the length of the three phase wires that went up to the motor, but they were quite long, much longer than anyone would have on an E-bike. Do not fear long phase wires, but keep the ESC next to the battery.

Here are capacitor purchase links from my files, $2 each:
http://endless-sphere.com/forums/viewtopic.php?f=30&t=22194&start=45#p382837

similar threads with pics which are helpful:
http://endless-sphere.com/forums/viewtopic.php?f=28&t=35420
 
Have to keep an eye on this topic!

Blew an 160HV once, after 5 minutes, with cheapo capacitors from China. Got an 160HV2, bought some good capacitors (the one recommended in another topic) and it ran good!
 
VRdublove said:
Also, perhaps most importantly, the PWM rate must be set at 24kHz. That will essentially cut your voltage ripple in half compared to the factory setting of 12kHz.
Click to expand...
What an amazing insight, I never would have thought of - or noticed that. I'm going to set this up on my HV-160s!
 
MattyCiii said:
VRdublove said:
Also, perhaps most importantly, the PWM rate must be set at 24kHz. That will essentially cut your voltage ripple in half compared to the factory setting of 12kHz.
Click to expand...
What an amazing insight, I never would have thought of - or noticed that. I'm going to set this up on my HV-160s!
Click to expand...

You can get away with this on a low pole count motor like the astro...on a high pole outrunner the switching losses will have you in flames in a suprizingly short time. Use the outrunner settings in the castle link for outrunners! :p

I know Recumpence blew a few controllers in his early days(outrunners)....but I am curious, how many blown an HV160 with the astro motor?

(has to be a small % compaired to the bigger outrunners)
 
I've read that re-terminating a motor from Delta to Wye will reduce inductance, I seem to remember something about that helping controllers quite a bit. i don't understand it, but it may be worth reading up on to ensure your ESC has the best possible chance to perform well without coming close to frying.
 
Thud said:
MattyCiii said:
VRdublove said:
Also, perhaps most importantly, the PWM rate must be set at 24kHz. That will essentially cut your voltage ripple in half compared to the factory setting of 12kHz.
Click to expand...
What an amazing insight, I never would have thought of - or noticed that. I'm going to set this up on my HV-160s!
Click to expand...

You can get away with this on a low pole count motor like the astro...on a high pole outrunner the switching losses will have you in flames in a suprizingly short time. Use the outrunner settings in the castle link for outrunners! :p
Click to expand...
Thanks for that clarification Thud... In my case I am using the Astro inrunner, but it's a great caution... I was ready to rush off and implement a best practice without carefully considering what was similar and different between my system and someone else's.
 
Wow quick replies! Thanks guys! I think I'll run it then. I guess I better get some caps. I am planning to use the cheap non hk 80/85 motor 170-180 kv motor. I thought it was 170 kv, but the motor sticker says 180, whatever! I still need to check the windings, the magnets, and the bearings. :roll: I am planning to run it geared 8.18:1 to a mid mounted Nuvinci then 2.5:1 to the rear 26" wheel. So a total reduction in low gear of 40.9:1, and high gear of 11.68:1. I figure 9000 rpm's unloaded (12s) and about 7200 loaded, so 13 mph in low and 47 in high.

How does the 80/85 180kv compare to the 80/100 130kv both on 12s?

The 130 kv would be a lot simpler. I would do the same 8.18:1, but the Nuvinci would only spin 635 instead of 880, making gearing up the pedals way easier, plus in high it would only spin 1050 instead of 1500, and I could drop the ratio to the rear wheel to less than 2:1 which is easier.

I will build a custom full suspension mountain bike frame, I want to go up the ski runs!
 
You must log in or register to reply here.

Similar threads

B
    • Like
Intro guide to TSDZ2 & BBS02 (updated 2021)
2
Replies
35
Views
23,562
docw009
docw009
rberger
My Experience Converting an ICE Trike to a DD Hub Pedalec
Replies
5
Views
3,340
chvidgov.bc.ca
C
ElectricGod
My kick scooter project
15 16 17 18 19
Replies
462
Views
86,175
ElectricGod
ElectricGod
Beuhji
Castle Creation RECALL
Replies
1
Views
753
AussieJester
AussieJester
B
DUAL E-MTB BUILD
2 3
Replies
68
Views
12,303
torqueboards
T
Back
Top