Efficient solar powered (electric) kayak

Discussion in 'Boat Design' started by ElectricKayak, Mar 20, 2015.

  1. ElectricKayak
    Joined: Mar 2015
    Posts: 78
    Likes: 0, Points: 0, Legacy Rep: 10
    Location: Canada

    ElectricKayak Junior Member

    Yes, electric drive at short burst is definitely faster than we can paddle and better control too when the current is running near our maximum effort. Previously it could get a little exciting trying to power through a narrow spot, now I just twist the knob a bit more and relax while we cruise up. I run to about 200 watts, beyond that the power goes up dramatically but the speed doesn't (on the inflatable). Doubling the power might only get me another 1kph and by then the shaft is beginning to buckle which dramatically shortens life.

    I think the shaft will almost certainly fail at my coupler since there is a significant stress concentration there. At 30-50 watts I wasn't too concerned but expectations rise to meet available power (!) so a redesign that more properly deals with the stress raisers is in order.
     
  2. portacruise
    Joined: Jun 2009
    Posts: 1,115
    Likes: 32, Points: 48, Legacy Rep: 218
    Location: USA

    portacruise Senior Member

    Thanks, EK
    Comments and questions below. Hope this helps.

    PC

     
  3. ElectricKayak
    Joined: Mar 2015
    Posts: 78
    Likes: 0, Points: 0, Legacy Rep: 10
    Location: Canada

    ElectricKayak Junior Member

    I'm using a 20 amp fuse. I think the shaft is strong enough provided it is operated well below the buckling loads. I could be proven wrong though, we'll see.

    (EDIT: I've stalled the motor on a weed ball, it didn't blow the fuse though. I'm not really sure the fuse is protecting much. I suspect it'll be my $50 ESC that will blow to protect the fuse (one reason I over spec'd the controller at 120amps). I've also left the motor stalled accidently (prop stuck in the mud) while eating lunch. Luckily it was on low power from cruising into shore so survived.)

    My main concern with the current shaft/coupler is fatigue failure due to the cyclic bending which is a characteristic of the flex shaft approach. A million cycles doesn't go very far (100-200km) so the design really needs to be based on infinite life criteria for a reasonable shaft life (this includes no significant stress raisers). So I'll sheath the shaft and change the coupler accordingly and see how it goes.
     
  4. BertKu
    Joined: May 2009
    Posts: 2,480
    Likes: 43, Points: 58, Legacy Rep: 223
    Location: South Africa Little Brak River

    BertKu Senior Member

    My recommendation is to use a semiconductor fuse. They are normally available in electronic components shops. The speed of those fuses are special to protect electronic equipment, while a normal fuse only blow fast, when the overload is 10 times or more.

    I have in my own made controller given special attention to this problem and check every 1/1000th a second the current and should it be more than (my case 60 Ampere), the controller disconnect the motor before any great damage can be done. Maybe you should make your own electronic fuse whereby you can set the maximum current rating. Bert
     
  5. ElectricKayak
    Joined: Mar 2015
    Posts: 78
    Likes: 0, Points: 0, Legacy Rep: 10
    Location: Canada

    ElectricKayak Junior Member

    Thanks for the suggestion. That would be an easy improvement but don't these produce a high arc voltage when tripped? I'd be concerned they would damage the ESC. Otherwise they do seem better, maybe 5x overload versus 10-20x of a regular fuse.

    Despite my flippant comment in the earlier post, my ESC has built in temperature and current limiting so should be safe. It will put an upper limit on the motor overload until the fuse blows. At these power levels I'm thinking it is good enough but time will tell...so far so good.
     
  6. BertKu
    Joined: May 2009
    Posts: 2,480
    Likes: 43, Points: 58, Legacy Rep: 223
    Location: South Africa Little Brak River

    BertKu Senior Member

    If a motor stalls, I don't know whether your ESC has a time delay build in, or tries and tries to start the motor when the motor has stalled. i.e. your motor will still get hot and hotter and hotter. In my case in my program as soon I have exceeded the current by 1 Ampere, I say, there is a problem do no longer switch the PWM (Pulse width modulators on any more) and I psychically must switch the system off and on again to clear the fault. Should the fault persist, before the current can do any damage, the motor (PWM) does no longer get switched on. Therefore I cannot damage anything even if the motor is stalling. I do not know what your ESC is doing when an over current is sensed. However you don't have to be worried about spikes. The motor has such a low inductance (100 pieces of thin copper wire connected parallel per coil x normally x 3 coils ) that there is no difference in switching the coil on and off by the very fast PWM or by switching the system off with a fast semiconductor fuse. All MOSFets have zenerdiodes build in, to cope with those small spikes. You could thus with a MOSfet IRFP064N and a small microprocessor make your own very fast fuse, which can be reset and used over and over again. The small microprocessor senses any current exceeding the limit you have set and it switches the MOSfet off. The MOSfet can handle up to 100 Ampere continuous ( 100 x 8 milliOhm = 0.8 volt loss only), but you will set the limit to 20 Ampere in anyway , thus a loss of 0,16 Volt. If you are interested, even just for the sake of interest, I could make a new thread in http://www.boatdesign.net/forums/onboard-electronics-controls/ and show you how to do it. Bert
     
  7. ElectricKayak
    Joined: Mar 2015
    Posts: 78
    Likes: 0, Points: 0, Legacy Rep: 10
    Location: Canada

    ElectricKayak Junior Member

    Perhaps but I think then the fuse will blow. The fuse I2t is much lower than the motor limit.


    Yes, I like this approach. Much better than a fuse.

    Although my ideal solution might be based on motor rpm. If the motor rpm drops by say 10% then shut down.


    Are you sure? With semiconductor fuses my understanding is the peak arc voltage is proportional to the system voltage and set by the design of the fuse (by controlling the arc stack, higher arc voltage = faster current limiting). I've attached the arc voltage graph for a A15QS20 amp fuse showing that with a 12 volt system the peak arc voltage is over 100 volts.



    So about 3.2 watts loss at 20 amp rated current in the MOSfet. Not bad.

    What is the current draw for the rest of the system (Arduino etc)?


    Yes, this would be very interesting. I'm sure others would have an interest also. Please do and thanks!
     

    Attached Files:

  8. BertKu
    Joined: May 2009
    Posts: 2,480
    Likes: 43, Points: 58, Legacy Rep: 223
    Location: South Africa Little Brak River

    BertKu Senior Member

    I have decided to make one in anyway for my own use. I estimate about 40 milliAmpere in total ,when the buzzer is on to warn that the current has been exceeded. i.e. at 12 Volt = 0,5 watt.

    Yes, the power dissipation the IRPF064N can withstand/have is 200 watt per MOSfet, thus 20 Ampere x 20 Ampere x 8 milliOm = 3.2 watt. Cooling is not an issue and only a small piece of cooling plate will do the job.


    I will do. The graph confuses me. A fuse which cut an large inductance will arc and as a result slowly the melting wire will pack up. Your motor has very low inductance, when I measure my motor spikes, at 24 and 36 Volt I can hardly see any spikes because of to the PWM which we both have in our ESC's . However, we switching the plus 12 Volt (24 Volt) off when 1 bit (19 milliVolt) exceeds the set current limit and the Capacitors in the ESC will smooth it out. (maybe only 0,5 Volt because of the low inductance and not up to 100 Volt. I will start a thread in the control section. bert
     
  9. portacruise
    Joined: Jun 2009
    Posts: 1,115
    Likes: 32, Points: 48, Legacy Rep: 218
    Location: USA

    portacruise Senior Member

    Hi, Bert, I also would be interested to see your post on motor protection.

    When my system stalls, it is at the maximum ratings of the motor, but that is different from EK's situation- which stalls at maybe 10% of maximum motor ratings. So I expect that means much less heat buildup in EK's setup.

    I also like your manual reset, circuit breaker analogue type approach, compared to the slower fumbling with fuses. Fuses are not great, if you need to get back on line quickly.

    PC



     
  10. BertKu
    Joined: May 2009
    Posts: 2,480
    Likes: 43, Points: 58, Legacy Rep: 223
    Location: South Africa Little Brak River

    BertKu Senior Member

    Hi Porta, I had a sleepless night and worked the various methods and software out. Having a bad night sometimes happens when my mind is too busy. Today I will start putting it onto the control thread.
    Bert
     

  11. BertKu
    Joined: May 2009
    Posts: 2,480
    Likes: 43, Points: 58, Legacy Rep: 223
    Location: South Africa Little Brak River

    BertKu Senior Member

    Porta, I have posted the thread in OnBoard Electronics & Controls. Bert
     
Loading...
Similar Threads
  1. highamperage
    Replies:
    18
    Views:
    5,010
  2. Annode
    Replies:
    106
    Views:
    3,005
  3. 07MAM
    Replies:
    18
    Views:
    2,735
  4. johnnythefish
    Replies:
    34
    Views:
    5,290
  5. xichyu
    Replies:
    22
    Views:
    2,614
  6. xichyu
    Replies:
    7
    Views:
    1,616
  7. Wavewacker
    Replies:
    58
    Views:
    7,667
  8. SailorDon
    Replies:
    17
    Views:
    6,677
  9. johnnythefish
    Replies:
    24
    Views:
    8,569
  10. JosephT
    Replies:
    37
    Views:
    4,524
Forum posts represent the experience, opinion, and view of individual users. Boat Design Net does not necessarily endorse nor share the view of each individual post.
When making potentially dangerous or financial decisions, always employ and consult appropriate professionals. Your circumstances or experience may be different.