Jump to content


Full Members
  • Content count

  • Joined

  • Last visited

Community Reputation

72 Excellent

About Seba

  • Rank

Profile Information

  • Location
    near Gdansk, Poland

Recent Profile Visitors

61 profile views
  1. It's because windings have much less contribution to BLDC motor noise. Most contributing factors are Maxwell forces and magnetostriction - both cause stator core (and consequently the whole motor) vibrations.
  2. I'm also a happy owner of KS18L. I've already clocked over 500 km without any issue.
  3. There is quite easy & effective method for detecting cold joints, and this method can be used by anyone. You just need to buy special freezer spray - for example https://uk.rs-online.com/web/p/freezer-sprays/0846682/ - it can rapidly lower temperature of electronic parts on controller PCB board. Just remove cover, power up the wheel and apply spray on different parts of PCB. There are lot of examples of use on YT, here is one of them:
  4. This high-pitched sound is linked to pulsed magnetic field created by motor windings. They cause small vibrations that are heard as a sound (due to Maxwell forces, Lorentz forces and magnetostriction). It's common with PWM control - many electric motors with PWM inverters also make that sound. This whining may be limited by using sinusoidal current instead of pulsed, but this will require significant changes to control boards and such EUC will cost more. From my point of view this sound is good, because it causes attention to nearby pedestrians
  5. This kind of hardware issue may have very serious consequences. Previously, there was a consensus that this issue is related only to trolleying, so it's safe to ride affected wheel. Now it's clear to me, that this kind of failure (cold solder joint on crystal pad) may occur at any time, also during high speed riding. This cold solder joint may cause intermittent clock failure, leading to complete microcontroller freeze - and this is why PWM stops with certain MOSFETs open, causing constant high current flowing thru the motor winding. And this kind of hardware failure will degrade further - temperature changes and vibrations will evetually cause permanent connection break. So it's a kind of bomb ticking... @tinawong, please forward following idea to your R&D team members: Please consider employing means for quick crystal oscillator failure detection and failover. Some MCUs have special hardware means for clock failover (for example CSS in some STM32 MCUs). When crystal oscillator will fail, clock will switch to another working oscillator preventing system lock-up. If such means are not available, it may be a hardware watchdog with relatively short timeout. It may be cleared within dedicated ISR, to be sure that WDT will be reset within timeout period. Resetting WDT in ISR is normally a bad idea, but in this particular situation clearing WDT in ISR will be beneficial. When crystal oscillator will fail, watchdog will reset MCU which on subsequent boot should switch to internal RC oscillator (HSI in STM32, for example) and enter "failsafe mode" (omit startup checks, BT etc. and go for basic wheel function - maintain balance). If all done quick, it should be enough to regain and maintain wheel/rider stability, preventing faceplant what is most important in this situation. Next, wheel may force rider to safely slow down (using tiltback) and stop or continue riding with limited, safe speed (or just to allow trolleying). I don't know what MCU is used in KS18L controller board, I don't even know detailed hardware architecture of this board. However, it's very likely that at least hardware watchdog and internal RC oscillator is available on MCU used. Watchdog timeout and boot up time should be short enough to minimize motor lockup time, because its braking moment will work against inertia of wheel/rider. If MCU will be able to recover quick, chances are that it will be possible to prevent faceplant or just minimize its consequences. It may be also worth to consider permanently using calibrated internal RC oscillator, if exists on MCU. Higher jitter and lower frequency stability should not be an issue with motor-control application and even with UART timing (BT module has its own crystal oscillator).
  6. 2.4 bar is rather low pressure, don't go lower. Wobbling may especially occur with new tire, because of recessed center part of the tire. After about 400-500 km tire will wear enough so this recessed element will be level with the rest of the tire, giving much more contact and stability. Of course wobbling may be contributed to tired legs.
  7. Seba

    18L orders

    I've "accomplished" faceplant at about 35 km/h at the beginning of my EUC career... Now I enjoy riding at moderate speeds, because I prefer to wear only most important gear - light helmet and soft wristguards.
  8. Seba

    18L orders

    I'm 80 kg netto, but always riding with my 5+ kg backpack And I get consistently more than 60 km on a single charge (at least from my three rides, measured with GPS because Kingsong app is unreliably optimistic in this matter). My average speed is ~20km/h.
  9. Seba

    18L orders

    @Unventor, I'm happy that you're happy :) I'm sure you will love your KS-18L. I've clocked about 200 km on this wheel and as I was very happy after my first ride on this wheel, I'm getting more happy with each ride. Of course I'm aware of trolley mode problem and waiting for confirmed information that the problem cause has been identified and resolved. However, with lift sensor disabled and by avoiding trolleying as much as possible, I can enjoy my rides without any problem or glitch. From my point of view, choosing KS-18L was a good decision. Now I have KS-16C for local, city riding (up to 40km) and KS-18L for longer (60+ km) rides.
  10. Seba

    Another Example of Slime Saving the Day

    You can get original Kenda inner tube for your KS-16S (with angled valve) for about 14 EUR from here - https://eunicycles.eu/en/accessories/11-detka-16-cali.html - but I don't know what shipping cost will be. It may be a good idea to check at your French dealer.
  11. Seba

    Another Example of Slime Saving the Day

    This may be worth to consider - https://www.decathlon.co.uk/universal-puncture-repair-can-id_1278315.html - I always keep one can in my backpack, for emergency 🙂
  12. There are two things to distinct - how many amps battery would get from a charger and how many amps charger can deliver. For example, if battery is near to full charge it wil draw less than one amp, even from 100A capable charger. 175 watts at 84 volts gives us about 2,08 amps. So it may be just a 2A charger, but advertised as 4A. It happens, I've seen a 50 000 mAh powerbank equipped with just one 4 000 Ah cell inside 🙂
  13. Seba

    KS18L Range Test - 200 lbs rider

    It's normal that battery discharge rate is faster when battery is low on charge. It is because battery voltage is going lower as the battery discharge, however power demand is still at the same level. To remind - power is a current multiplied by voltage. To keep motor delivering the same power, current drawn from battery must be higher. And this is what battery really holds as its capacity - ampers per hour. By drawing more amperes, you discharge battery more. This explains why discharge rate at the same load is nonlinear. The more battery is discharged, the higher discharge rate is. This characteristics is quite complex to provide reliable charge/mileage data.
  14. Seba

    KS18L Range Test - 200 lbs rider

    I also think that 100 km is an exaggerated range, because I don't believe it's achievable by anyone in real conditions on KS-18L. But if we imagine a light, delicate girl that like to ride slowly (on wheel, of course ) so I'm pretty sure that she could make a quite long ride - even more than 70-80 kilometers. I got 60 km and I think if I reduce mass of my gear in backpack (at this particular ride I was returning from my work, carrying my big&heavy Dell M4800) and reduce speed to about 15 km/h, I will get another 5-10km more. PS. I get 20Wh/km brt. Brutto, because this has been measured at charger input - I need 20Wh drawn from wall outlet to drive 1 km.
  15. Seba

    KS18L Range Test - 200 lbs rider

    Today I've done a range test on my new KS-18L. Ambient temperature was about 25 °C (77 °F), 80kg of rider weight + 10 kg in backpack (about 200 lbs in total). Light wind, mostly flat terrain. Tire pressure about 300 kPa (43 psi). Fully charged battery, wheel with 1.0.7 firmware and set to medium riding mode. I was able to make a 60 km (37 mi) trip with average speed of 20 km/h (12 mph) until wheel started to show charge levels below 10%. Wheel was still able to ride gently with small speeds, but it was clear to me that my ride was finished Tiltbacks and "Please decelerate" messages was frequent enough to prevent me from further ride. From my point of view this is exactly what I expected, so I'm happy with that I was getting 40 km (24 mi) in similar conditions (but with smaller average speed) from my KS-16C with 680 Wh battery. So I think by riding with lower speed I can get considerably better range from my KS-18L. IMPORTANT NOTE - this was "real" 60 km, measured with GPS. It's important, because mileage reported by the wheel application is considerably inflated, even 20-25% up. This does not apply to speed, which seems to be measured correctly.