Live data from Hacker News

Show HN: Long Range E-Bike

jacquesmattheij.com

531–540 of 626 posts

Re: Show HN: Long Range E-Bike

#531
post #462
post #459

Earlier quoted context omitted.

I hope you realize you're one pebble or pavement imperfection away from breaking every bone in your body at those speeds.

At those speeds bikes are incredibly stable. The only way a pebble of pavement imperfection will harm you is if the bicycle fails mechanically (e.g. a wheel folding in on itself). I'm pretty sure that the risk of being hit by or hitting a vehicle is multiple orders of magnitudes higher. Which isn't exactly better, but let's at least worry about the correct sources of danger.

Cornering grip at speed and some gravel can really ruin your day, in spite of the bike itself being stable at speed.

Re: Show HN: Long Range E-Bike

#532
post #38

Awesome project Jacques! I saw a comment of yours on an article a day or two ago and was hoping to see this pop up soon. I've found the most difficult thing about riding an e-bike is the other motorists have no idea how to react to you. You're not really a regular bicycle anymore due to your speed, but you're also not a motorcycle that deserves its own lane. I have at least one car turn in front of me almost every tr…

Cars misjudging the speed of oncoming bicycles is a huge problem, even without pedal assist. I have added significant length to my bike commute to avoid places where cars turn across my path. I can take a 1% risk once in a while, but not twice a day.

This is rule #1 of bike safety: avoid risky situations.

Re: Show HN: Long Range E-Bike

#533

Earlier quoted context omitted.

Electric moped that works like a bicycle.

Still, the "mo" in "moped" comes from "motor". It's a vehicle with a non-human power source, just like a car. So I find it kind of weird that they're often allowed in bike lanes, especially where those are separated from pedestrians by just a painted line on a sidewalk (or sometimes even worse, explicitly mixed in with them!).

Because there are limited resources and e-bike cyclists will be much nicer to other cyclists than cars are to any kind of cyclist. Besides that if there is an accident it won't be a fatality right away.

Re: Show HN: Long Range E-Bike

#534

There are alternatives that don’t involve DRM or Bosch at all. I built our ebikes around regular bikes using a TSDZ2 torque sensing motor and an em3ev 52V battery. Total cost around 600$ + bike + tools. The battery is already assembled so no risk of setting your house on fire. The torque sensing makes it really natural (like having bionic legs), the whole thing has an open source firmware if you want to replace the C…

Where I live you can't legally cobble together your own s-pedelec, you need type approval.

Re: Show HN: Long Range E-Bike

#535

Earlier quoted context omitted.

Yes, that's at 5% state-of-charge, which is a bit low. If you charge up to 100% and discharge to that cut-off point you will get 300 to 400 cycles out of your battery. If you go from 80% state-of-charge down to 20% then you can get thousands of cycles out of your batteries. That's one of the reasons I designed that pack as large as it is: to stay within the 80-to-20 range on a 100 km trip.

The “charge to 80%” advice isnt generally useful to PEV riders in 2021 It is true that a lithium _cell_ lasts longer when charged to 80%, but a _battery of cells_ must be “balanced” and PEV BMSes only do that balancing in the constant voltage stage of charging in the hours after the charger indicates the battery has been fully recharged. For anyone just learning about this, an unbalanced cell is usually the cause of…

There is an active balancer in the pack I built. It keeps the whole pack within 2 mV.

Re: Show HN: Long Range E-Bike

#536
post #526

Earlier quoted context omitted.

The “charge to 80%” advice isnt generally useful to PEV riders in 2021 It is true that a lithium _cell_ lasts longer when charged to 80%, but a _battery of cells_ must be “balanced” and PEV BMSes only do that balancing in the constant voltage stage of charging in the hours after the charger indicates the battery has been fully recharged. For anyone just learning about this, an unbalanced cell is usually the cause of…

By charging and balancing individual cells, you mean group of cells that are connected in parallel (the 4P in 12S4P)? I don't see how or why would you balance individual parallel connected cells, unless you had a different battery configuration. Simplest "balancers" drain cells that have too high a voltage as you are approaching full charge. You do need to occasionally top the battery off, to keep it balanced. Seeing…

The pack I built has a balancer, that's what all the red wires on the photograph are for. It uses a charge pump to move some charge around so that the high cells get clipped and the low ones brought up. It keeps the pack within 2 mV. The unit it monitors is the cell group, all 10 of them have a wire to the balancer and a ground for reference. It is very accurate and it wills substantially contribute to the life-span of the pack.

Re: Show HN: Long Range E-Bike

#537

Earlier quoted context omitted.

It adds about 7 kg.

So that’s 7kg extra for your 2150Wh, over the original 500Wh battery? So around 250Wh/kg? I’ve been planning a velomobile that I can live out of, and I’ve been interested to observe that by the time you get to one kilowatt hour, I can’t find any Li-ion batteries at all: they’re all LiFePO₄. https://au.renogy.com/renogy-smart-lithium-iron-phosphate-ba... is then about 1280Wh (12.8V/100Ah) for 11.8kg, 110Wh/kg, around…

Yes, LiFEPO4 would have been preferred but the Bosch BMS doesn't know how to deal with that chemistry. There are some interesting developments, have you looked at these?

https://www.youtube.com/watch?v=kYx097cVR48

If I had free pick of chemistry I probably would have gone with those. 250 Wh/kg sounds about right, I'm running the cells in a very conservative regime though, topping them up to about 80 SOC and discharging to about 20, so my effective range is smaller but the battery life-span will be very long, many thousands of cycles. The balancer also helps with that.

Re: Show HN: Long Range E-Bike

#538

FYI: spot welding is a lot easier and precise using this welder: https://www.keenlab.de/index.php/portfolio-item/kweld/ With the capacitor module, you can easily power it with a lab bench power supply.

Finally, a constructive comment instead of a bunch of legal stuff and fantasy performance figures. Thank you very much, I'll look into this.

edit: Hm, the power supply looks to be a LiPo pack, the other two options are no longer available. But very nicely executed kit, now to find a way to power it from a wall socket.

Re: Show HN: Long Range E-Bike

#539

Earlier quoted context omitted.

They are super dangerous too. Very low to the ground, nearly invisible from the perspective of a motorist and far harder to overtake with on narrow bike paths than regular bikes. We have a couple of them near here, also a few battery assisted ones. The velomobile would be an ideal vehicle if everybody rode one.

I haven’t yet ridden a velomobile (though I may purchase one soon and am planning on building one for riding round Australia and living out of for a year or so), but I’ve been riding a recumbent tricycle (Greenspeed GT3 Series II) since 2014, including various touring in Australia (Victoria and some South Australia), America (through California, and from St Louis to Philadelphia) and New Zealand (Auckland to Kaikoura…

That sounds like an amazing thing you are doing there, do you have a write-up somewhere?

Re: Show HN: Long Range E-Bike

#540
post #476

Earlier quoted context omitted.

How powerful and effective could that be without launching the lighter vehicle backwards with dangerous force?

From a purely theoretical point of view, if there was a perfect feedback control loop the energy of the explosion could be controlled to bring the front of both vehicles to a perfect stop. Both vehicles would then crumple about that stationary point. From a practical point of view ...

Trying to imagine two such vehicles colliding head on...
Post reply on HN