Number Two

[anotherkiwi]

I love the Chinese...

Ordered my hidden wire brake switch, a new S06SC torque simulation controller with input for external speed sensor and other bits and bobs from bmsbattery. So they take the money and package the order and... nothing for 5 whole days until I send an e-mail "please ship my package" and then it goes into warp drive and arrives about one and a half days later from their secret Dutch warehouse (and you thought it was coming from China... ). The good news is that this is a 16 Amp peak controller (15 + or - 1 Amp) which means a potential 664 Watts with the battery hot off the charger.

Of course on arrival that which should have been plug and play isn't. The plugs for the brakes on the photo on the site are the right ones.



And the right number of wires for the three wire switch! But my controller has two ugly huge white male plugs instead. One two wire and the other three wire black, red and yellow and yellow and black. At least all the wires are there. Cut plugs and solder are the answer to the problem.

I thought I had a spare male XT-60 for the power supply but ho hum nope, so I will just do a splice from my extended power cord which I don't need any more.

Could this be the end of the electrics phase? Stay tuned as tomorrow is night off so fitting and testing are planned.

 

[anotherkiwi]

Oh joy!

So I soldered an XT-60 on and went down to mount the controller. Well the male LCD3 connector on the controller doesn't plug into the male LCD3 connector on the LCD, what a surprise!

From the bmsbattery site:
"2. It fits all S series LCD meters. Please select a S-LCD1 or S-LCD3 as well."

Fail bmsbattery!

Back to the soldering iron...

 

[anotherkiwi]

A tail of two controllers...



A bit of soldering and crimping and splicing later. I kept the original connector "just in case" I had to fall back to something I know works.

Tony's top tip for soldering JST-XH male connectors

Plug it into the female connector which is already crimped to the controller wires. The soft plastic melts like butter and the wires act as a heat sink. Since the above photo heat shrink tube has been applied over insulating tape.



Those huge white plastic connectors are for brakes and throttle, not quite the same as those shown on the Bmsbattery site...

OK so now that the dirty work is done and the controller is tucked away in a Bmsbattery plastic controller box (large - vanity photo soon) and connected what's it like?

Bl**dy awesome! Fortunately all the soldering grunt work was not in vain the S06SC is a fabulous controller when mated to the GSM. There is absolutely no transition felt when going past cut off speed, it is as smooth as silk. The first few times I looked down and I was doing 30-32 km/h and the Watt meter was showing 0 W, I hadn't felt a thing!!! I thought the motor wasn't working until I let speed drop to around 22-23 km/h and it slowly kicked in again just as smooth as during acceleration.

I LOVE IT!

There is quite a bit of run on and the PAS still kicks in when the pedals turn in reverse so all is not perfect but the riding experience is fabulous. Now I understand what some have been saying that a torque sensor mid-drive is more bike like. This isn't a torque sensor of course but it appears to be doing a top notch imitation of one.

If you let the cadence and the speed drop when in assistance level 5 it continues to provide power up to about 30 km/h, maybe because it is configured for "stronger assistance" in the C14 parameter of the LCD-3? This is very handy for going under the main road and back up a steep little bit we have here for example.

Yesterday I did a test run to the station and back and averaged 26.7 km/h without really forcing things. I must insist that I am getting a great kick out of wondering if the motor is broken or not because I can't hear it and looking down and seeing that I am above the cut off speed without realizing.

PS gear changes are also even smoother than the smooth they were when running unrestricted!

Next chapter: finishing the last little things...

 

[trex]

I am surprised that the torque immitation controller is not more popular.

 

[anotherkiwi]

http://www.roswheel.eu/roswheel-shop/all-items/top-tube-double-phone-bag

Now how much lipo can I fit in there...

 

[danielrlee]

View attachment 15307
http://www.roswheel.eu/roswheel-shop/all-items/top-tube-double-phone-bag

Now how much lipo can I fit in there...

Probably not as much as you think. I bought one ages ago with a mind to fit 2x 6S 20C packs either side, but it would only take one (published dimensions don't take into account the wedge shape). Undeterred, I cut the phone holder out and sewed on a couple of small camera bags instead. You might have a bit more luck with Multistar packs though.

 

[anotherkiwi]

Thinking about a small 8 Ah pack for every day use. And as a range extender for getting home when the main pack runs out.

 

[anotherkiwi]

Big Fine Battery



I now have a 20 Ah battery composed of two different types of Hobby King Lipo packs (same size, same cell count). The Multistars are out of stock for months now so I went for Graphene on sale. Bigger, heavier but not by much. I was really scared that I wouldn't be able to solder the XT60s on because the wire is so thick. In fact it is 10 AWG on the inside of much thicker insulation than the Multistar I only just managed to coax the gray insulation cap over the wire. Above are the 4S wired up in parallel with my new lip alarms.

Road test

So I charged up to 41 V (5 hours in all) and set off on a road test. 48 km in 1 hour 48 minutes. That used 8.364 Ah at 6.79 Wh/km which I estimate is about half the useful capacity of the battery. Happy bunny!



I love this bike! It does need a suspension seat post, several km of gravel road down the valley to Bera convinced me of that! It breezed up the pass at Ibardin (the big bump above) at 18-20 km/h average in second and a couple of hairpin bends in 1st. I did manage to conk it out on a ramp over 15% up to the top Venta at Ibardin, bottom gear not low enough.

 

[anotherkiwi]

Gearbox update:

The owner of the LBC said my little slope that I couldn't climb is over 20%!

In any case I now have an 11-32 cassette I know, doesn't exist, you didn't think I was going to mount the 12 and give up on my cheap downhill thrills did you?

So I went out and tested on my local 1 km 15% test hill. I had to work hard but didn't draw more that 300W all the way to the top. The cassette is better than the 11-28, 3rd is a very comfortable slow speed around town gear. Second is 26 so a couple of teeth more than my previous 2nd gear and will climb most normal hills (my normal is bits at 12%).

If you are thinking of getting a GSM the ideal cassette on a 28" - 700C wheel bike is an 8 speed 11-34 if you are reasonably fit. 11-32 would probably be the best bet on a 26" wheel but I haven't crunched the numbers. If you are less fit you don't need a high top gear so a 7 speed 13-34 hyperglide would do the trick.

 

[Deleted member 4366]

I am surprised that the torque immitation controller is not more popular.

Torque imitation means current control. All that blurb about huge pedalling database is just made-up sh!te. I have a logger that shows that for each level, you get a fixed current (or watts - same thing). These controllers have been around for a while now and are more or less universal in kits and they're installed in many OEM bikes.

Current control is more or less essential for crank-drive bikes for obvious reasons.

IMHO, these current control controller with a simple on/off cadence sensor give a more than adequate riding experience for normal general purpose riding. I wouldn't want anything else. A torque sensor is only necessary if you want to do difficult off-road riding where you need a more sensitive power response. OEM systems, like Bosch, use torque and cadence sensors, and we're starting to see over-rides of the torque sensor for easier hill climbing, like on the Kalkhoff "Climb assist". A torque sensor on it's own for general purpose riding would be worse than a cadence one.

 

[anotherkiwi]

This one is so different from the other two controllers I have. The ramping up and down of the current is so smooth. The initial supply of power is strong but once at speed very gradual. It also seems to go beyond the current limit, I think this is due to the C14 setting of 3 "Stronger assist strength of motor"?

If only it was 36/48V, now that I can do 12S...

 

[anotherkiwi]

While mucking about on bike-discount.de of course I had to discover the CS-HG50 13-34... I'll try my 11-32 as soon as the weather permits but now I know I can make an 11-34 in 7 speed. 11-15-17-20-24-29-34

http://www.bike-discount.de/en/buy/shimano-7-speed-cassette-cs-hg50-13-34-535334/wg_id-8169
http://www.bike-discount.de/en/buy/shimano-sprocket-11-t-with-spacer-ring-for-cs-m771-bj-bk-162015
http://www.bike-discount.de/en/buy/shimano-distance-ring-pu-3mm-for-7-speed-cassettes-37437/wg_id-8175

 

[anotherkiwi]

New Lipo charger arrives

And first Lipo charging incident ensues...

No! No white smoke, no projection of flames, no fried fingers. Sorry no real drama. A burning plastic smell told me something was amiss and the 10 year old converted PC power supply promptly made a "pop" sound and ceased to function.

Maybe my drawing 18 Amps constant was a little beyond its means?

It did charge the two 4S packs to 16.6 V from storage voltage at 1C (10 Amps) in just over 35 minutes. It did bravely manage to provide 9.6 Amps for a short while on the 6S packs.

So it looks like tomorrows long ride is cancelled "due to unforeseen circumstances" A new power supply is on order.

Tony's top tip:

A 300 W charger really does need a 30 Amp power supply to charge 6S at 10 Amps...

 

[anotherkiwi]

Power supply autopsy:

A condenser (? composant which looked like a brown chewing gum rather than a tiny coke can) has failed, spewing grey goop (is that word copyright free now Gwyneth?). I had bet on a capacitor failing so I lost that bet... <pays self 1€>

I have chosen to go with an R/C 540 W 30 Amp power supply. I may buy a 20€ PC power supply as a backup. My choice was motivated by two 15 Amp x 18 V outputs from which I will be drawing just over 12 Amps each.

Of course charging the 4S Multistar pack to 4.15 V per cell was always going to be a test of iffy cell number three. Test failed... The cell is losing voltage by the hour. When I get back from shopping later I will put it in discharge on the Battery Medic just to keep it safe. It charged to 4 V and behaved itself for the last several charges. It has held storage voltage for weeks on end. Just doesn't like full charge. Looks like a replacement 4S pack is on the budget for next month... Cost per month of broken one: 4.65€ or 0.05€ per kilometre (estimated)

The good news is that charging time for the whole 20 Ah battery will be about an hour from storage voltage to full capacity, two hours from empty. That is a breakfast time charge, or out on the road, a lunch break (hey I'm European - two hours is normal lunch!) charge.

 

[malkie0831]

New member here, been reading lots but this is my first post!

I admire your determination anotherkiwi to use RC packs instead of a regular eBike pack, I decided I wanted the plug and play simplicity of a Li-On battery with a built in BMS when I decided to order my conversion kit.

In my RC hobby I have recently modded a Server Power supply for charging car RC packs. Follow this link:

http://www.rchelination.com/setting-hp-dps-1200fb-power-supply/

and you too can make a very small 100A 12 v supply (yes, 100 amps!) or even stack two together for 24v at 100A.

 

[anotherkiwi]

I now have a VERY BIG (zut won't fit in the battery box...*) RC power supply which is finishing a 10 Ah Graphene pack as I type. Charging at 1C the Multistar packs puff ever so slightly, the Graphene packs don't. 35 min from storage to full voltage (4.15 V per cell) on the Multistar and 31 min for the Graphene. So they are different.

The choice of lipo was first of all one of cost - I have been running a 10 Ah pack since March and it only cost me 180€ with the charger and all the wires and bits you need.

Then I discovered the lack of voltage sag - most of my rides have slopes of up to and over 12%. My entry level Li-ion bottle battery doesn't like that and a replacement Li-ion battery that has enough power to handle the hills is eye wateringly expensive. Or from China, but I prefer a good old European consumer protection guarantee when I buy stuff that costs over 200€.

Third I go places where there is no mains electricity. There is a solar panel and a 12 V battery I can hook an R/C charger up to however.

Fourth seeing how my >260€ Li-ion battery is behaving after about 3400 km I don't mind the idea of getting 100-150 charge cycles from Lipo then recycling. It is going to be more cost effective on a per kilometre basis. Swapping out a 27€ 4S pack that is showing weakness after 6 months use is not bringing a tear to my eye.

* The power supply, two 300 W chargers and 20 Ah of Lipo do fit in the battery box but not in the way I would like them to fit

 

[anotherkiwi]

A what if? ride

So my 20Ah battery was all charged up to 41.5 V time to try something new. I did my usual station run but left the assistance set to 5 all the way and tried not to pedal too hard. Kept the right cadence of about 70 and used the gears properly.

Result 8.5 Wh per kilometer. Cool as a cucumber and not tired at all. So it does work OK as a means of transport this electrical bicycle thingy! I'll go back to being my normal push hard on the pedals self next time out. The difference in power use between the two riding styles is about 2 Wh per kilometer. The difference in range (80% battery use) is 21 km.

My iffy cell stopped misbehaving and stopped dropping volts at 4.1 V so it was 0.05 V lower than the others.

The Graphene pack cells are within 0.01 V of each other whereas the Multi stars are within 0.03 V.

That is all of the (I hope) useful data of the day.

 

[Alan Quay]

Interesting how close to the often quoted 7.5wh/k (12wh/m) you are.

8.5 wh/k when being lazy, 6.5 when putting effort in, neatly straddling the average. (in imperial the numbers are 13.7 wh/m and 10.5)



Sent from my HTC 10 using Tapatalk

 

[anotherkiwi]

I'm feeling more "normal" than last year when I was learning the ropes and really giving it some welly I was down around 5.7 - 5.8 Wh per km once I got fit. flecc said I didn't even need a pedelec! But then there are those hills...

I think during a normal ride when I am really needing the motor most times I am using over 10 Wh going up and of course 0 Wh coming down, the average of 6.78 Wh the other day resulting from the not so flat bits where the motor helps a little bit. On the gravel section I was at 23 km/h average in assistance level 3 so drawing about 90-115 W especially when stop/starting through the bits with flood damage.

 

[Benjahmin]

Floods??? You have floods? I thought you said the sun always shone?
Just teezing.
Do you think those Wh averages would be approx. the same for a hub?