10 year old lead acid 36V e-bike needing an upgrade!

[amberwolf]

If you can find one with multiple shunts inside, like say, 2 shunts and 40A, cut one shunt and now its' a 20A controller.

[monkeychops]

If you can find one with multiple shunts inside, like say, 2 shunts and 40A, cut one shunt and now its' a 20A controller.

There are a couple of 36V 40A controllers on the TNC Scooter site but neither of them say anything about 2 shunts

[dogman]

Some of those might be potted, and hard to do modifications like that. The ebay one looks as good as any to me.

Also, you don't have to ride full throttle. Well I do, but you don't.

[Samd]

If it were me, I wouldn't but a 36v rated controller. That's not to say you shouldn't do it! But I found a problem with this when i did my conversion.

As soon as your lead acid battery sags below 31.5 volts it will cut power to the motor. And on a lead acid battery that is bound to happen very quickly as soon as you try and accelerate hard.
I'd buy a 30 amp 24v controller instead. They will work fine when "overvolted" at 36 volts and you won't know the difference.
I recommend the LBD 14 and have now used two of these daily for about 6 months at up to 1100 watts. Barely gets warm on the casing.
I think I have a third at home - could have a look inside the case in a few hours to see if it is potted or not...

If you don't want a keyswitch just short those two wires out. Or put a small switch across the two wires somewhere out of sight to deter theft.

Then fit a voltmeter in a spot that's easy to see to watch how much your voltage sags - use it to understand range vs speed. When you convert to LiPo you will see a much lower sag, but it will help you avoid going below 3.7v per cell and killing your batts. A much better fuel gauge than those throttles with an LED. And as dogman says it is important to understand the tradeoff between speed and range for your bike.
http://www.ebay.com.au/itm/5PCS-Digital-Voltmeter-DC-4-5V-30V-Voltage-Red-LED-Mini-Digital-Panel-Meter-/270836381709?pt=LH_DefaultDomain_0&hash=item3f0f1b380d
i recommend this sort. Don't worry about the stated voltage range of that meter being 30v max, mine is accurate to about 0.25volts at 33.6 volts. Happily works up to 40v and there is only two wires on this one - just pop it across two the power input wires on your controller. You cna put a small switch in series to turn it off, or just leave it on, they use very little power.

And the warm red glow shows your e-grin as you quietly whirr past people at night...

Which country are you in Monkeychops?

[monkeychops]

Thanks for that samd.
I am from England.
It does sound then like a 24v controller is the thing to get. So is the low voltage cutoff the only significant difference between the 24V models and the 36V? In my ignorance I assumed that they would output 36V or 24V to the motor and 24V going into a 36V motor wouldn't be enough. I guess it doesn't work like that?
The voltmeter is an excellent tip too. Currently I have an analog voltmeter without any numbes on it. The digital one looks good. Is it reasonably weatherproof?
Like your blog too!

[Samd]

Thanks monkeychops - there is a lot of similarity between the aprilia rebuild on my blog and what you want to do - essentially strip everything off prior to the motor and update/simplify it.

My voltmeter is quite waterproof because of the way I reused the Aprilia's plastic overcap, then effectively 'potted it' by shooting up the back full of hotmelt glue to seal it up. It may pay to buy one and pop the bezel off, then glue the bezel back on to the plastic overlay in front of the screen so water can't get under the bezel/surround. I'm sure you can come up with other ideas. Cheap enough to buy two..

Because these voltmeters are an unknown when it comes to accuracy it pays to compare the stated voltage when you plug the battery at various charge states into a LiPo charger - A good liPo charger will tend to be more accurate on stated voltage. Or compare with another quality voltmeter if you can find one handy or borrow.

The good news is that when you are using for example an 8s LiPo pack, even if the voltmeter is ~0.4 volts in error, you are dividing that error across 8 cells in series, which is about 0.05 volts on average. Not a big showstopper when you are trying to avoid your cells going below ~3.70V.

I am sure that a 24v controller just takes any input above 21.5volts and chops it into a signal of varying pulse width output to the motor. The voltage will pass thru pretty much the same. The key is not to exceed the rated voltage of the FET's inside, which sounds like it is often 40V for the cheaper ones. *I think*. If I can find a decent CRO I might test in future.

As an aside my brushes wore out on my DC coffee grinder on the weekend. My dad was in town and showed me how to shape new ones using a file and some spare brushes I got with my last 20 dollar chinese angle grinder. It was so simple I can't believe I was ever worried about my brushes wearing out on the aprilia bike. Don't ever avoid riding hard for the sake of brushes. If you can get to them easily inside the motor , it's a 5 minute job to make more and pop them in.

[monkeychops]

I think what I might do is keep using my existing analogue voltmeter (which is in a weather proof case), but calibrate it with a multimeter. I think (but need to find out) that Full = 39V and Empty = 35V. My 15 mile journey to work drains it pretty much all the way from one to the other. The needle obviously drops under load but never significantly below the Empty line. Resting voltage after this journey is just under the mid-line, so around 36.5V I think.

I really want to measure current between the controller and the motor as well, to get a better idea of my current setup. Almost certainly I'm going to be replacing my brushed 200W motor at some point but I'd like to see how much current the existing motor is drawing, especially under load. Can anyone offer advice on doing this? I thought just connecting a multimeter in series would work. But mine says '10A' on it so don't want to overload it. I've also read it's not a good idea to measure current for too long (overheating issue maybe? I can't remember) but I would really like to ride around and see how the current to the motor is changing.

Maybe one of those 'clamp' style current meters would work?

Anyone?

[amberwolf]

Measuring current between motor and controller is difficult, because it is not constant nor DC. Depending on controller design, it may even reverse itself during throttle-down or braking, if it regeneratively brakes the motor.

But if yours is a Brushed DC motor, you can measure battery current and call it the same as average motor current at any particular time. (peak motor current may be MUCH higher) THere are some good sites to learn about that, including some threads in the motor section of ES, and http://www.4qd.co.uk/faq/index.html or http://www.4qdtec.com/pwm-01.html


Most multimeters have a rating for how long you can use it for higher amp readings; you'd need to check your manual for htat. If it's a Centech/Harbor Freight/ other cheap meter, it's proably good for 30 seconds at 10A. Higher currents than that, or longer, and it may unsolder the shunt from the PCB inside the meter (ask me how I know ).

removing the back cover to expose the shunt to cooling airflow may make it survive continuous 10A, but going over that may well still cause the shunt to fall out/off.

Better meters may be able to tolerate rated current forever, or might have similar limitations.

Clamp-on meters can do their ratings all day long, or at least till the battery runs out.

[dogman]

A cheap harbor freight, or equivilant 12v car ampmeter will work for crudely measuring your draw. Might be poorly marked like mine was, in 5 amp increments. But when I got a Cycleanalyst later, I found the readings were not off by enough to care about. Just harder to read on the fly.

I wouldn't sweat the 36v controller thing, if it's lvc is popping, it's because it's trying to save your batteries. Using a lower lvc so you can kill your batterys better is not a solution.

[monkeychops]

Thinking about all this measuring a bit more...

How is my bike limited to a powered speed of 15 mph? there is no feedback from the motor to the controller. When I’m going along downhill it still cuts out at 15mph. I’ve got a feeling that no amount of increased current (from a replacement controller and/or battery) will help with this. Top speed must surely be related to voltage in this case?

So to test this I could wire up another 12V source in series with my 3 x 12V SLA batteries (I’m learning the acronyms) to make 48V and see if this spins the wheel round faster. I now wish I hadn't chucked away (recycled) my old 12V SLA batteries. Would it be safe to wire in a 12V mains transformer in series to the batteries? Instinct tells me this wouldn't work or be safe... so I'd need an extra battery from somewhere.

[dogman]

It should spin faster with 48v.

Several things limit speeds for motors. Primarily the winding of the motor will yeild a set speed for a given voltage. So your motor may just be a slow type of winding. Another way to limit speed is amp limiting. You controller may allow a higher amp spike for a second to get going, then limit amps. If low enough, that low amp limit will limit speed. If there is x amount of power, a given weight and wind resistance is only going to go but so fast. That's the idea behind 200w limits for ebikes in some countries. Make it so the thing can't go fast, because 200w will only do but so much.

And then there is gearing, for motors that go through the gears. you can limit speed with low gearing.

You can keep adding volts and power to that motor, but at some point, a long ride will let the magic smoke out and it won't work anymore.

[monkeychops]

I think I'm going to try the 48V route. I've not got a lot to lose. If the motor or controller break, well I was considering replacing them anyway.
I've read a few posts on here from Powabyke owners who have overvolted successfully to 48V. I think my bike is very similar to these Powabykes.
It's only £25 for another 12V 12Ah SLA.
If it works I expect I will end up craving even more speed and power at which point it'll be time to move on from lead-acid.

I've attached a pic of the bike and battery case. I'm planning to plonk the extra battery on top, where I've drawn a box in the image. Any advice on how to attach it would be really helpful. I'm a bit of a novice at all this. The only solution I can currently think of involves using a lot of strong tape and drilling a hole in the current battery case for 2 wires to feed through. My main requirement is being able to lift the battery case out (as one unit) to take it into the office for charging at work.

IMAG0514_with_box.JPG

(206.73 KiB) Downloaded 2 times

[Gordo]

The only solution I can currently think of involves using a lot of strong tape and drilling a hole in the current battery case for 2 wires to feed through. My main requirement is being able to lift the battery case out (as one unit) to take it into the office for charging at work.

If you are concerned with water getting into the battery case, drill your hole on the side and use what is called a "drip loop." Form the wire down slightly in a "U" and then back up and into the case. Duct tape is good for a temporary test, then use Velcro plus a nylon strap. Velcro between the extra battery and current case, acting in shear, will not come loose if a strap around the case is holding it down.

[monkeychops]

If you are concerned with water getting into the battery case, drill your hole on the side and use what is called a "drip loop." Form the wire down slightly in a "U" and then back up and into the case. Duct tape is good for a temporary test, then use Velcro plus a nylon strap. Velcro between the extra battery and current case, acting in shear, will not come loose if a strap around the case is holding it down.

Thank you. I'll probably need to cover the extra battery with opaque tape or something, so it doesn't look too much like a bomb or something as I'm lugging the whole pack into the office!

[Samd]

Hows the build going?

[monkeychops]

Thank you for asking.
The extra 12V SLA arrived today.
It all seems ok on 48V except with zero throttle the rear wheel still wants to turn at a speed of about 1 turn per second.
Road tested it very briefly in the dark tonight and early indications are good.

with extra 12V SLA

Comparison of rear wheel rotation speed:
36V http://www.youtube.com/watch?v=xx_OJI_J348
48V http://www.youtube.com/watch?v=BT5Ej36rNZk

Will do a slightly better road test tomorrow in daylight.

[Samd]

Hey MC, thats good news. Speed increase would be interesting.

Same controller? If so, perhaps the old dedicated controller is looking for a certain minimum voltage in from the hall sensor/throttle and the input voltage to controller has now been amplified a small amount - because of the 48v stepup. Maybe a new controller is warranted to try.

[monkeychops]

Took it out for a quick test ride up and down the road in the light rain. Confirmed it reaches 20mph under power on the flat. Acceleration is noticeably better (but not greatly improved). now I'm annoyed it's the wrong season for riding to work. My route goes through a field for about 1 mile too, so really muddy now, this being England. I will give it a longer test if it stops raining tomorrow.

I may replace the controller, but I'm going to give the existing one a go too. I think that if I upgrade the motor to brushless then I'd need to get a new controller anyway at that point.

New digital voltmeter and charger on its way now, courtesy of eBay.

Does anyone know what voltage the 48v SLA pack should read and what an 'empty' one would read? Just ballpark figures.

[amberwolf]

Empty depends on how long you want them to last, but for each SLA, 10V is dead empty. 12V resting is probably a good place to stop if you want to keep them around at all, especially if you can't immediately recharge most of the time.

13.6V is usually "full" for a typical SLA in good condition. Some of them less or more.

So for four in series, 54.4 "full", 48V empty, 40V dead empty.

[monkeychops]

Good answer - thank you.

[Samd]

In the long run these are handy for measuring how many watt hours you have, as well as peak amps. Cheap. http://www.youtube.com/watch?v=kfs4Bs3H ... ata_player

[monkeychops]

Thanks for the link. I was already thinking about getting one of these somewhere along the line. Probably not yet though.
I wondered if they were easy to connect up? Since they measure current as well as voltage, don't they need to be connected in series and parallel at the same time? Or can they literally just be connected up to your battery as a normal voltmeter would be and they just work? From the video, I *think* he's got it connected in series...

[monkeychops]

Update:
Took it out for a 6 mile test ride yesterday. All went fine. Average speed 17mph, a few gentle hills. Didn't notice any heat from the motor or controller.
Battery voltage was sagging to 42v under load at the end.

This morning:
Charged the batteries up to full, reading about 52v. Motor seemed dead, throttle did nothing. Back wheel was still turning at one turn per second when I turn the ignition key to on. I had to get my normal bike out, which felt awful
Came back to it later, took the cover off the controller and started testing things with a multimeter. All of a sudden the motor seemed to come back to life just as I was testing the throttle voltage on the controller. Shortly after that, noticed the controller was getting *very* hot. So hot I could only touch it for about 3 seconds. I wonder if it suddenly started working when the input voltage had dropped off a bit. When I tested it yesterday the batteries were probably only about half charged.

Maybe I need to replace the controller. What do people think? I know Samd has already suggested this.
Something like this one: http://www.ebay.co.uk/itm/Electric-Scoo ... 0590002122
350 watts might not be enough?

This one better? http://www.tncscooters.com/YK40-4.php

Any advice appreciated

[monkeychops]

Further update: if the battery voltage is below 51.5v when I turn the ignition, the throttle works fine. If it's above 51.5v I can get the throttle to work by turning the key whilst turning the throttle at the same time. It seems like the controller needs to 'see' less than 51.5v before it starts working.

I went for another 6 mile ride after I'd got it going and all seemed fine. Average speed 18mph with some pedal assist.

[monkeychops]

Oh dear, further update: I think the controller broke.

I was just setting off and had to turn the ignition off and on again in the first mile of the journey (to get around the 51.5V cutoff problem). Then while stationary, the motor kicked into life of it's own accord. The throttle would not respond. In other words when I turned the ignition key I had instant full power, regardless of what I did to the throttle. The only way to cut power was to turn the ignition back off. Made for an interesting 15 mile journey back home! I don't think the throttle itself is malfunctioning because the brake levers (which previously would cut power when pulled) also didn't stop the motor.

Time to buy a new controller I think.
Something like this be ok? http://www.ebay.co.uk/itm/Electric-Scoo ... 0714562156
or this? http://www.ebay.co.uk/itm/Electric-Scoo ... 0590002122

The motor is 200W (or possibly 250W I'm not sure). So presumably the 350W controller would work ok?

Any advice greatly appreciated.

<edit>
I've got a feeling those controllers are for brushless motors, so might not be any good.
This sort of thing would work I guess? http://www.ebay.co.uk/itm/Scooter-Elect ... 0569778580
</edit>