Motor speed control

[allotmenteer]

Hello all

While surfing the web I found a guy on a forum who has the same hub motor kit as me. He had discovered that on his controller there was what he called a pot which had been shrinkwrapped and cable-tied out the way. Apparently he did something to this and it made his motor go faster.

I went and checked my controller and found the same thing. On removing the shrink-wrap I found that there was a small (5mm*5mm) square thing with a round yellow dial on it which has a cross slot in it (presumably for a screwdriver). There are two blue wires connected to this thing. Of course I turned the dial and found that indeed it controlled the speed of the motor. It used to spin at 18mph unloaded and I found that I could get this to 21.5mph by turning the dial clockwise as far as it would go. If I turned it anti-clockwise then the motor went slower.

Is this thing a potentiometer?

Is it safe to tinker with this dial? I assume that what I have is a dial with which I can derestrict my motor.

Could I replace the tiny dial with a larger dial so that I could adjust the motor speed on the fly? My idea here is to use the high speed setting for flat terrain and a lower speed setting for hills.

My belief (mostly ill-informed ) is that moving the dial will place me at various points along the motor's power band and thus adjust the speed and torque output. This would seem to be a jolly useful feature I think. I wonder if anyone else's controllers have this dial?

 

[flecc]

Yes, it is a potentiometer, but what it controls can vary according to motor type. Most bikes don't have them, having fixed components in their place as part of the control circuit design.

At it's fullest setting you may experience cutting out under load, especially when the weather gets hot and the system heats with it. The range will also drop markedly. There's nothing wrong with having a variable control, but not with that component which I recognise from your description. It's what's called a preset, designed to be set once and left alone, it won't stand repeated usage.

You could take it out of circuit, check the stamped on value or measure it, then replace it with a variable pot, but be aware that even these are unreliable. The "works" are a resistive carbon track with the current connections at each end, and a sliding contact which can pick up the current from intermediate points. The carbon wears as the sliding contact passes over it repeatedly, and carbon particles can cause intermittent contact. It's this which causes "noisy" and "crackling" volume controls at times on radios, TVs etc.
.

 

[allotmenteer]

Thanks for the info on pots flecc. I'll read up about them on the internet.
I've tried various settings on the pot:
At minimum when I opened the throttle it wouldn't pull the skin off a rice pudding.
At 1/4 setting it pulls much more but seems to stop assisting at about 11 mph. At 1/2 setting (the original setting) it goes well up to 15mph.
The 3/4 setting seems the same but with maybe 1-2 mph more top end.
Max setting didn't improve on 3/4, in fact it seemed gutless at top speed (I guess this must be at the very limits of the power band where torque and watts drops off a cliff.

I've put the pot back to 1/2 setting now where I will leave it as I think that's the setting best matched to my motor as I found that as you said the range will suffer. Also, like you said, the controller heated up somewhat. Not burning hot but still much more than normal.

Incidently, the motor has never got even slightly warm even after I've run the battery to exhaustion. I wonder then whether I can run the motor at 36V? Now there's an idea, hmmm...

 

[flecc]

Incidently, the motor has never got even slightly warm even after I've run the battery to exhaustion. I wonder then whether I can run the motor at 36V?

Almost certainly yes, most of these motors will accept much higher voltages, 36 volt motors accepting 48 volts and sometimes even 72 volts. Not the controllers though, which fail if overdriven very much. You'd need a new suitable controller, and if your manufacturer doesn't do one, it's difficult to identify one that will be ok.

However, you could up the voltage just a bit using extra cells. The US Batteryspace company for example, supply 39.6 volt packs for owners of 36 volt bikes to pep things up a bit with only a negligible risk of failure. So 10% up is generally ok with the same controller. Do at own risk though, there is some.

 

[allotmenteer]

Not the controllers though, which fail if overdriven very much. You'd need a new suitable controller, and if your manufacturer doesn't do one, it's difficult to identify one that will be ok.

However, you could up the voltage just a bit using extra cells. The US Batteryspace company for example, supply 39.6 volt packs for owners of 36 volt bikes to pep things up a bit with only a negligible risk of failure. So 10% up is generally ok with the same controller. Do at own risk though, there is some.

How much play do you think there is in a 24V controller? At the moment with the li-ion pack it copes with a 29V fully charged pack (though the new pack is only reading 28.7V for some reason). I believe a 37V Torq battery pack would be 42V fully charged so I imagine that the controller wouldn't handle it. What about Nimh though, what is the real voltage of a nominal 36V pack (I think each cell is about 1.35V so I make it 40.5V - too much maybe?)

How much difference is there with controllers? Mine has 3 large wires from the motor and 5 hall effect wires. I imagine all controllers have connections for the throttle and also the battery. As long as I get a brushless controller then I imagine any old 36V controller will do

 

[flecc]

The 10% I mentioned would hold good for 24 volts as well. My comments were for NiMh like that Batteryspace pack. A nominal 24 volt pack of NiMh has 20 cells, so adding two gives the 10% extra. Adding three cells to 36 volts as on that 39.6 volt pack is the same 10%. It's the on load voltage that counts, the idling voltage (like that 29 volts) won't affect the controller as long as it didn't rise to very high levels.

A nominal 36 volt pack is far too much for your kit, a 50% increase, and the moment load power was applied the controller would probably go west.

Nominal 24 volt lithium packs have 7 cells on most types, but it's not practical to add a cell since it wouldn't have the same protection and control circuitry.

I don't know how flexible controllers are when applying them to different motors, it's something I've never tried. The main things are that the same type needs to be used, 3 wire or 8 wire, and the amperage rating will need to meet or exceed that of your motor.

It's a shot in the dark though as to whether one would match up with the rest of your kit. Having a word with Mark at Team Hybrid or Tony Castles at Electrodrive might get some information on using other controllers successfully.

One problem with adding cells to NiMh is charging. I suspect many chargers would adequately charge the higher number of cells, since chargers usually have enough voltage headroom for that. The 39.6 volt batteryspace pack copes in a different way, It's sold as three 13.2 volt coupled packs with three 13.2 volt chargers, so they obviously have a question mark on using existing chargers.
.

 

[allotmenteer]

A nominal 36 volt pack is far too much for your kit, a 50% increase, and the moment load power was applied the controller would probably go west.
.

I experimented today with some 12V 7Ah SLAs I rescued from an old UPS at work. I don't know if they're up to the task of powering an electric bike but they were free so I had a go. First I put them in 2S2P to give 24V 14Ah and hooked them up to the bike. It went ok but after a short while (like 2 mins) they sagged badly and were no good. After a little rest they recovered a bit.
Next I put 3 in series to give 36V 7Ah and connected them up. Interestingly my throttle LED finally left the red minimum indication at last! I got the amber medium light on . I think my throttle is set up for 36V.
Next I lifted the rear wheel and opened the throttle and the wheel spun much faster than at 24V. No sparks yet so I applied the rear brake to simulate load and did that for a couple of minutes whilst checking on the heat from the controller. All seemed ok so I took it out for a spin. I didn't get to run on the flat, only up and down some small hills (say 2-5%). The bike accelerated to 15mph no pedalling up the hill on my street which was quite impressive. After the short spin I found the controller was getting a bit warm to the touch so I stopped as I didn't want to fry it.
The motor was stone cold as usual.
Well, it looks like my controller won't handle 36V for long (unless I can fan cool it or something) but it has made me want to upgrade to 36V. I'll look into a new 36V battery pack and a 36V controller to go with it. As you say flecc, I'm sure the motor can take the extra voltage so I'll give it a go. I've got some research to do I think.

As for the charger I will buy the battery pack and matched charger to ensure compatability. I'll price things up but I think if my company does the cycle to work scheme again then a Torq will be the cheapest option I reckon.

Paul

 

[flecc]

I thought the motor would take it ok, and it seems like you have a quite tough controller. I can think of some that probably wouldn't have stood that. Forced cooling is an option with a fairly powerful fan, though the Summer hot weather might test it beyond the limit.

As you say, the Torq is probably a less expensive option ultimately.
.

 

[allotmenteer]

As you say, the Torq is probably a less expensive option ultimately.
.

Yes, and it is off-the-shelf too. With two young children I don't really have the time to mess around with kits at the moment (or for the next several years I imagine ). I shall sort out a test ride on the Torq and then badger my work until they run the cycle scheme. I'll probably sell the kit quite soon to free up a bit of cash for a Torq.
Thanks for all your advice.

Paul

 

[allotmenteer]

Just thought I'd update you on my ebike tinkering.

After using my electric bike for a few commutes now and wanting a bit more speed to cut the 54min to 1 hour 3 min trip (depending on route) down to 45 - 50 minutes (or less ) I decided to have another go with the SLAs. This time I added a 12V SLA in series with my 24V Nimh and went out for a 4 mile spin. What a difference it made . The bike really came alive. It felt much more powerful and the speed increase was excellent. 20 mph was no problem! With pedalling 22-25mph was achievable. On downhill slopes I could feel the motor pushing to about 24mph. At 24V the motor alone was good for 15 or 16 mph and 18 mph with pedalling.

The controller seems to be ok. The aluminium case got fairly hot (not enough to burn but I estimate it was about 35-40 C. I'm not sure if it will cope with 15+ miles of full throttle though. If I mount it on the bike with good airflow it may improve cooling.

Before running it at 36V I opened up the controller and with the help of the guys on the endless sphere forum decided that the components were able to handle 36V. The MOSFETs were rated at 60V and main capacitors at 50V.

I'm going to try running the bike for longer distances and mounting the controller where it'll get a lot of airflow and see how it copes. If it does handle it long term then I'll have to buy a 36V Nimh battery for it as I doubt if a 12V SLA boost battery is a good long term solution.

Mind you, at 36V and 20-25mph I am going to have to sort out the brakes as they are not suited to stopping the bike from those sort of speeds.

Paul

 

[flecc]

Sounds good Paul. That sort of temperature isn't unusual on a 36 volt motor, so it will probably be ok. It seems your system is a mainly 36 volt one running as a 24 volt, given that you've also found the twistgrip/meter was 36 volt.

You won't be able to play these sort of tricks with a Torq if you still buy one, 3 volts extra and it signals overvoltage and refuses to run!
.

 

[coops]

The clues to the Torq-radical project were there for all to see... all seems so obvious now

EDIT: More accurate to say, he told us (the battery part), gave it away, and still we didn't twig

How's your bike doing now Paul? Still running well? Hope your brakes are "up to speed"

Stuart.

 

[flecc]

The clues to the Torq-radical project were there for all to see... all seems so obvious now

Stuart.

That's right Stuart, I left many clues littered all over the place, but to my amusement not one was picked up. I did the same with the Q bike as well.

For example there was the person who asked about multi sprocket freewheels and I answered with every minute detail, right down to the millimetre separations of the sprockets and the exact overall widths. The clue was there, I had to have one in my hands!
.

 

[coops]

yes, with the Q bike it was only when you mentioned how quickly, from repetition, you could remove & replace the Quando rear wheel, that the project "subject" became apparent: with the T we knew all along the bike being adjusted, and still couldn't guess the method!

 

[allotmenteer]

How's your bike doing now Paul? Still running well? Hope your brakes are "up to speed"

Stuart.

I haven't had much time to use it over the last week or two. What with DIY in the kitchen (nearly finished ) and then every other spare moment at the allotment, harvesting strawberries (15kg+ over the last 2 weeks ) and potatoes mainly, I just haven't got enough hours in the day. My parents are on holiday leaving me to tend the allotment alone. Serves them right if there's no strawberries left when they get home!

I have finally put a better wheel on the front, I had to true it though and also adjust the bearings as they were loose + change over the tyre from the old wheel. With the front brake adjusted as best as I can it's still pretty rubbish. This is due to
1) the pads being c**p quality
2) The v-brake lever arms being made of plastic and thus flexing under load
3) The actual brake levers being rubbish too

No doubt the cable is made of elastic instead of metal too I honestly don't feel safe going much more than 10-12 mph on the bike as I really can't stop it in a hurry.

I've also mounted the controller (temporarily with straps) out in the open to aid cooling and after a quick spin tonight it barely got warm after 1.5 miles.

I've got a list of things to do to the bike to get it working to my satisfaction. Briefly they are:

1) Put a CPU heatsink on the controller case to further aid cooling.
2) Replace the front large chainring with a 48 tooth one.
3) Weld / braze a bracket on the downtube for the battery.
4) Weld a mounting plate on just behind the seat tube for the controller.
5) Shorten all the cables as much as possible and tidy them up / hide them.
6) Get a 3mm spacer to replace the 10mm one on the non-drive side of the hub to shift the hub over and improve chainline.
7) Get the wheel rebuilt and dished properly.
8) Put a better rear sprocket on - with a 12 tooth smallest cog.
9) Fit some decent v-brakes to it.

There'll probably be more when I think of it but it's enough for now. In the meantime I aim to keep it on the road and make the changes incrementally. The batteries can stay in the panniers for a while yet but the controller has properly cooled asap.

I'll take some pictures and do a write up when I'm done as others may find it helpful and even be inspired by an amateur's attempt to improve his ebike.

Paul

 

[coops]

Thanks for the full update Paul, sounds like you're busy then?! For a self-confessed "no time for tinkering with kits" kind of person, you are certainly getting your share of that done! I hope its giving you valuable hands on experience and look forward to hearing the details, shuld you ever find time to post them

Stuart.

 

[allotmenteer]

Thanks for the full update Paul, sounds like you're busy then?! For a self-confessed "no time for tinkering with kits" kind of person, you are certainly getting your share of that done!
Stuart.

I didn't intend tinkering with it again but those spare batteries were just sitting there staring at me and I thought I'd have to try it at 36V, it'd be mad not to. I think the main problem is I'm not very busy at work and spend more time than I should browsing ebike forums like this one and endless sphere and they're encouraging / corrupting me. Fleccs Q-bike projects have also been an inspiration to me.
I don't have a lot of spare time but I find that I rarely watch TV these days so that helps.

Anyway, I didn't specify how long the alterations are going to take. My kitchen is only nearing completion and I started that in Jan 2006

Paul

 

[coops]

browsing ebike forums like this one and endless sphere and they're encouraging / corrupting me. Fleccs Q-bike projects have also been an inspiration to me.

Yes, same here totally: good to hear you're getting good use of your bike. My latest project "idea" (thats the pre-"concept" stage ) Did you see this one on your forum visits? I think it looks promising.

P.S. By the way, you said before your hub motor resembled the Q motor, does it look anything like the puma in design & ~6:1 gearing, and do you get that sort of performance from yours @ 36V now?! (well done for discovering that one!)

Stuart.

 

[allotmenteer]

Yes, same here totally: good to hear you're getting good use of your bike. My latest project "idea" (thats the pre-"concept" stage) Did you see this one on your forum visits? I think it looks promising.

P.S. By the way, you said before your hub motor resembled the Q motor, does it look anything like the puma in design & ~6:1 gearing, and do you get that sort of performance from yours @ 36V now?! (well done for discovering that one!)

Stuart.

I'm having trouble keeping up with all the posts at the moment, there's so many (a good thing though). I've read about the puma motor before, knoxie is using them. Knoxie is a British guy on the endless sphere forum (and was / is on the visforvoltage forum). Knoxie is heavily into powerful motors and large voltages - 72V He's made quite a few videos of his bikes which are worth watching. Anyway, Knoxie has said that the puma motor has some serious torque and such a beast of a motor has to use proper torque arms or it'll rip your dropouts off

I've taken a few pictures of my motor if anyone is interested.

As for the gear reduction I have the following information:
The main shaft of the motor has 22 teeth (as best as I can count).
The 3 planetary gears have 36 teeth each.
The hub casing has 93 (odd number that ) teeth.

Is this a 3.875:1 reduction? ( (33/22) * (93/36) ). It doesn't sound right does it?

Paul

 

[flecc]

No, the planetary wheels are irrelevant to the measurement Paul, treat them as idlers which they are. Its just the first and last that count, 22 to 93 which is 4.227 to 1.

That's high, no wonder it's fast!
.