How the windings are connected. There are 3 phases, or sets of coils . Lets call them A, B , And C . So a third of the coils belong to each phase . Suppose we trace one input wire, and trace which coil it enters . Call That coil 1 and suppose it is the first coil of phaseA . current then exits this coil 1, misses two coils and enters coil 4 , misses two more coils and enters coil 7 . Suppose , during winding that only phase A has been completed. So we get a wound coil, two bare coil cores , a coil;, 2 bare cores and so on, until we get back around the motor to near where we started, We wind the first coil of phase B next door to coil1 and wind every third coil, back around to the start point . Phase c , we just fill in the remaining gaps.

 

So now every phase has a beginning wire and an end wire. The beginning wires all go to the controller . The end wires are all connected together . This method of interconnecting the phase is called"Y" connection . We can represent it diagramatically , by drawing each set of phase

s coils as just one long coil . So draw a letter Y with coils instead of lines.

 

A less common way to interconnect the phases is called Delta. Draw an equilateral triangle, but use coils instead of lines . Connect the wires from the controller to the points of the triangle. So each flexible input wire connects to TWO phase wires .

 

Whether Y or Delta an ohms reading between any 2 input wires should be the same . Draw the diagrams . Hope this helps.

Edited June 14, 201213 yr by neptune

How the windings are connected. There are 3 phases, or sets of coils . Lets call them A, B , And C . So a third of the coils belong to each phase . Suppose we trace one input wire, and trace which coil it enters . Call That coil 1 and suppose it is the first coil of phaseA . current then exits this coil 1, misses two coils and enters coil 4 , misses two more coils and enters coil 7 . Suppose , during winding that only phase A has been completed. So we get a wound coil, two bare coil cores , a coil;, 2 bare cores and so on, until we get back around the motor to near where we started, We wind the first coil of phase B next door to coil1 and wind every third coil, back around to the start point . Phase c , we just fill in the remaining gaps.

 

So now every phase has a beginning wire and an end wire. The beginning wires all go to the controller . The end wires are all connected together . This method of interconnecting the phase is called"Y" connection . We can represent it diagramatically , by drawing each set of phase

s coils as just one long coil . So draw a letter Y with coils instead of lines.

 

A less common way to interconnect the phases is called Delta. Draw an equilateral triangle, but use coils instead of lines . Connect the wires from the controller to the points of the triangle. So each flexible input wire connects to TWO phase wires .

 

Whether Y or Delta an ohms reading between any 2 input wires should be the same . Draw the diagrams . Hope this helps.

 

Wow! Thanks Neptune. What a great explanation that is. I had found the information elsewhere, but yours is SO much easer to digest. I'm going to dive in with my multimeter shortly. Fingers crossed that all is well with the motor. From what I understand, these motors are pretty robust and there is very little to go wrong with them.

Glad you could understand my explanation. There is an additional test to do . Put your multimeter on ohms.If it has several ohms ranges, choose the highest one. Put one meter probe on any phase wire. Put the other probe on the metal outer case of the motor . The meter should give a very high reading, exactly the same reading as it gives when the probes are connected to nothing. If it passes this test, there are no earth faults on the windings.

Neptune, I've looked at the windings up close and your explanation makes sense - there are 18 individual windings, so that means 6 sets of 3 as per the sequence in your explanation. What is less clear is where I'm meant to probe. Do I need to remove the insulation of the input wires or can I test from the end of the cable protruding from the axle. How about the output wires? I'm not even sure I can see those. Do I have to further de-construct the motor to access them?

 

As a new member of this forum, I'm overly aware at the amount of leeching I've done so far. I have requested much advice and help and really feel I need to follow this through to the end or I'll let you all down

 

As offered earlier, I can also make a video of the problem occurring if you think it would help.

 

Thanks so much everybody.

Glad you could understand my explanation. There is an additional test to do . Put your multimeter on ohms.If it has several ohms ranges, choose the highest one. Put one meter probe on any phase wire. Put the other probe on the metal outer case of the motor . The meter should give a very high reading, exactly the same reading as it gives when the probes are connected to nothing. If it passes this test, there are no earth faults on the windings.

 

I have tried this and can report no faults as described in your description (multimeter reads infinitely high).

No need to remove any insulation . All tests can be made from the ends of the wires protruding from the axle. A video might help. Do not hesitate to ask for help . We shall have helped you enough only when you ride into the sunset sporting a big grin.

Right, I think I've got what is good news. I tried something that I know I shouldn't have, but did it anyway.

 

Without having the wheel mounted to anything, and with the hub cover off, I powered it up and tried a quick blip on the throttle. The central core of the hub (can't remember the propper name) spins freely, smoothly and fast, without any of the clicking heard previously. It sounded like an electric motor should

I only blipped the throttle for a fraction of a second and pre-wound the motor cable to minimise stress to the cable. It only span a few times, so no harm done.

 

This means that the motor's good right? Why does it not operate like this once mounted?

Realising that I was probably pushing my luck, I tried the spin test again while holding the axle tight in my hand against the floor for longer and the WHEEL spins great. Throttling up slowly, I ran it for about a minute at full speed and it's pretty smooth. There is a different sort of clunking noise, but I think it's due to the wheel wobbling and the bearings not being loaded.

 

This thing looks like it should propel a bike pretty well.

 

Need I continue with the motor winding test now? Is this possibly an issue with the controller?

WTF? IT WORKS!!!!!!!! I feel like I've just invented the wheel

 

Following my last test, I reattached the wheel to the frame, powered it up and to my surprise, all is as it should be. Could there be a loose connection in the motor somewhere?

 

Hopefully my pannier rack arrives tomorrow and I can get this thing wired up properly. Can't bloody wait!

 

EDIT: Also, I just realised that I had the controller in 'restricted mode'. De-restricted, this thing looks like it might take off!

Edited June 14, 201213 yr by danielrlee

Excellent news - Well done!

 

It could have been a bad connection somewhere, or possibly a clutch problem.

Whatever, fingers crossed it doesn't happen again.

Having been woken from a gratuitous lay-in by the postie this morning, I now have the pannier rack in my hands, so full steam ahead today.

 

I've had time to contemplate the source of the problem I experienced with the hub motor. Could it have been caused by over-tightening the axle nuts?

@danielrlee. I was thinking about your problems last night. I would say that the problem is unlikely to be elecrical. As has been said it could be a clutch , bearing or gear issue. If you feel competent to do so , you could haver a look inside the other side of the hub, and clean and grease things. On the other hand if it is running OK now , leave it alone for now, and do some road tests. Check that when fitted to the bike, and the wheel nuts are tightened that it still spins freely by hand , and that no part of the wheel or hub is fouling the frame. Good luck and keep us informed. Seems that you will ride grinning into the sunset soon after all.

@danielrlee. I was thinking about your problems last night. I would say that the problem is unlikely to be elecrical. As has been said it could be a clutch , bearing or gear issue. If you feel competent to do so , you could haver a look inside the other side of the hub, and clean and grease things. On the other hand if it is running OK now , leave it alone for now, and do some road tests. Check that when fitted to the bike, and the wheel nuts are tightened that it still spins freely by hand , and that no part of the wheel or hub is fouling the frame. Good luck and keep us informed. Seems that you will ride grinning into the sunset soon after all.

 

I'll soon be smiling about having everything operational, but I'm not sure about that sunset - the weather looks like it's going to be atrocious this weekend.

 

I agree. The motor is probably best left be until I experience another problem with it. It's strange because when I was experiencing the motor issue, the wheel did spin freely both by hand and by turning the crank. I did apply quite a lot of torque to the axle nuts, but I can't imagine it would have affected anything within the motor casing. It's working now and that's the main thing. I just wish I knew what the problem was.

 

I've now swapped the old handle bar grips with the new throttle grips and secured the wiring with zip-ties and velcro straps where appropriate. I've fitted the pannier rack, but am now having difficulties working on the bike when standing the right way up since I don't have a bike stand. I've got a kickstand on order that should be here tomorrow, so that should make things easier.

 

I'm not going to fit the PAS sensor yet, since I can potentially see myself removing the front derailleur and switching to a single chainwheel - more discussion on that later though.

 

That just leaves fitting the batteries and controller for now. The bag I've bought is a little smaller than anticipated. It fits four 7ah 12v SLA batteries perfectly, but doesn't leave any room for the controller in the main compartment unless I lay it on top of the batteries. I'd rather not do this though since it could easily cause a short. I'm not sure whether I should install the controller permanently attached to the frame, making removing the batteries easier, or house this in the bag in maybe another compartment. I don't plan to ride much in the rain if I can help it, but I am concious of leaving the controller exposed to the elements. I know it is weatherproofed to a degree, so should I be concerned?

Hey, glad about the progress that you are making. I usually hang my bikes from a beam in the garage to work on them. Not ideal but it works for me. If your seat is usually adjuster to a high position, you may be able to mount the controller on the seat pillar, plenty of ventilation, and a bit of shelter from rain. Just attach if with cable ties for now. You can finalise such details later , but for now, get it together and ride it.

Finally, I had a chance to test drive the bike today. As a new experience, there was so much for me to take in. I didn't have my GPS logger enabled, but under motor power only, it managed about 15 on the flats and 8 on a moderate incline. I have somehow managed to screw up my brakes (one pad on front and rear not returning to unsprung position) and the tripple chainwheel is almost unusable, due to the amount I had to spread the dropouts. It is pretty heavy with 4 7ah SLA batteries on the rear rack (yet to weigh the whole setup) and a bit unbalanced with more weight on the rear. Even with all the faults though, it was great fun....... until the motor stopped working again. I managed about 2 miles and then rather randomly, lost power.

 

Luckily, I had the foresight to not wander too far from home, so didn't have far to cycle back. I can definitely say that the motor is of great assistance (esp. uphill), since without it, I was absolutely knackered by the time I got back.

 

I'm going to try running the wheel off the frame again and see if I can replicate my previous (albeit temporary) fix. I'll also try and post some photos this evening. I'm really hoping I can work out what the fault is, since that brief taste of e-biking goodness has whetted my appetite for plenty more.

Edited June 16, 201213 yr by danielrlee

@danielrlee. At least you got a tryout and had a chance to see the potential. A good idea for road testing is to find a local test route whereby yo never go more than half a mile from home. Regarding battery weight . If most journeys are local, and the batteries you have are in fair condition, you may find that 2 batteries are enough. Carry the extra 2 when you need them . Consider making a battery box to fit within tha frame triangle to hold 2 or even 4 batteries. I would be interested to know what battery fuse you are using now, and if it blew when the fault occurred. If the fault is stalling the motor it will likely draw more amps than usual. Next time you get it running on the road, fit a suitable amp meter in one of the battery leads, and then you will know what normal is. If the fault then recurrs ,you can tell if it draws more current or less.

Once again, the motor runs well off of the frame. I decided to deconstruct the motor a little and removed the central core from the hub casing. I did note that the securing nut for the axle was very loose, although not undone and wondered if this was the cause of the problem (maybe introducing a small bit of play and causing something to bind).

 

Having taken a look at the flywheel and gearing, I can say that everything looks okay. The internals are pristine and the gears have pretty much zero wear (to my eyes and fingers anyway) with well defined, even teeth. There is lots of lube, the flywheel is true and nothing binds. Everything moves very smoothly.

 

I have reassembled the motor and now tightened the securing nut propperly. I'm going to try the motor mounted to the bike again and see if the problem re-appears.

If you get to the stage where the motor will run off the bike but not in the frame, Try to see if it runs with ONE wheel nut tight but not the other. Try each in turn. [careful with that throttle.] It might point you to exactly where the problem is.

Alas, after about a mile the motor gave up on me again.

 

@neptune I have have been running the motor with a 10A fuse and so far it has remained intact.

 

Once more, I unmounted the wheel, powered it up, and this time nothing. I popped off the motor cover. Still nothing. I then realised that this is actually a good thing, since I can now try prodding and poking one thing at a time to test for open connections. After a while, I prodded all of the black shrink tubing under the netting in the photo below, tried the throttle and it was working again. I have found the problem in the form of a bad connection on one of the windings:

 

[ATTACH]3760.vB[/ATTACH]

 

I re-mounted the wheel, tested the motor and it still worked. To verify that I had found the fault, I disconnected one of the motor phases and the motor runs as it does when faulty. JACKPOT!!!

 

Whichever one it was, I gave the faulty connection quite a prod, so maybe with a bit of luck any open connection has now been closed for good. If not, at least I know how to progress.

 

All in all, a successful afternoons troubleshooting. Time for a brew now.... then maybe I'll fix those brakes

Edited June 16, 201213 yr by danielrlee

@neptune It just occurred to me that if I'd managed to test the windings with a multimeter as you said, I might have identified the problem sooner. An intermittent problem can be so difficult to diagnose at times though.

 

Thank you EVERYONE for your help so far. This is such an awesome forum and has been key to the progress I have made. Taking on a DIY project can be extremely challenging, but ultimately rewarding and I already feel like I might already be able to give something back and help others in a similar situation.

 

Miles of smiles to come.

@danielrlee. You are making great progress, both physically, and dare I say it mentally.The fact that the fuse is still intact points to an open circuit fault rather than a short circuit fault. If and when the fault returns, proceed as follows. Do not fight shy of using your multimeter, it is unbelievably easy, and all tests will be made on the ends of the wires emerging from the axle. Set your multimeter on its lowest Ohms scale .I do not know if your multimeter is digital, or analogue[having a scale and a needle] . Either will do . Some meters have a built in buzzer, which buzzes when the meter detects a low or zero ohms reading . If yours does not , you will have to keep one eye on the meter . So, with the fault on, proceed as follows .

Label the three wires A, B, and C in any order at random. Test resistance between A and B . It should be zero or a very low value. Test between A and C . Reading should be the same . Test between B and C, reading should be the same . With the fault on , one of the above 3 readings will be a very high value or infinity. Let us suppose it is the test between A and B . Using crocodile clips, or similar , leave the meter connected to A and B . Now, whilst watching the meter, or listening to the buzzer, poke those connections one at a time . Your meter will tell you when you poke the dodgy one .

You will probably have to remake the joint . Before you cut the lacing, what you call the net, check if it has been varnished to stick it to the wires . If it has , you can get away with cutting it without it all unravelling . You will have to cut away the sleeving to get at the joint . After repair, if there is not enough room to resleeve it , insulate it by painting with nail varnish, but do not splash it about , in case it reacts with the normal varnish on the wires [this is very unlikely, but not impossible]. If space permits you might need to add some new lacing to secure the repair . Check that any new lacing cannot foul moving parts . When resoldering the joint , use a large iron, and some fresh solder . Ask if you need further help . Looking at the photo again it looks like there are 4 joints in the wiring. 3 of these are where the flexible leads from the controller join onto the actual windings . The forth will be the Y point where the 3 sets of coils meet.[see earlier post]. I would bet the fault is most likely at this Y point, because as there are 3 wires in the joint it is physically harder to solder successfully.

Edited June 17, 201213 yr by neptune

@neptune Ahhhh! That's what the fourth piece of insulation is - the Y point! Seems so bleedin' obvious now it's been pointed out. I was under the impression that I had to test from the Y point to the ends of the wires, not end-to-end. That is easy. Also, great idea to use nail varnish to insulate where there is no room to resleeve - will definitely use that tip in the future. Thanks for all your brilliant explanations Neptune. Ever thought about going into teaching?

 

I took the bike out today for a good old road test and did about 12 miles on it. I still hadn't charged the battery since before I first got the motor running a couple of days ago, so adding that to the 3 miles I did yesterday, the battery has given me about 15 miles with liberal throttle usage and has only just started to feel like it's delivering less power. The voltage of the pack reads 24.5, which is at the 12.25 I like to recharge SLA's at. I could probably get another 5 miles out of them before the voltage drops below 12v, but doing that regularly would likely limit the life of them.

 

Great news - no more motor problems today I did have a couple of issues, but they were only minor ones.

 

The first was when the motor had a total loss of power (not even the jolting as when running on 2 phases). I thought I'd blown the fuse, but couldn't be bothered to check at the side of the road since I was only a few minutes from home. When I got home, I realised that the fuse was in fact fine and one of the battery leads (stupid spade connectors) had managed to work itself loose.

 

The second problem must have been rather funny to any onlookers. I obviously hadn't attached my pannier rack securely, because all of a sudden I heard a great thud and a loud scraping noise and looked around to see the rack and batteries being dragged along the road behind me. I scavenged a few minutes for some long grasses and used them to jerry-rig the rack back onto the frame until I got home to make a proper repair.

 

Slowly but surely, I am working out the remaining problems and the bike is becoming more reliable. Shouldn't be too long now before I attempt my first 12 mile round trip commute to work.

@danielrlee.Boy is my face red. You praise my description , and then I realised I made a boo-boo. I can not modify my last post. Everything is right as regards the method of testing, and the test results with no fault condition .

However, suppose the fault is on, and is as we suspect a break in one of the strings of coils [phases] between its wire coming out of the axle and the Y point , or even at the Y point. What we shall find then, is that the multimeter shows a low ohms reading between just 2 of the wires, and the third wire will not give a low reading to either of the first two . So clip one lead of the meter to this third [dead wire] and the other meter lead to either of the other two. Then do your poking about whilst watching/listening to the meter . The repair proceedure is the same of course . Just pleased I got a chance to put this right before you [might] want to use it! You will understand better if you draw the diagram, as I should have done . Just draw a large letter Y , but instead of drawing it with straight lines, draw it with curly lines to represent the coil strings or phases

@neptune The funny thing is, I understood your explanation as you actually meant. As soon as you told me to test the windings in pairs, end-to-end, the method seemed pretty obvious - as with any troubleshooting, it's just a systematic process of elimination.

 

Before I switched to my current career as a photographer a few years back, I had done a 10 year stint in IT support and management, where fault finding was a daily occurrence. I have always found that technical troubleshooting is a skill that transfers well to many things, including now it appears, building electric bikes.

I've had a couple of days to ride the bike properly and apart from a few problems all involving the battery, all is well.

 

I was intending to take some photos today, but following the latest problem, I have managed to rip the pannier rack from the frame, along with the controller. The battery bag has also taken a serious beating. No permanent damage done, but extremely annoying nonetheless.

 

The source of my problems all stem from the batteries being too heavy to mount on the rear rack securely. I am now intending to mount them onto the frame triangle using a modified football boot bag, which seems a perfect size for the task. I could knock something up from plywood, but I don't want to add any unnecessary weight to the setup and want to retain the ability to easily remove the batteries since I live in a first floor flat and need to carry it up and down a flight of stairs.

 

I've just been to a couple of sports shops to find something suitable, but all I could find were garish bags, all covered with team liveries. I also tried a couple of cycle shops for a frame bag, but none stocked anything that comes close to my requirements. One thing I noticed when talking to the cycle shop staff, is their ignorance to electric bikes in general. None had any knowledge on the subject and the cycling purists pretty much sneered at the idea - seems like a missed business opportunity IMO. One lad in Halfords however, was quite interested in doing a conversion on the cheap himself and said to pop in with it I'm ever passing, since he'd love to see it.

 

I've also been giving some thought to both the PAS sensor and brake cut-outs. Even though I've already bought a PAS sensor, I'm not sure I like the idea of it now. I kind of like the fact that I can peddle independently of the motor. If I were to fit it, I'd also install a switch so I can enable it as necessary and to be honest, I just cant be bothered. I also can't see the point of brake cut-outs at the moment. My throttle has a cramp buster built into it and as soon as you reach for the brakes, the throttle naturally rolls off. Does anyone else see either of these as a necessity? Also, is the bike legal without a PAS sensor?