Giant lafree twist electrical breakdown

My Lafree Twist has died again for the fourth time. When I turn on the switch the red lights comes on for a few seconds then goes off again, indicating all is well, but as soon as I apply pressure to the pedals - nothing, no motor activity at all. This seems to happen when undue stress is applied to the motor - this time I was struggling in bottom gear up a very steep hill when suddenly the motor cut out. I got off and pushed it a few hundred yards, but it failed to start again and I ended up pedalling back manually for 8 miles. Turning the motor on and off again, removing the battery and swapping it for my wife's made no difference. At the time of failure I had three blobs left on the battery. When I got back there was only one blob flashing, despite the motor having failed to do any work.

Now this has happened before and the motor has recovered eventually - one time it wouldn't come on again for a week so I took it to the bike shop, who then phoned up and said "what's supposed to be wrong? its working normally!", as indeed it was after a trip in the car. When I phone Giant they said if the motor cut out, then pedalling backwards a few turns would reset it, but it doesn't. It implies there is some sort of internal cut out in the motor unit, but if so, where is it and how do I reset it? If it were a thermal cut-out it would come back within an hour, but this time it's been three weeks and it still doesn't work. If it were a loose contact then it would surely cut out at random over bumps, but it doesn't do that.

I have the technical manual but it doesn't go any further than exposing the motor unit and replacing it. I was surprised to see such a complicated circuit board hidden within the unit on flecc's stripdown page, which poses the question - what does the circuitry do?

So, I have dead bike -does anyone have any clues as to what's wrong?

There is a thermal cutout function, but it's not apparent as a sensor. It's almost certainly a function of the driver circuit of the power output transistors which you see on the heatsink at the foot of the webpage you refer to Alan.

 

The reason I believe that's the case is Panasonic's background, this being the way thermal cutouts operate on class B amplifiers, which are very much nearer to Panasonic's normal design field. Forgive me for the following technical description, but there's no other way to say it. In the past the cutout circuit was a drain to earth of the bases of the two driver transistors, cutting their drive to the output transistor bases. As such it could be repaired by component replacement. Now, most of that main circuit board consists of miniature surface mount ICs (chips) which aren't possible to trace and diagnose, it's a board replacement if that's the fault, or whole unit replacement, depending on Giant's policy.

 

As for what it all does, as well as the drivers for the output transistors, the circuit organises the power functions according to the readouts of the pedelec sensors and pedal pressures. It's also just possible that a faulty pedelec sensor is involved, since a failure to signal pedalling would cut the power of course. That's what the backpedalling should have reset.

 

That will involve opening the unit and stripping the main pedal shaft out to gain access to the sensor cluster for repair or replacement.

.

Edited February 17, 200719 yr by flecc

Battery terminal condition?

 

Before going down the path of investigating a possible fault within the Panasonic power unit (which I've NEVER heard of happening before), it would pay to ensure the spring terminals on the bottom of the battery are pressing fully against the pins in the battery holder. The spring terminals should not be bent back at all when the battery is removed from the bike. I had this problem and fixed it by bending the spring terminals back flat again and dressing the contact areas. Possibly your wifes battery also has bent spring terminals - hence no difference when fitting that.

Wouldn't have thought it was that as, when I had dodgy terminals, I got not lights at all. In this case the red light is coming on initially. This suggest power.

 

Sav.

a possible fault within the Panasonic power unit (which I've NEVER heard of happening before)

 

It is known. Before the upmarket models were produced, a large batch of Twists had an odd fault which caused both switch positions to be Eco, the fault internal to the unit. With no resources to find the solution in Giant Holland or the UK, they sent complete new motor units out to dealers with instructions to swap them over and bin the old ones. A shocking waste, but that was what happened.

.

Wouldn't have thought it was that as, when I had dodgy terminals, I got not lights at all. In this case the red light is coming on initially. This suggest power.

I had both happen with mine (no red light at all and a couple of times a red light indicating power and then when I started of it died shortly after). This could be explained by the increased current drawn causing an already weak connection (if the spring terminals aren't properly pressing onto the holder pins) to form a localised hotspot which burns an insulating film into the joint. Dressing the terminals was important to get rid of any traces of this burning. This problem seemed to start after I was engulfed by a wave splashing over the wall on a cycling and walking path. Although I washed the bike when I got it home, I did use it to get home under power. Fortunately the main connection blocks are covered inside little plastic bags but I guess some salt water got into the unsealed battery connection area.

 

I haven't had a repeat of the problem since I cleaned and straightened the spring terminals. Having said that I'd not heard of what flecc reported. That may be why they introduced revised powerplants in later model Twists. Thanks flecc for the heads up.

Thanks for suggestions

 

Thanks for all suggestions so far - I did take a mild abrasive to the battery terminals as a first shot, and I've started stripping the motor unit down. I'll check the voltage is getting as far as the motor, and the switch is working as it should, with a meter at the weekend -its too cold to go out to the garage in the evening! I'll report back.

I did take a mild abrasive to the battery terminals as a first shot

And did you also check that the spring terminals in the battery aren't bent inwards at all? Even the slightest deformity in the case of my battery made a big difference and its not that obvious.

Stiil no joy

 

I've just spent a couple of hours checking all the wires to the hand control and the motor and comparing them to my wife's identical bike. Surprisingly they did not correspond the the readings I should have got according to the technical manual, but as they were mostly in agreement, and my wife's bike is working normally, I think I can safely ignore the differences with the manual.

The hand control readings to both balck and white connector were identical. The readings on the white connector going to the motor were different however. What I got was:

Yellow and black wires -open circuit on both bikes (says short circuit in manual)

Green and black wires- open on both (says short in manual)

Green and yellow wires- short on wife's bike, but open circuit on mine (says short in manual)

Battery terminals were cleaned and bent forward a a little - red light still coming on and then off again on both normal and eco, but still no life from motor.

Any more ideas?

In all three cases you have open circuit where it should be short circuit Alan, so a simple loss of connection within the motor casing is a possibility. I'd suggest you remove the sprockets and open it up to check the internal connections to the circuit board, unplugging and replugging to remake the connections. Likewise check the pedal sensor white connector security at the crank spindle behind the crank freewheel pawls. Beware of touching the circuit board components while doing this, and preferably earth yourself to discharge any static.

 

You could check the readings again while the case is still open.

 

Let me know if you'd like any additional advice on opening up the casing. It is quite straightforward once the unit is off the bike.

.

Thanks again. I've now successfully removed the motor unit. I can see five hex headed screws I need to remove, but first -do I need to remove the pedal cranks and if so how? I've taken the plastic covers off and removed the nut but they still seem firmly attached. Do they just need tapping with ahmmer, or ids there something else holding them on?

Also do the two gear wheels need removing? I have some circlip pliers, but I don't have a gear remover - do I need one?

For the cranks you need a crank extractor tool Alan. Good bike shops have them. It's basically an outer sleeve with an external thread for the crank and an inner with another thread which drives against the spindle. The outer is wound into the crank thread first, then the inner tightened with it's tommy bar or a socket driver, they differ a bit. That pulls the cranks off the taper fit.

 

The sprockets come off easily once the circlips are removed, no problems there, just rock them slightly to loosen the bond.

.

Screws out, circlips of and case partially apart, but I can't get any further -looks like I need a cotterless crank remover, so I'm off to see if I can buy one

Judging by the times, looks like our posts crossed Alan. No problem since you've picked up on what's needed.

.

I've been and bought a crank remover - good job I took the motor with me, as the first one I picked up, (fits any modern bike I was told), didn't fit, and I had to exchange it for a more expensive Park one. So now the cranks are off, the casing is open and .... there's nothing apparently wrong. No loose wires or burnt out components, but with all the wires being soldered to the circuit board and then the whole board being covered in a layer of silicone rubber, there's no way I can check the continuity of the wires. The sensor board is the only part that isn't encapsulated, but again it looks fine physically, so I can't see there's anywhere else to go other than replace the whole motor unit. Anyone know where I can get one for less than £450?

Re-reading your initial post Alan, I'm wondering if it's the electric motor itself which is the problem, since you mentioned it happening when under high stress.

 

It could be worth unbolting that and disconnecting the three "faston" connectors and checking the continuity through the phase wirings. It's best to leave the alloy case open to do that, since you need access from both sides.

 

Let me know what you find and I'll do comparison readings at this end.

.

Edited February 24, 200719 yr by flecc

I've just measured the resistance between the three terminals, and it shows zero ohms between any two terminals. I'll be interested to see what you get.

I've got the same Alan, there should be continuity through all three phase windings, so it appears your motor is ok. So the fault you have is either the main circuit board or the pedelec sensor unit on the crank spindle.

 

As ever Panasonic have varnished in the white connector for the pedelec sensor unit at the mainboard, but I found by prising at the back and side areas between the top and surround I managed to get it unplugged without damaging the main board. See if you can achieve the same and post back, as there's a further avenue for you to try.

 

P.S. It was only a bit of the silicone on the connector which has peeled off, so yours may be easier to unplug.

Edited February 25, 200719 yr by flecc

More readings

 

I think I've been struggling to get the wrong connector off! There are two white connectors that lifted off easily -no varnish on them -the smaller one goes to the crank sensor (three wires) and the other to the motor sensor (5 wires). The one I was trying to prise off without success goes to the finger switch on the handlebars. Here are the readings I got (and the last two digits on my multimeter are playing up, so these may not be exact):

Crank Sensor:

White - Blue 150 ohms

Blue - Yellow 75ohms

White -Yellow -75 ohms

Motor Sensor:

-ve probe of meter on black:

Black - each colour - open circuit

+ve probe on black wire:

Black - Green 4.4K

Black -White- 4.4K

Black -Blue 4.4K

Black -Red 3.8K

+ve probe to Green wire

Green to white blue and Red -all open circuit

-ve probe to Green wire:

Green -White open

Green - Blue open

Green - Red 13K

+ve probe to white wire:

White - Blue open

White - Red open

-ve probe to White wire

White - Blue open

White- Red 13K

+ve probe to Blue wire:

Blue -Red open

-ve probe - Blue Wire

Blue - Red 13K

I hadn't taken the motor sensor readings before, but have done so now. My readings are with the sensor board and motor in place to ensure the correct magnetic conditions. There are some discrepancies which I've shown in red against your readings below, and it seems the motor sensor board is what's faulty if your readings on that meter are ok. (I'm getting megOhms where you have kOhms in some cases)

 

Crank Sensor:

White - Blue 150 ohms

Blue - Yellow 75ohms

White -Yellow -75 ohms

Motor Sensor:

-ve probe of meter on black:

Black - each colour - open circuit

+ve probe on black wire:

Black - Green 4.4K 5.6M

Black -White- 4.4K 5.6M

Black -Blue 4.4K 5.6M

Black -Red 3.8K 4M

+ve probe to Green wire

Green to white blue and Red -all open circuit

-ve probe to Green wire:

Green -White open

Green - Blue open

Green - Red 13K 21.9M

+ve probe to white wire:

White - Blue open

White - Red open

-ve probe to White wire

White - Blue open

White- Red 13K 21.9M

+ve probe to Blue wire:

Blue -Red open

-ve probe - Blue Wire

Blue - Red 13K 21.9M

 

If you double check and are sure that's the answer, mail me an address and I'll post a replacement motor sensor board to you to try out.

 

I'm getting almost identical Crank Sensor readings, so that would indicate your meter probably is ok.

.

Edited February 25, 200719 yr by flecc

I'll try and get myself a new meter in the next couple of days, then I'll let you know for sure, but many thanks for the offer. That little circuit board only has three active components -three identical chips labelled 410 WHO, so in theory it should be possible to replace them -it would take a fine soldering iron, but at least the board is only one sided. I wonder what they are and how they detect the motor speed?

I'm sure they're Hall effect sensors Alan. They sense the rotational position of the motor magnets and trigger the timing of the phase current pulses. Whether that 410 WHO will be sufficient to identify suitable replacements is open to question, I'll have a look.

 

It's usually not worth the bother though, and you're welcome to the new board which at least you'll be sure of. I'm here all of this coming week but will be away from home for some of the following week, so that week might cause a delay in sending it if needed.

 

P.S. I've drawn a blank on the 410 WHO search as I thought, not recognised at RS Components. Lot's of the stuff Panasonic use in these units isn't available here as I've found in the past. Much of it is "in house" manufacture in that Matsushita group.

 

If you look at this page on my Torq site, you'll see an animation of the operation. The Hall sensors are numbered H1, H2 and H3 within the motor, but in the Twist case, the sensors are external and two at least must have sequence switching and a delay line to permit correct operation. This sort of complication is typical of this unit's design.

.

Edited February 25, 200719 yr by flecc

Hall sensor board in the post to Alan now, hopefully will do the job.

.

Many thanks for getting that sensor to me so quickly flecc. I've just spent the morning reassembling the motor, but I'm afraid to say its made no difference. I've just been out for a test ride and although the red light comes on as before, there's no output from the motor, which I guess means its a faulty main circuit board. I'll try pleading with Giant on Monday.

Alan

Ok Alan, sorry to hear that. There is still the possibility that it's the pedelec sensor unit at fault, especially as it occurred when you were putting the maximum effort onto that.

 

There are three leads into there, and the motion sensor is a multi element type with about 40 sensing pins, but there must also be a strain gauge on the spindle for load sensing. It's a sealed unit though.

 

That can be unplugged and the whole crank spindle assembly removed from the left hand housing complete with it's bearing which is a force fit on the spindle. It wouldn't be worth struggling to get that off, but if you'd like me to mail you the entire spindle assembly with pedelec sensor and bearing to try, you're very welcome. Just say the word.

.