I was riding home this evening in the rain when I had "engine trouble". Because of the sudden loss of power and gear noise, I first suspected the planetary reduction gear, but that seemed less likely as I let off the accelerator and coasted on normally. After I pulled over to the side of the road, I tried the accelerator again and ... nothing. The bike didn't even budge.
This seemed a little odd to me because I was getting some acceleration when moving, although much less than normal. On an impulse, I tried again as I pushed the bike forward and the accelerator was working again...sort of. Speed was limited and with the disturbing noises coming from the back hub I wasn't going to go any significant distance. Clearly whatever had occurred has the potential to do even more damage to the motor and reduction gears, so I called for a rescue.
Thinking back to my induction motor theory, I recall that single phase induction motors require a starting circuit to start rotation of the rotor. If this is not done, rotation may be commenced by manually giving a slight turn to the rotor. That seems to be what's occuring now for me, so does that mean I've lost 2 of 3 phases? Is that more likely caused by a loose connection or a blown motor controller (IBGT)?
At this point there's no indication of a problem on the dash. Would the onboard diagnostics even know if an IBGT was blown?
(It's really a pity, since I was just about to roll over 10,000 miles.)
"At this point there's no indication of a problem on the dash. Would the onboard diagnostics even know if an IBGT was blown?"
Speculation here:
looking at Mik's pics of his motor control board, it seems 2 of the 3 phases pass through inductive pick ups. They might be used to measure motor currents. I think the bike would be aware of a lost phase
I think only one phase goes through the Hall sensor.
The other hall sensor measures the total current from/to the battery.
This information may be used entirely at your own risk.
There is always a way if there is no other way!
looks like one sensor around the DC after the fuse and two sensors for the 3 phase.
Hi guys, before going to the worst possible scenario, let's go to the most common. The problem described could be the throttle has gone out of alignment. A quick check can be made by turning the bike on, placing the 'Kill' switch in the 'OFF' position and squeezing the left or rear brake lever. The speedometer needle will move to where the throttle is presently set at, it should be on 43 km/h. Whilst holding the left lever, also squeeze the right lever and the needle will go to where it should be correctly set to. Adjustment is relatively easy to do if you know which end of a screwdriver to hold and have a 1.5 mm allen key. if the speedometer needle does go to 43 km/h, then the next culprit is the 'ENCODER' which I will get into later.
To recalibrate the throttle, leave the bike switched on as above, then begin by removing the bar end weight. Most can be removed by firmly holding the throttle handgrip to stop it rotating and grasping the bar end weight with your other hand and unscrewing it. If it is on there tight, wrap a heavy rag around it and use some slip-joint pliers to get a better grip to aid in loosening. VERY IMPORTANT - DO NOT use the stop to keep the throttle from rotating as the bits inside can get bent or broken. Inside you will see a buff coloured nylon piece with a slot and a grub screw, this is the throttle magnet. Place a flat blade screwdriver in the slot and using the 1.5 mm allen key, loosen the grub screw a turn or two. Use an elastic band, cable-tie or friend to hold the left brake lever so the speedometer needle is at the current setting. Then, holding the throttle in the 'REST' position, gently turn the throttle magnet and you should see the needle move. Adjust to the 43 km/h then turn lock the grub screw, be gentle as the threads are easy to strip. 'BLIP' the throttle a few times letting it come back to the REST positioin to check the setting. Try the bike to make sure it works properly then re-install the bar end weight. DO NOT put Loctite or similar on the threads as you will likely never get it back off again without damaging something!
If one does remove the throttle magnet, when re-installing turn it in until it just bottoms out, then turn it back ONE FULL TURN and begin calibration from that point. DO NOT adjust in more than 1/2 turn as the throttle could stick during use or damage the Throttle Position Sensor.
ENCODER -
The Hall Effect encoder reads the motor position, rotation and speed through a series of slots and holes sending the information to the Motor Controller. Occasionally thest very small holes and slots get some dust or debris in them or the mirror and readers on the sensor board get dirty. Dismantling and re-installing the Encoder components requires special alignment tools to position the bits absolutely correctly and the Scooterdiag programme is needed along with the CanBus reader to recalibrate it. IMPORTANT MESSAGE HERE - DO NOT TAKE THE ENCODER APART OR LOOSEN ANYTHING UNLESS YOU HAVE THE ABOVE EQUIPMENT!!! THE BIKE MAY NOT WORK AT ALL OR MAY RUN IN REVERSE!!!!
I do not recommend the following as a regular or routine proceedure but a bit of light housekeeping can be done without disturbing anything. Remove the Chrome cover on the left side of the swingarm followed by the ivory colored cover. BEFORE removing each cover use a natural hair (horse, deer, goat, whatever) brush to clear away the dust and grit which will be there. Then blow it off with CLEAN air (no oil residue) I recommend using the canned air used for computers and electronics unless you have an oiless compressor. Once the covers are removed you will see the Encoder, pinch the release clip on the wire harness and pull straight out. Inspect the wiring for any chafes or nicks, this includes where it goes into the swingarm. Repair any wires if damaged. Back to the Encoder... you will see the mirror part of the reader, easily identified as it is the part that covers part of the reluctor wheel (the bit with all those tiny holes in it) remove this with a pair of needle nose pliers gently squeesing on the flats on each side, pull it straight out. Inspect the mirror prism for scratches and blow off with compressed air and set to one side. Now using the brush gently clean the entire chamber avoiding the Encoder. A vacuum cleaner will aid in extracting the dust. Again using the compressed air blow everything off in cluding the Encoder. Visually inspect the reluctor wheel for any scratches or damage without removing it. Now spray some contact cleaner on the reader board and the reluctor wheel then blow it off again with compressed air. Using the needle nose pliers, hld the mirror prism, take note of the locating dowels and the holes in the reader then with careful alginment fit it straight on. BE VERY CAREFUL not to have the pliers slip and strike the reluctor wheel as damage may occur. Reconnect the wire harness. REMINDER - DID YOU CLEAN THE INSIDE OF THE COVERS??? then reinstall the gaskets and covers. Seal the wire harness grommet with a generous amount of RTV silicone sealant.
Cheers!!
Doug Townley
info.DTES [at] gmail.com
(Formerly Vectrix Europe Technical Supervisor)
One door closes, another door opens
Ooops..you are right!
This information may be used entirely at your own risk.
There is always a way if there is no other way!
Okay, I can rule out some issues at this point.
Throttle checks are fine. The diagnostics software shows no faults. (Low Bus Voltage, IGBT Fault, Throttle Fault, Heatsink Temp, CANbus, Phase Current, Bus Current, Motor Temp, and Capacitor Temp all seem fine.) I also seem to get the usual green boxes in Motor Control Status as I just drive back and forth in the garage (with a push now necessary to get started). I see only a few amps of bus current, but that's probably to be expected given the low speeds. Unfortunately, I can't really monitor for high temps without warming it up over a longer distance, and I'm reluctant to do that given the nasty sounds coming from the motor. (In the past months, I've received a few CAPHOT warnings, but only briefly when climbing steep hills at low charge.)
So my hypothesis at this point is that I've got a bad connection on one of the three power cables going from the motor controller to the motor. Losing one connection (e.g. A) would eliminate two phases (e.g. AB, AC), leaving one remaining (BC), right?
Has anyone had a bad connection on one of the three motor wiring cables? Are the connections at the motor or motor controller more prone to failure? Is it common to have a bad connection within the motor housing? What portions can I safely test without disconnecting the battery?
Hi, I have heard of hall censers coming loose and causing the need to turn the motor for the controller to know were the fields are ! the noise you hear ? could be the loose censer, grinding on The rotor? what does it sound like when It runs with the wheel off the ground? I don't think the problem is in the power connection ! LaTeR
thank GOD I wake up above ground !!!!
Since the noise only occurs when using the throttle at low speeds, I'm not inclined to think that it's a mechanical failure. Any loose sensor or damaged components in the rotor or reduction gear would cause grinding noises even when coasting, and I'm not seeing that. All seems fine until I attempt to accelerate.
The noise responds more like an engine running on one cylinder, where the unusual noise and shaking is more noticeable when you hit the gas. It's not really a grinding sound, but more like a vibration from something dramatically out of balance.
I have to agree with Doug. look at the simple bits first.
When the encoder is dirty or not aligned correctly it makes a loud "mechanical" noise and can cause a juddering.
Maybe, but I had my encoder cleaned last summer. While I'll still get encoder faults on occasion, they're now less frequent and never accompanied by any loud mechanical noise or juddering. (I will get a stutter from CANbus faults, but we've been assured by X Vectrix that those are harmless.)
Given the symptoms, it seems more likely that I've got what Mik had on the Phase C power connector of his Vectrux. He's got great photos on his repair job here, but if mine's anywhere near that bad I think I'll see if I can get the motor controller replaced under warranty first.
Mik, are my symptoms consistent with what you experienced when your Phase C connecter burnt out?
Hi again,
I notice you have the diagnostic software which, is a bonus for you. Have you checked the Encoder alignment? To do this the rear wheel must be clear of the ground and free to turn. Make a note of the setting before you start then do a number of readings making note of numbers. These numbers are actually degrees and should fall anywhere between 0 and 359. There is no number which is best or worst. The numbers collected should be no greater than plus or minus 10 degrees, ideally they should be 0 to 2 or 3 degrees. If they are all over the place and not in the parameters described then I would strongly suspect the Encoder which is ususally in need of the cleaning. It only takes 1 or 2 small particles of dust to mess it up. If you have the alignment tools and know how to use them, then by all means remove it completely for a thorugh inspection and cleaning.
A problem I have only seen twice was the reluctor wheel had become loose, the adhesive had let go, allowing the wheel to shift around ever so slightly. One other was a slightly bent wheel which I suspect had been damaged by someone poking around in there and not really sure of what they were doing.
I have never seen a Motor Controller cause the symptoms you describe but, that's not to say it could never happen! Normally any problems with the M/C is accompanied by a slight electrical burning smell which is most noticeable with the trunk open and sniffing around the seat latch area; without someone sitting on the bike as anyone watching may think you a little strange!!!! LOL
Cheers!
Doug Townley
info.DTES [at] gmail.com
(Formerly Vectrix Europe Technical Supervisor)
One door closes, another door opens
No, totally different.
As Winged Racer wrote, it's the smell that gives away a smoldering connection at the MC board. But it is not a slight smell, it stinks to high heaven! You could not miss it if your nose works at all, particularly just after riding.
The motor controller board was still smelly weeks after the smoldering!
And there was no other symptom whatsoever, just the smell.
I have made some simple adapters to align the encoder, it worked well for me. It's somewhere in the handbook, I think.
This information may be used entirely at your own risk.
There is always a way if there is no other way!
Hi Mik,
You are absolutely correct in the smell being rather fierce when the terminal does go pop! But, as the bike still moves I have seen the terminal just beginning to burn, that's why I said 'slight'. Also, most of the time the bike has been parked for a while allowing the terminal to cool and the stench to dissipate. I really don't think it is the M/C but I have been wrong before!!
Cheers!
Doug Townley
info.DTES [at] gmail.com
(Formerly Vectrix Europe Technical Supervisor)
One door closes, another door opens
I don't have a clue as to what the problem is, but...
A) I'm surprised a Encoder problem would not show a ENCODR warning on the dash.
B) The connection to one phase could be broken at a place where there is no plastic to burn and smell.
C) A Hall sensor (on the MC board, not in the motor) could malfunction and give wrong readings.
Anyway, I agree that cleaning the encoder disk is the first and least invasive thing to do to get to the bottom of it!
This information may be used entirely at your own risk.
There is always a way if there is no other way!
Well, I just inspected the encoder. There was a bit of black dust, but nothing like what Mik showed here. I didn't touch the mirror or disc, but I used a little (zero residue) TF solvent to clean off the components and wiped dust off the housing with a clean rag. There was no change.
Putting the bike up on blocks, I could turn the back wheel freely and nothing sounded loose inside. The throttle worked just as on the ground, requiring a little manual boost to get started. Once spinning, it sounded much smoother unloaded, with less of the disturbing vibrations.
The only encoder programming that I'm familiar with is the Encoder Mode button under Show Instruments. When I clicked this the autoprogram sequence started and the speedometer gauge started to rise to about 40 kph, but then it quit. There was a slight hum and occasionally the fan would start, but the rear wheel did not rotate. I'm guessing that isn't normal.
When you say I should collect numbers to manually check alignment, is that done from the Encoder Offset field on the Memory tab? I wasn't seeing much change when I hit Read while turning the wheel. How do you determine the allowable parameters?
I managed to only get one Encoder fault during all my trials on the blocks this evening, and I'm guessing that was because the TF solvent hadn't yet fully evaporated.
Mag
In the diagnostics click on the "AUTORUN" button near the "ENCODER" button. Make sure the KILL box is checked. This runs the motor in open loop mode. That is, it does not use any feedback from the encoder. It simply walks the stator field 360 degrees and the rotor (wheel) should follow. It is a slow rotation. If this does not produce a slow continuous rotation then the controller has lost control over one or more of the phases. This could be caused by a number of things on the MC but generally it is one of the transistors in the IGBT. Let us know what happens.
X X Vectrix
Sadly, the fan just cycles on and off four times when I do this. There's a slight twitch at first, as if it's trying to rotate the motor but can't get starting torque. I imagine if I give it a boost, I'd be able to get it to rotate, but it won't do anything on its own.
Well, so much for the easy fix. It's time to turn things over to the Vectrix authorized service tech to see if he can isolate the problem and obtain parts. If the problem is in a IGBT transistor, however, it seems strange that I'm not receiving an IGBT fault in the diagnostics.
Maybe I'll get lucky and it will just be a bad power cable or terminal.
Keep us informed, it is interesting to know how to diagnose this kind of problems. Hope you'll soon have your bike. Mine is at the authorised dealer checking why my battery is failing. mmhehehehehhe. We all know why all the batteries are failing, but without BMS the authorised dealer can do very very few...
Keep in mind that "authorized dealer" has very little meaning until agreements have been put in place with the new Vectrix. Here's what I heard this morning from Dana DeCosta:
Frankly, it seems completely understandable to me why those 100 old dealers might be reluctant to sign another dealer agreement with Vectrix, so you might want to contact your local dealer and lobby in support of reestablishing official ties.
Okay, it's long past time to close the loop on this thread. I gave up waiting for an "authorized" service rep and opened the bike up to get some eyes on the motor controller. It turns out that I've got a problem very similar to Mik's Phase C terminal burn-out. I've limited my inspection so far to what I can see without removing the back battery, but it seems clear the lug on the Phase A terminal is heavily corroded.
It's hard to tell from this angle, but do you think this could be fixed by simply replacing the terminal lug?
While I find Mik's work inspiring, I just don't know if I'm ready to tackle repairing the motor controller PCB.
Looks like it has been getting very very hot!
The current sensor (blue) might also be damaged.
It may have been the connection from the "post" on the board to the cable lug, but in any case some damage to the board is likely.
You may be able to fix it the same way as I have, but only if the current sensor is still good.
This information may be used entirely at your own risk.
There is always a way if there is no other way!
This doesn't look good! I am sure you will find the post has melted due to the heat and also damaged the current sensor. The cable itself is also useless now as the heat will have caused it to become very brittle and any bending or movement whilst removing or installing it will break strands within the cable. This was a problem with early boards where the post would become loose and cause arcing thus leading up to the situation present here.
The bad news is, it requires a new motor controller and a motor assembly. It is necessary to change the motor because the cables on the motor end are embedded in epoxy so cannot be swapped out.
I'd get in touch with Dana DaCosta ASAP if he is coming to see you with a battery set and make sure he brings along these bits as well. Sounds like he is going to be a busy boy. Maybe he should hire me on a consultancy basis and fly me over to help out!!! ;-)
Good luck!!
Doug Townley
info.DTES [at] gmail.com
(Formerly Vectrix Europe Technical Supervisor)
One door closes, another door opens
Well, we replaced the motor controller today. Here's how it looked:
You can see that both phases A & C were burned out, and the heat also melted the IGBT on the other side of the PCB, just like in Mik's case. The Phase A cable was trimmed back and given a new lug, which seemed reasonable given the expense and effort of replacing the entire motor assembly. Except for the first inch or so, the copper still seemed relatively flexible and undamaged. We were also fortunate that the cable run was a few inches longer than was really needed.
Apparently, they've redesigned the stand-offs used in the terminals to prevent this from happening. It still seems odd to me that there's no locking washer, but we'll see how it holds up. For now, I'm just happy to be back on the road enjoying the summer sunshine.
Well done! What an amazing on the spot diagnosis!
This information may be used entirely at your own risk.
There is always a way if there is no other way!
That day when Dana DeCosta was swinging through the Pacific Northwest on a mission to rescue broken down VX-1's, mine wasn't the only bike in desperate need of his attention. While my motor controller was getting replaced, I was texting and calling a friend whose 2008 model had died about 6 months earlier. Unfortunately, it being 4th of July weekend, we never got ahold of him and Dana had to move on. Three years later, my friend’s bike is still collecting dust in his garage and he needs to make room for his new Model S. I towed it to my garage and managed to get the battery recharged only to discover the same symptoms documented in this thread. I opened up the bike to visually inspect the motor control, and saw a familiar sight:
So, I’m now two for two. The arcing caused by these loose posts is known by Vectrix and has been fixed in newer motor controller assemblies, but if you’ve got one of these older models I highly recommend checking to ensure these are torqued to spec at 3.5 N•m (31 lb in). Once the sparking begins, it’s going to be very difficult (and expensive) to repair the damage.
This is one of the worst pieces of design in the entire bike. The entire assembly of these posts has a positive temperature co-efficient, (like an old Russian nuclear reactor) IE: the hotter it gets, the hotter it gets. I have had to fix three of these now and once fixed they don't play up again. Like Mics well documented repair but I use pure copper washers to pad the post back to the correct length and a metal lock nut. These copper washers expand more than the steel post and cause the posts to get tighter as it heats up, not looser. But as previously warned in these pages, If you smell burning circuit board, STOP, tow it, It can really save the board and IGBT (the cause of most of the smell)..
The arcing doesn't appear to have spread to the other board parts - it's probably fixable
I use a dremel with a router attachement, and cut away all the plastic surrounding those terminals to reveal the copper lugs inside the IGBT module
once you do that, just re-make the connection and you're back in business :D
Matt
Daily Ride:
2007 Vectrix, modified with 42 x Thundersky 60Ah in July 2010. Done 194'000km