Vectrix died at 22,000ks..Help!

You stated that the vehicle was "hard to push" after the failure, which leads me to believe there is a problem with the motor/gear drive unit. I would see if the rear wheel turns freely, and if not, I would dissassemble the motor/gearbox, and check for damaged/loose items. A bolt could have loosened/broken, or one of those pieces of epoxy used for balance could have come loose, and gotten caught in the works. If the wheel does not rotate freely, never keep trying to apply power! Doing so risks causing further damage. Let us know what you find, send photos if possible. Good luck,-Bob

PS:Try disconnecting the three large motor wires from the controller and then turn the rear wheel, if it now turns freely with the wires unhooked/disconnected, but not with them connected, you have a controller problem. If, on the other hand, rotation is still difficult/uneven with the wires disconnected, there is a problem in the gearbox/motor assembly.-Be sure to keep loose metal items away from the motor magnets, if you open the motor!

I suspect that one DC phase of your controller has been fried up... check the wires to the engine, then check the voltage between the three wires when you apply some little power... you may need controller replacement?... please kep us posted with pictures.

I suspect that one DC phase of your controller has been fried up... check the wires to the engine, then check the voltage between the three wires when you apply some little power... you may need controller replacement?... please kep us posted with pictures.

That can easily get you killed!

Definitely a job for which electrician gloves should be worn, to avoid accidental touching of voltage carrying parts. And one hand behind the back!

The whole system would be "hot" for this measurement, and due to the unknown nature of your bikes fault there might be very unexpected potentials.

It seems to me that the fuse has blown for a reason this time (this might be a first!). I have so far never heard of a fuse failing during a stalled motor at full throttle. That does of course not mean too much, but please wait until you get some feedback from the Vectrix technicians around here so that you can avoid "poking around in the dark", so to say.

If you have to stick your hands down into the guts of a Vectrix with a closed Andersons connector and an unknown serious electrical fault, you want to be sure it is worth the risk!

.

My best bet so far would be that one or several of the IGBTs have fried. Maybe that can lead to a situation where a short causes a braking action when you pushed the scooter, causing the resistance you felt. That same short would have blown the original fuse and caused the unusually strong spark when you closed the Andersons connector.

If the reason for the high pushing resistance was a mechanical problem in the wheel or gearbox, then I would expect that the motor can at least turn the wheel slowly, otherwise you could not have pushed the scooter by hand.

Another area to inspect, is the motor WIRES, especially where they are subject to flexing/abrasion as they make their way to the controller. If they have become damaged/shorted, this will cause a problem. (If it were me, I would first disconnect the batttery, then remove the motor wires from th controller, and check for smooth, easy rotation of the motor/wheel assembly.) This would quickly isolate the problem, to the controller, if it turns easily when disconnected, or the motor/wheel, if it turns rough/difficult.-Always observe safety procedures around high voltage!-Bob

EITS
Sounds like you definitely have a blown IGBT. Any time any 2 phase wires are touching (in an unpowered state) it will cause a braking action in the motor. Since the IGBT is blown it is basically shorting at least 2 of the phase wires. Altho I think you may have gotten rid of the short that first time you re-connected the battery cables! I dont think connecting w/o an ICL did the damage. If it were just a blown fuse initially the bike wouldnt be hard to push.
If you have good soldering skills you could purchase an IGBT and replace the burnt one, as long as there wasnt some other problem that caused the IGBT to fail...ie a faulty current sensor etc.

BTW...before this happened did you notice if the MC fan was working? Did it come on full blast under heavy acceleration?

EITS
Sounds like you definitely have a blown IGBT.
...

If you have good soldering skills you could purchase an IGBT and replace the burnt one, as long as there wasnt some other problem that caused the IGBT to fail...ie a faulty current sensor etc.
...

You would have to remove this black box from the back of the MC board:

By de-soldering this row of golden pins:
That takes a bit of skill!
It also looks like the black box might be glued shut, or potted, I don't know for sure. Tricky job, might be easier to buy a new Vectrix at rock-bottom prices!

Those are the ones. Its really not that bad if you have the nads to try it. You clip the pins, then use some solder wick on each pin (you do need a good soldering iron tho, no cheapies here). Use an exacto knife or similar to cut thru the silicon on the back around the edges of the iGBT, remove the phase and V+, V- lugs and it should pop out. Or check your area and see if there is a electronics repair facility nearby...should be a 10 minute job for them.

Well thats light at the end of the tunnel! I did notice the MC fan working'
I will start dismantling the MC and give it a go, where do a get another IGBT from
what is its unabbreviated name?

There is a link to the spec sheet for the IGBT module here: http://visforvoltage.org/forum/3635-vectux-part-2-open-source-vectrix#comment-20624

Hello,
I found the part here;
http://search.digikey.com/scripts/DkSearch/dksus.dll?WT.z_header=search_go&lang=en&site=us&keywords=PM300CLA060&x=21&y=14
$468.00, not a cheap part!

hi pete,
firstly we need to set a few things straight. with regard to the other comments relating to your post, they clearly show a lack of knoweldge and if followed could lead to further tears, please do not give advice unless you are absolutly sure that you are right the repair that is about to be undertaken.
The repair here is not just a simple case of removing the broken bit and putting in a new one, these IGBT cost around $1000. AUD, my question to the other respondents out there is are you willing to replace the IGBT after it failed a second time because Pete took your advice and simply replaced it? don't bother i know the answer NO. I am supprised by this considering the ammount of people out there that own a vectrix and are now working on them due to no other choice.

So for the benifit of everyone out there here is right way to replace the IGBT module.

The motor controller board is a dual layer fibergalss board with thru plated holes and must be treated with a great deal of respect or perminate dammage will result.

The equipment needed are:
1) "zero voltage tip" temperature controlled soldering iron, anything that pulgs into the wall and has the iron at the other end has the potential of static discharge and blowing up the input stage of the IGBT (insulated gate bipolar transistor) and though the name implies that they are safe to handle it is generally only the output that is robust.
2) a desoldering station (vaccume pump) the reason for this is because the board is thicker than most requiring the solder to be sucked out with minimal heat,using solder wick will result in damaging the board as the ammount of heat required is excessive and there is no garentee that the hole would be free of solder when prying the igbt from the board.

My sugestion would be to take the board to a TV repair technician who would have the right equipment to safely extract the IGBT, BUT it dosent end there, READ ON as the real interesting bit is next.

The correct proceedure is as follows.
1)remove the motor controller module INCLUDING THE HEAT SINK that it is attached to. this is whole block and dice MC board, heat sink, fan unit.
1a)claen the heat sink compound (white paste)of the heat sink, the heat sink needs to be spotless.
1b)clean the heat sink compound off the MC board. there are two items on the back of the MC board that connect to the heat sink, the IGBT and the DC/DC converter. the DC/DC converter needs to be spottlesly clean.
2)remove the MC (motor control) board from the heat sink and carefully extract the IGBT from the board
3)insirt the new IGBT into the MC board BUT DO NOT SOLDER IT.
4)replace the MC board and floating IGBT module back onto the heat sink, put in all of the screws and tighten them up, from memory the torque settings are 5Nm, the screws should be firm but not over tight. THIS IS CRITICAL AS IT SETS THE HIGHT TOLLERENCE FOR THE NEW IGBT. if this step is not carried out correctly or the board and heat sink are not totally clean then you have a better chance than most of kissing goodnight to the new IGBT as the terminals will be stressed and will fracture, if not straight away within a very short time.
5)with everything back in place resolder the IGBT using gravity (the reason for having the whole unit out). because this is a repair you can no longer rely on capilatry action for solder flow and a good connection.
6)remove the MC board and apply an even film (.5mm - 1.0mm) of NEW heat sink compound make sure that there are no lumps in the compound and replace MCboard and tighten down in multiple steps to the heat sink in the following order.
a) IGBT module
b) DC/DC converter
c) MC board screws
7) replace the Control unit back into the bike and put every thing back together, be careful to replace the motor cables in the same order.

The reason why the bike was hard to push would have been a short between the motor windings causing a back EMF resistance to the magnetic rotor. It would appear that the short was a bridging through the IGBT module, regen braking in on mode.

I hope this helps, cheers paulyd.

Hi Paul, good to have you back! You have learned a lot about these motor controllers! ;-)

This thread has been added to the Vectrix Collaborative Handbook, please stay on topic!

the black boxes are in fact ceramic fuses in a little back box filled with sand, the gold blocks are capacitors, pray to god that these dont fail, to my knoweledge they are made of unobtainium.

the black boxes are in fact ceramic fuses in a little back box filled with sand, the gold blocks are capacitors, pray to god that these dont fail, to my knoweledge they are made of unobtainium.

What are the chances that the primary fault is something else, and that the IGBTs failed as a result?

And what are the chances that the IGBTs "took other parts of the MC down" when they failed?

Is it worth taking the risk of replacing the expensive IGBTs when one could buy a new VX-1 for around AU$5000.-?

Could it be possible to replace single IGBTs in the black box, or is a complete new box with three IGBTs needed each time?

the black boxes are in fact ceramic fuses in a little back box filled with sand, the gold blocks are capacitors, pray to god that these dont fail, to my knoweledge they are made of unobtainium.

What are the chances that the primary fault is something else, and that the IGBTs failed as a result?

And what are the chances that the IGBTs "took other parts of the MC down" when they failed?

Is it worth taking the risk of replacing the expensive IGBTs when one could buy a new VX-1 for around AU$5000.-?

Could it be possible to replace single IGBTs in the black box, or is a complete new box with three IGBTs needed each time?

mick mate the IGBT is an intergrated unit last time i checked it was worth around $650US. the chances that the IGBT took out other parts are slim how ever the cahnces that another faulty part took out the IGBT are good. if the input signals are not right then the MOSFET's that drive the high power transistors can suffer from runaway and cause the IGBT to fail as a unit, buying a VX-1 for $5000. will give you 102 stuffed batteries and one MC card which were listed new at $2400. i would try sending the listing on e-bay a question to see if there are still parts available. ??? worth a try eh.

It keeps coming dowm to the question when the first owners are so fed up that they start selling their bikes remains for cannibalizing parts!

So far, everyone seems to keep hanging on because (when it works) the Vectrix is special.

I thought so far I might start buying "New" Vectrixes for AU$2000.- once available. At higher prices the risk of endlessly sinking more money down a bottomless pit seems too high.

Sooner or later there will be "cannibalized" parts available.....

When you say "102 stuffed cells" - do you think the cells are irreparably damaged, or just in need of a lot of careful attention? NiMH cells might be resurrected with the right methods, sometimes.

buying a VX-1 for $5000. will give you 102 stuffed batteries and one MC card which were listed new at $2400. i would try sending the listing on e-bay a question to see if there are still parts available. ??? worth a try eh.

the $5000 vectrixs are brand new, i know because i bought one :D

personally, id be buying one of the new cheap ones, and keep the old one for spare parts

Matt

As previously mentioned, I would thouroughly inspect the motor and all associated wiring for any possible damaged/loose/oveheated hardware or electrical items, to reduce chances of blowing any replacement module or controller. SOMETHING went wrong to cause your failure, you do not want a repeat! I do not know if it would be practical to run initial testing with a resistance in series with the phase leads, but doing so, with the wheel spinning free, might be a good idea.--I would use high-power 1 ohm resistors, something rated at 250 watts or higher. Fault current would thusly be limited to about 125-150 amperes per phase, with the resistor blowing if a malfunction occured, instead of the controller/module.--I see that POWEREX has a similar, higher rated module, the PM600CLA060, which can operate at TWICE the current of the Vectrix module, but I do not know if it would be feasable to retrofit this more robust unit as an upgrade. (It also COSTS much more!)--Lets see now-Install a 45 cell, 40 AH lithium-polymer battery, using KOKAM cells, rated for 20C or higher discharge, with the 600 amp controller module, and more powerful cooling fans?-193 volts at full charge, about 135 at discharge end? Hot Rod Vectrix, anyone? ZOOOM! (Might need a water spray to cool the motor, if operated too long!---Bob

It seems that new vectrix will extend the warranty of our vectrix. Is the bike under warranty? Maybe you can buy a new controller.... cheaper than a new bike...

It seems that new vectrix will extend the warranty of our vectrix.

I'm sure exactly what this means. Does that mean that our warranties will last longer than just 2 years?

It seems that new vectrix will extend the warranty of our vectrix.

I'm sure exactly what this means. Does that mean that our warranties will last longer than just 2 years?

Please discuss this in another thread, this one is about the motor controller / IGBT module!

It keeps coming dowm to the question when the first owners are so fed up that they start selling their bikes remains for cannibalizing parts!

So far, everyone seems to keep hanging on because (when it works) the Vectrix is special.

I thought so far I might start buying "New" Vectrixes for AU$2000.- once available. At higher prices the risk of endlessly sinking more money down a bottomless pit seems too high.

Sooner or later there will be "cannibalized" parts available.....

When you say "102 stuffed cells" - do you think the cells are irreparably damaged, or just in need of a lot of careful attention? NiMH cells might be resurrected with the right methods, sometimes.

the australian bikes were charged from the factory then left to sit in a warehouse for 9 months and then charged before delivery and so on. this was due to the lack of information from vectrix and no fault of the current / latest management. basically the batteries would have been dammaged and this is why the specs have been degraded on the e-bay site. its too hard to say if they would all have been damaged irriversabily but it would be pot luck to get a good one.

Having read the posts, isn't hindsight wonderful, If you had used an 'inrush current limiter, you would have saved yourself a lot of work and some of the resulting damage

I don't know what the official current limiter looks like electrically but all that is needed is a single 240 Volt 40 watt or less filament lamp and an insulated socket with the leads taken to a test probe (insulated) or similar.

THE FOLLOWING PROCEDURE IS SERIOUSLY HAZARDOUS. I HAVE DONE IT THIS WAY. I AM NOT SUGGESTING THAT YOU SHOULD.

Instead of reconnecting the Anderson plug socket, Make the positive supply lead your final connection point. YES, I know that there are serious safety hazards doing this, but care can overcome all hazards. Connect the test lead a couple of cells back from the battery positive (this gives more space to work at the positive terminal) and you will observe the following.

An open circuit fuse will give no light (the lamp will not light up).
A short circuit or an unacceptably high current flow will cause the lamp to light and stay lit.
If all is well, the lamp will light and then dim quickly. Now you may complete the connection of the battery.

Taking this idea a little further. Before making the final connection it may be possible to repeat the above process with the ignition switched on. This may show up any problems which would otherwise not show until you you switch on. A higher wattage lamp may be necessary. I have not tested this latter idea.

THIS IS, OF COURSE, A SERIOUSLY (POTENTIALLY) HAZARDOUS PROCEDURE. I AM NOT SUGGESTING THAT YOU TRY IT. THE POWER AND VOLTAGE OF THIS BATTERY MAKES ALL WORK WITH AND ON IT HAZARDOUS. UNLESS YOU KNOW WHAT YOU ARE DOING AND TAKE ADEQUATE MEASURES TO PROTECT AGAINST ELECTRIC SHOCK AND SHORT CIRCUITS, DON'T EVEN THINK ABOUT IT.

If anyone feels that these ideas should not be published here, then PLEASE complain to the webmaster and request their removal as a matter of urgency.

The Laird

Having read the posts, isn't hindsight wonderful, If you had used an 'inrush current limiter, you would have saved yourself a lot of work and some of the resulting damage

I don't know what the official current limiter looks like electrically but all that is needed is a single 240 Volt 40 watt or less filament lamp and an insulated socket with the leads taken to a test probe (insulated) or similar.
...
...

Yes, as far as the use of a light globe goes.
As I wrote:

Eyeinthesky's recent drama (frying the IGBTs) has shown another good reason to use an ICL which gives feedback (that the capacitor charging is finished).

Simple resistors will work as ICL, but if there is a short which caused the fuse to open, then you will still get a nasty big spark when you connect the Andersons! That is because with a short somewhere, the capacitors will never be charged and the ICL does nothing useful.

A light globe as ICL will remain lit if a short condition is present in the IGBTs, thus showing you that the Andersons connector must not be connected!

.

http://visforvoltage.org/forum/7741-vectrixs-recharge-halted-due-water-sensor-temp#comment-45163

But I disagree about connecting the positive terminal last. It just does not add any useful function, just severe risk!

.

The light globe as ICL is described here: http://visforvoltage.org/book/ev-collaborative-hand-books/6155

Maybe you could add your experience and then we can elevate it's status to "Vectrix Inrush Current Limiter with considerable testing".

Seems like at least TheLaird, Eyeinthesky and myself have used this type of ICL now.

Hi Eyeinthesky,

how did this go? Did you get it fixed?