Went high power, destroyed something...

[rjc34]

So I recently posted here asking for some advice on a LiPo pack and controller. Settled on a 59.2V nom 16S4P pack made up of 16 4S hard pack LiPo bricks paired to a Lyen 9-Fet controller.

After some sweat and soldering, managed to get everything all wired up and ready to go. Stupidly I assumed the hall and phase wires for the motor matched colour to colour, so I just hooked it up that way and in my eagerness took it for a test ride. What I know now is that I had 2 of the Hall wires revered (Blue to Blue and Green to Green when it should have been Green to Blue and Blue to Green), and 2 phase wires mixed (should have been either yellow and blue switched, or green and blue).

So anyway, in my eagerness to test it, everything looking good, I gave it a bit of throttle. It was loud and a bit rough, but again very stupidly I just brushed it off as being the motor groaning under the higher voltage and current. Well... seems that wasn't the case. As the speed maxed out (admittedly slower than my 36V pack would take me), the groaning stopped, so I assumed it was alright. Slowed down, started up again through the same loud groaning and then suddenly the CA is reading 8000W being used, then the motor simply cuts out and I guess the term 'cogs' would be correct, as it provided a ton of resistance pedaling back with the controller off. When I got back the motor was extremely hot (too hot to touch), and the hall wires were equally hot. The controller also had a significant amount of heat coming off of it.

Now after fixing the wiring problems, the controller still reads 8000W or so when turned on, and I get absolutely no response from the motor.

So my question is... did I kill the motor, controller, or both... If there's any other information any of you might need to help diagnose, or anything you think I should try and report back on, I'm all ears.

 

[VoKuS]

So I recently posted here asking for some advice on a LiPo pack and controller. Settled on a 59.2V nom 16S4P pack made up of 16 4S hard pack LiPo bricks paired to a Lyen 9-Fet controller.

After some sweat and soldering, managed to get everything all wired up and ready to go. Stupidly I assumed the hall and phase wires for the motor matched colour to colour, so I just hooked it up that way and in my eagerness took it for a test ride. What I know now is that I had 2 of the Hall wires revered (Blue to Blue and Green to Green when it should have been Green to Blue and Blue to Green), and 2 phase wires mixed (should have been either yellow and blue switched, or green and blue).

So anyway, in my eagerness to test it, everything looking good, I gave it a bit of throttle. It was loud and a bit rough, but again very stupidly I just brushed it off as being the motor groaning under the higher voltage and current. Well... seems that wasn't the case. As the speed maxed out (admittedly slower than my 36V pack would take me), the groaning stopped, so I assumed it was alright. Slowed down, started up again through the same loud groaning and then suddenly the CA is reading 8000W being used, then the motor simply cuts out and I guess the term 'cogs' would be correct, as it provided a ton of resistance pedaling back with the controller off. When I got back the motor was extremely hot (too hot to touch), and the hall wires were equally hot. The controller also had a significant amount of heat coming off of it.

Now after fixing the wiring problems, the controller still reads 8000W or so when turned on, and I get absolutely no response from the motor.

So my question is... did I kill the motor, controller, or both... If there's any other information any of you might need to help diagnose, or anything you think I should try and report back on, I'm all ears.

Dude slow down!
High power lipo is no joke!
High voltage is dangerous!
You can not just assume things.
Is your pack in a safe fire proof place right now? (If it in anyway got damaged you can have an un-expected fire!)
Please be safe!

Before doing anything else, I would check the pack it self and make sure its in proper order.

Other members can chime in on the controller/hub because I only have experience with lipo chemistry it self.

When testing an ebike for the first time, its a good idea to flip it over, or hoist the hub motor side in the air instead of jumping on to the bike and giving it max load...

 

[rjc34]

Dude slow down!
High power lipo is no joke!
High voltage is dangerous!
You can not just assume things.
Is your pack in a safe fire proof place right now? (If it in anyway got damaged you can have an un-expected fire!)
Please be safe!

Before doing anything else, I would check the pack it self and make sure its in proper order.

Other members can chime in on the controller/hub because I only have experience with lipo chemistry it self.

When testing an ebike for the first time, its a good idea to flip it over, or hoist the hub motor side in the air instead of jumping on to the bike and giving it max load...

Thanks for the concern about the batteries. They're all disconnected and safe currently. Left them in the garage that night disconnected just to be sure, but they're definitely OK.

 

[dogman dan]

Most likely fried fets in the controller. Unplug the motor, if it stopps cogging, the motor is still ok most likely.

 

[DAND214]

You killed the controller for sure, maybe the hub too.

Fist disconnect the phase wires and se if you still have the cogging. If not then the hub should be ok. You will also need to check the halls to be sure they didn;t melt.

As posted earlier, you always need to be sure the color combo is correct before putting a load on the system.

As for the batteries, have you checked the cells not just the packs. With such a large load you might of killed a cell or two.

16s4p should be fine as you should of melted someting else before you killed the battery.

8000w? Didn't you think that was a little high? 3000w would be high for the controller.

Choices, repair the controller, yourself or Lyen or get a new one. If you don't have a spare i would do both.

Dan

 

[wesnewell]

8000W is nothing to a 16s4p (20ah 20C pack) That's a 400A limit, at 60V, 24000W. He only hit it at 133A for 8000W. Might need to check the wiring/connectors though.

 

[DAND214]

8000W is nothing to a 16s4p (20ah 20C pack) That's a 400A limit, at 60V, 24000W. He only hit it at 133A for 8000W. Might need to check the wiring/connectors though.

I agree and disagree! 8000w with no power is a mistake. 8000w with wheelies is nice.

I agree the wiring was wrong and you still need to check for melted wiring. 132v with that kng of current available I would look at all the wiring.

Dan

 

[dogman dan]

Batteries themselves should be ok, but wires and connectors could heat up. Could be the short is actually in a heated up wire in the axle. But a short there would likely still kill the controller, so I don't have a lot of hope for it.

 

[rjc34]

I don't think it ever actually drew 8000w from the battery. When it first started it was just normal around 2000, then it started showing 8000 but no voltage drop or anything else. I'm pretty sure the motor was cogging after I unhooked everything, so I guess I blew that too.

If the motor turns out to be blown... I'm not sure what I'll replace it with yet. I've been thinking about one of those crystalyte rear hubs from ebikes.ca. Open to anything else (motor and controller wise) that'll go well with my pack.

 

[wesnewell]

What motor do you have and what are your goals? 16s is kind of a weird pack. Too high for most 48V controllers with 63v caps, and too low for a 72V controller. I'd change the pack to either 12s for a 48V controller or 20-24s for a 72v controller. I run my 4s hard packs as 24s2p on a 1000W motor with a $35 72V 40A controller. Works fine for me.

 

[dogman dan]

Quite possible that the only damage to the motor was melting insulation on the wires as they enter the hub at the axle. So don't give up on the motor till you check for excess cogging after cutting the wires inside the motor. Maybe all you need is new wiring.

 

[rjc34]

Quite possible that the only damage to the motor was melting insulation on the wires as they enter the hub at the axle. So don't give up on the motor till you check for excess cogging after cutting the wires inside the motor. Maybe all you need is new wiring.

Thanks for the advice. I'll crack open the motor when I get a chance and see what's going on in there.

Any recommendations for a new controller?

 

[liveforphysics]

Classic false positive phase/hall wire combo. The battery is likely fine. Unhook all motor phase leads and try to ride/spin the wheel, if it spins freely, you likely just have a dead controller, if it spins like it's filled with peanut butter after the leads are unhooked, you likely have a dead motor and controller.

A false positive phase/hall combo can wipe out any motor/controller combo.

 

[rjc34]

Classic false positive phase/hall wire combo. The battery is likely fine. Unhook all motor phase leads and try to ride/spin the wheel, if it spins freely, you likely just have a dead controller, if it spins like it's filled with peanut butter after the leads are unhooked, you likely have a dead motor and controller.

A false positive phase/hall combo can wipe out any motor/controller combo.

Thanks for the help.

As an update to the thread, I've now pretty much concluded that the motor is 99% likely completely fine, that I need to properly tension my spokes and replace a broken one, and that the controller is either completely dead or just suffering from a single wire becoming disconnected.

After I had concluded that the controller was pretty much toast, I opened it up just to see what was up. To my surprise everything looked fine except for a single disconnected wire that belonged to the cycle analyst connection.

Picture: http://i.imgur.com/MwWAz.jpg

Now, I'm not sure if that's the cause of all my problems, or just an incidental problem to the FETs being blown or whatever else. Anyone got any help/advice? I just sent a message off to Edward Lyen asking what he thinks of the problem.

 

[dnmun]

did you check the mosfets on each phase? they may be shorted.

maybe the wire came loose when the solder melted.

 

[rjc34]

did you check the mosfets on each phase? they may be shorted.

maybe the wire came loose when the solder melted.

Would I be able to test that with a multimeter?

Edward got back to me and said the loose wire should be connected to the shunt on the side with the orange capacitors. Later today I'll solder it on and see what happens.

 

[dnmun]

yes, that should be the first thing to do if the controller blew up. measure between the high side where the big red wire from the battery attaches to the controller to each of the phases and then measure from the ground or negative black wire on the controller to each of the phases.

you can use the diode function on the voltmeter and measure the source drain forward bias.

the drain is always on the upper leg. the red to the phase is drain to source, and the phase to ground is drain to source. you wanna measure the forward bias of source to drain.

in reverse bias it should be open circuit.

 

[rjc34]

yes, that should be the first thing to do if the controller blew up. measure between the high side where the big red wire from the battery attaches to the controller to each of the phases and then measure from the ground or negative black wire on the controller to each of the phases.

you can use the diode function on the voltmeter and measure the source drain forward bias.

the drain is always on the upper leg. the red to the phase is drain to source, and the phase to ground is drain to source. you wanna measure the forward bias of source to drain.

in reverse bias it should be open circuit.

Alright, so I understood enough of those words to grab my multimeter, set it to diode test, and do a bunch of measurements. I'll list the measurements that I got here, and hopefully someone will be able to tell me whether they're good or bad

Red lead from multimeter to battery input red wire, and black lead from multimeter to the ends of the FETs gives me a quick climb up to ~1900 for 8 of them, but one of them starts at -1000 and drops to 0.

Black lead from multimeter to battery input red wire, and red lead to the FETs gives me ~610 on 7 of them, with the one with the negative reading from last test pretty much doing the same thing, and a different one reading ~910 off the opposite side lead.

Red on battery red and black on phase connections leads to similar behavior to first test (climing to ~1900 then going back to 1). Black on battery red and red lead on phase wires leads to steady ~610 reading on all 3.

 

[dnmun]

opposite side lead means the gate? should be open circuit except the gate is tied to the source through a resistor i think.

sounds like there is a 2k ohm resistor across the S/D cap you are reading as 1900 ohm, and the 610mV sounds ok so just the ones that measured zero appear dead, except i really didn't follow the verbage so well. but i think you can see which one is bad now.

you can pull out the dead one and then remeasure and the 1900 ohms should reappear on that phase.

there are experts around so they may comment too.

 

[rjc34]

I'm seriously considering just ordering the 48-72V 25A controller from Ebikes.ca. It would certainly be a lot easier.

 

[dnmun]

sure, you can worry about fixing this when you feel more confident about repairing stuff.

so it could be the backup controller project. but you have an idea of what's wrong now. people regularly talk about fixing their lyen controller so you can just follow those threads for the time being.

 

[rjc34]

Could do.

I think it's more that I just want to get back to riding again. I'm sick of taking the bus to college every morning! Not to mention that I can go right back to plug and play; no having to solder all those connectors again.

 

[rjc34]

Went ahead and ordered the 48-72V 25A controller from ebikes.ca along with some other goodies like a full set of replacement spokes, a spoke tool, and a freewheel removal tool.

Any advice regarding the FET testing numbers I posted or what to do with the busted controller is appreciated.