8000w Novice Build! Advice Needed!

[narakutaishi]

Hello all,

Long story this one,
Bought a used stealth bomber clone 8000w T5, 150A Sabvoton controller, 72v 24ah battery with all other wrighting on this unit in chinese worked for my daily commute for a year in all weathers, plenty torque and would pull all the way up to 60mph. Never skipped a beat and always guaranteed a smile!

Was heading out on it one day coming up to a corner I locked the back break up coming in to a corner (no cut off switch in break leaver, hope tech 4s I'd fitted myself on this bike) the motor made a heavy grinding noise and the controller knocked off. When I turned it back on there was no green light on the controller but the screen came one.

Bought a new 8000w motor and controller
8000w T3 motor
150A Sabvoton

Realised the battery BMS had blown, so had the battery balanced and the BMS replaced like for like with 150A unit

I've put it all together and it just kept cutting out when I get to certain speed and I have to disconnect and reconnect the battery to get it to go again. I have found some settings for the MQCON app from someone programming the same size motor and controller on yiutube. It's definitely better but still cuts out at a certain point.

So I'm kind of stuck here,
The setup I originally had felt bulletproof, could keep the throttle wide open all the time, it never let me down till the day it did haha

Any help would be much appreciated!

 

[E-HP]

!lWas heading out on it one day coming up to a corner I locked the back break up coming in to a corner (no cut off switch in break leaver, hope tech 4s I'd fitted myself on this bike) the motor made a heavy grinding noise and the controller knocked off. When I turned it back on there was no green light on the controller but the screen came one.

Realised the battery BMS had blown, so had the battery balanced and the BMS replaced like for like with 150A unit

Seems like some missing info. If the controller shutdown, but the display stayed on, then the BMS didn’t trip. What made you think The BMS died? The shop would have told you that, so there’s some missing info in what you’ve described.

Could be your battery is just dying. What is your charging/maintenance routine? (E.g. charge to 80% discharge to 20%, etc.). Do you know the previous ownwers charging routine? Can you monitor your voltages in real-time?

When you locked up, were you still on the throttle, so you stalled the motor under load?

 

[narakutaishi]

The screen stayed on initially, when I got it home nothing was working at all, when I plugged the charger in the screen came back on but the battery wouldn't hold a charge. After getting the BMS replaced the battery is working fine now. My charge routine was get to work/home 35-45% battery left then charge it to full just in case I had to take evasive action "again" not sure about previous owners charge routine.

Yes I do believe I was still on the throttle slightly when I locked the back break up, I think that's what broke it.

After a bit more tinkering around with MQCON today I have found a nice balance where modes up to 3 will let me give it full throttle, mode 3 powers up to about 40-45mph

I tried the same on setting 4 and it blew the 80A fuse from the battery to the controller, I've just replaced it and rode it hard in mode 3 for 10 miles without a hiccup.

I know this motor is capable of more speed my question is what's limiting me here, is it the battery or just some simple setting on MQCON?

 

[narakutaishi]

 

[E-HP]

I tried the same on setting 4 and it blew the 80A fuse from the battery to the controller, I've just replaced it and rode it hard in mode 3 for 10 miles without a hiccup.

Do you have more information about your battery, the cells used and the configuration? My first impression was that a 72V 24Ah battery is undersized for a bike using a 150A controller, but was guessing it is made with some really good cells. But, the fact that the previous owner placed an 80A fuse between the battery and the 150A controller, tells me it was known to be undersized.

 

[narakutaishi]

I have no idea about the battery other than that very brief 72v 24ah description. With the battery having a 150A BMS wood it he safe to say I could increase the size of this fuse to anything less than 150A so that would blow before the BMS. I have blown a 150A controller that came with the new motor, so I'm actually on my 3rd controller now and really don't want to be buying any more of the expensive components

 

[E-HP]

If the battery is capable of delivering 150A, then using a 150A BMS seems fine. If the battery and controller are rated for 150A, I see no reason for using an 80A fuse, unless your conductors are undersized.

I've put it all together and it just kept cutting out when I get to certain speed and I have to disconnect and reconnect the battery to get it to go again.

Next time it happens, measure the voltage at the battery input terminals of the controller, before disconnecting the battery. If the display works when your controller doesn't, then it's not the battery.

You probably need brake cutoffs if you keep killing your controllers. You're doing the ebike version of driving down the freeway and throwing your car into Park.

 

[amberwolf]

You're doing the ebike version of driving down the freeway and throwing your car into Park.

The difference is that a properly designed controller (especially an FOC type) ought to be able to detect the excessive phase current (especially vs motor speed, decreasing or zero) and shut that down, so it doesn't cook the motor or blow the FETs.

Even if it doesn't have phase current sensors (most non-FOC types won't), it should see a sustained battery current spike vs actual motor speed (decreasing or zero) and shutdown.

At minimum it ought to have some form of temperature monitoring of the FETs and shutdown when those get too hot so they don't blow up.

Most controllers dont' actually have any of those protections, though, and I don't recall any that have all of them. So...effectively it's the "same" as what you're describing.

 

[amberwolf]

I have no idea about the battery other than that very brief 72v 24ah description. With the battery having a 150A BMS wood it he safe to say I could increase the size of this fuse to anything less than 150A so that would blow before the BMS.

Is your wiring and connections from the battery to the controller able to all handle 150A continuous? If not, then use a fuse rated for what the wiring/connections can actually continuously handle.

If the fuse is at the battery end, it's able to protect the battery from a wiring or controller short.

If the fuse is at the controller end, it's able to protect the battery and wiring from a controller short.

Neither one is really there to protect against a system overload, because the way they rate fuses isn't really compatible with that. Fuses are rated so that they will last "forever" at the rating. Above the rating, there is a curve (published in the specs for that brand and model of fuse***) that decreases in time vs increases in current, for the point at which the fuse will blow.

It can take minutes to hours for a fuse to blow when used just above it's rating, but milliseconds when it's far enough above it. That's probably why the 80A fuse was used--if the system (or any part thereof) is really only able to handle 150A for a short time, the fuse would blow eventually if that 150A was sustained, but not if it was used in short enough bursts (like during acceleration).


The BMS is there to protect the battery cell block itself from any short or overload past it's output. It should detect an overcurrent and turn off the output once it's time limit has passed (varies depending on design, some are programmable).

Buuut...under extreme conditions near the spec limits for FETs especially, FETs tend to fail shorted, leaving the BMS stuck on, so it can't actually protect against overcurrents. It's a silent failure, so unless you test specifically for this, you won't even know that this problem exists until you encounter a situation where the battery should have turned off but didn't (like overdischarging it because a cell went too low), and probably already damaged the cells in the process.


The fuse won't protect against htat, but if it's sized low enough to blow *within* the limits of the BMS and cells, then if a high enough drain continues long enough it will blow and protect them against that, at least.


Also...just because the BMS is rated for 150A doesn't mean the cells are capable of that; it's not uncommon for batteries to be built of cells that can *peak* at the rating the battery is sold as, but not do that continuously, and possibly even more common for them to use cells that can't even handle that much, so they sag greatly in voltage even though they're within the "ratings".


***for actual brand name fuses. Clones and generics you can only guess what they will actually blow at; they may not even blow at all under the usage you need them to protect things at, or they may blow well below their "rating". Or they may be similar to whatever they cloned, if htey properly built the fuses mechanically to match the cloned device, *and* they did the correct metallurgy to match the cloned device. (both are not very likely even if you get one of them). Different batches of the same clone/generic may not even blow on the same conditions.

 

[narakutaishi]

If the battery is capable of delivering 150A, then using a 150A BMS seems fine. If the battery and controller are rated for 150A, I see no reason for using an 80A fuse, unless your conductors are undersized.

Next time it happens, measure the voltage at the battery input terminals of the controller, before disconnecting the battery. If the display works when your controller doesn't, then it's not the battery.

You probably need brake cutoffs if you keep killing your controllers. You're doing the ebike version of driving down the freeway and throwing your car into Park.

Yes the controller turns off but the screen stays lit up, I knocked on the brake setting I've set this so it decelerates when I let the throttle off this has a great effect before I pull the break. Before this on the set up I blew a controller at it felt more like a runaway train before I put the break on, I'm very confident with the set up I have now and the brains to not lock the back break up as if I was racing downhill on a MTB that I won't blow another controller thank you for the advice about cut off sensor in the break leaver much appreciated

Also the power lines are big thick cables these are the original ones on the bike, after the first hiccup that blew the motor and controller it blew one of these cables where it connects to the controller with a M6 screw

 

[narakutaishi]

Is your wiring and connections from the battery to the controller able to all handle 150A continuous? If not, then use a fuse rated for what the wiring/connections can actually continuously handle.

If the fuse is at the battery end, it's able to protect the battery from a wiring or controller short.

If the fuse is at the controller end, it's able to protect the battery and wiring from a controller short.

Neither one is really there to protect against a system overload, because the way they rate fuses isn't really compatible with that. Fuses are rated so that they will last "forever" at the rating. Above the rating, there is a curve (published in the specs for that brand and model of fuse***) that decreases in time vs increases in current, for the point at which the fuse will blow.

It can take minutes to hours for a fuse to blow when used just above it's rating, but milliseconds when it's far enough above it. That's probably why the 80A fuse was used--if the system (or any part thereof) is really only able to handle 150A for a short time, the fuse would blow eventually if that 150A was sustained, but not if it was used in short enough bursts (like during acceleration).


The BMS is there to protect the battery cell block itself from any short or overload past it's output. It should detect an overcurrent and turn off the output once it's time limit has passed (varies depending on design, some are programmable).

Buuut...under extreme conditions near the spec limits for FETs especially, FETs tend to fail shorted, leaving the BMS stuck on, so it can't actually protect against overcurrents. It's a silent failure, so unless you test specifically for this, you won't even know that this problem exists until you encounter a situation where the battery should have turned off but didn't (like overdischarging it because a cell went too low), and probably already damaged the cells in the process.


The fuse won't protect against htat, but if it's sized low enough to blow *within* the limits of the BMS and cells, then if a high enough drain continues long enough it will blow and protect them against that, at least.


Also...just because the BMS is rated for 150A doesn't mean the cells are capable of that; it's not uncommon for batteries to be built of cells that can *peak* at the rating the battery is sold as, but not do that continuously, and possibly even more common for them to use cells that can't even handle that much, so they sag greatly in voltage even though they're within the "ratings".


***for actual brand name fuses. Clones and generics you can only guess what they will actually blow at; they may not even blow at all under the usage you need them to protect things at, or they may blow well below their "rating". Or they may be similar to whatever they cloned, if htey properly built the fuses mechanically to match the cloned device, *and* they did the correct metallurgy to match the cloned device. (both are not very likely even if you get one of them). Different batches of the same clone/generic may not even blow on the same conditions.

The fuse is between the main power line of the battery and the main power line connector that goes to the controller. Thank you for your advice I will order some genuine fuses of a 120A rating and hopefully I can start riding in modes 4 and 5 and I can get that 60mph top speed without doing any damage other than the fuse blowing at worst. Thanks again for your help everyone!

 

[Ssk28]

J'ai eu exactement le même problème que vous...une coupure intempestive à chaque fois que j'ouvrais les gaz en grand.
Je devais débrancher les connecteurs batterie et les reconnecter.
j'ai remplacé le bms , ensuite le contrôleur et toujours ce problème.
Ma batterie 20s 13p bms 80a annoncée pour 40ah a été achetée sur aliexpress et je pensais qu'elle était de mauvaise qualité...après démontage, il s'agissait de cellules correctes ( Bak 2600mah ) et donc réellement 32ah de capacité, loin de ce qui était annoncé, mais une seule bande d'acier-nickel de 0.15 x 10mm pour les connexions en série.
Et c'était ça le problème....
Après avoir soudé 1 bande de nickel pur 0.2 / cuivre 0.2 en sandwich par dessus celles existantes, le problème à été résolu.
le bms ne coupe plus par sécurité et je peux profiter pleinement de la puissance de mon moteur qs 205.
J'espère que cela vous aidera.

 

[amberwolf]

Yes the controller turns off but the screen stays lit up,

That usually indicates a battery that is dropping in voltage so far that the controller LVC kicks in. That usually happens because the cells are incapable of handling the load, so they sag in voltage during high current draw.

It can also happen because of poor connections, which includes fuse holders along with all the other connectors and bolt-on connections and crimped connections and switches between the battery and the controller.

Which type of fuse holder do you have? If it's not one that bolts the fuse into place, I'd change to one that does.

Also the power lines are big thick cables these are the original ones on the bike, after the first hiccup that blew the motor and controller it blew one of these cables where it connects to the controller with a M6 screw

That means there was (is still, possibly) a poor connection, either in the crimp for the lug on the wire, or in the bolt-on connection to the controller itself. This is a common problem, and it causes voltage drop across the connection, which heats it up.

When it's really bad you can see the heat damage, anything from slightly deformed wire insulation to discolored wire / insulation to actual burned insulation, and if it's actually loose, there can be arcing burns from metal to metal. The last one can blow up electronics.

This happens in all high current connections, both battery and motor connections to/from the controler, for instance.

 

[E-HP]

That usually indicates a battery that is dropping in voltage so far that the controller LVC kicks in. That usually happens because the cells are incapable of handling the load, so they sag in voltage during high current draw.

I was trying to look up whether a Sabvoton controller requires disconnecting the power source in order to reset the controller after LVC is triggered, but couldn’t find anything. Do you have any experience or reference to the expected behavior?

 

[amberwolf]

The fuse is between the main power line of the battery and the main power line connector that goes to the controller.

That's where fuses for battery/controller connections always are--somewhere between the two.

But *where* between them is it? You may want to base your decision on fuse purchase based on that--the post I made about that distinguishes between the places for the reasons stated.

You should also be sure the cells in the battery are actually rated for the load they'll be under. If you only rarely use the high current then they only need to be rated for that peak, but if you use it a lot or for more than a few seconds at a time, they should be rated for that continuously to keep them healthy.

Keep in mind that as cells age, they become less capable, as well as have less capacity, so they can't be pushed as hard. If it's a used pack then how hard it's been pushed and it's actual age and condition are unknown, so this is something to keep an eye on. If your connections are all good but you see a lot of voltage sag under load, the cells are probably not able to handle it.

 

[amberwolf]

I was trying to look up whether a Sabvoton controller requires disconnecting the power source in order to reset the controller after LVC is triggered, but couldn’t find anything. Do you have any experience or reference to the expected behavior?

Not for sabvotons***, but all the controllers I've used so far just wake back up once the voltage recovers above some point over the LVC.

(the Phaserunners presently on the trike haven't been tested under that condition, as I haven't run that low with them)

***remember that that name, like "infineon", covers a lot of controller types and designs, as it has been used generically as well as specifically, and more than one manufacturer has claimed to be making that "brand" over the years, so whatever info you find should probably be applied only between exact model and firmware matches. Behavior between other versions could be different.

 

[E-HP]

Not for sabvotons***, but all the controllers I've used so far just wake back up once the voltage recovers above some point over the LVC.

That’s why the described behavior is puzzling to me. I was thinking LVC as well, except for the need to disconnect the battery. I was also thinking poor solder joint or something else internal to the controller.

 

[amberwolf]

The behavior I replied to didn't mention disconnecting the battery.

If that is a required step to repower the system, then the battery's BMS probably shutdown to protect the cells due to a problem with the cells and/or interconnects *inside* the battery, and the whole battery would require testing (and possibly disassembly to do that).

Either way, it's still a battery voltage supply issue. The second one just limits it to something inside the battery, whereas the first could also include poor connections between battery and controller.


EDIT: The second one actually also includes a faulty LVC setting in the controller--if it's set higher than it should be for the battery being used, it could shutdown when the battery doesn't actually have a problem, just because voltage sag at that stage in the discharge curve happened to dip below controller LVC long enough to trigger the controlelr shutdown.

 

[narakutaishi]

I

That usually indicates a battery that is dropping in voltage so far that the controller LVC kicks in. That usually happens because the cells are incapable of handling the load, so they sag in voltage during high current draw.

It can also happen because of poor connections, which includes fuse holders along with all the other connectors and bolt-on connections and crimped connections and switches between the battery and the controller.

Which type of fuse holder do you have? If it's not one that bolts the fuse into place, I'd change to one that does.



That means there was (is still, possibly) a poor connection, either in the crimp for the lug on the wire, or in the bolt-on connection to the controller itself. This is a common problem, and it causes voltage drop across the connection, which heats it up.

When it's really bad you can see the heat damage, anything from slightly deformed wire insulation to discolored wire / insulation to actual burned insulation, and if it's actually loose, there can be arcing burns from metal to metal. The last one can blow up electronics.

This happens in all high current connections, both battery and motor connections to/from the controler, for instance.

Hello thanks again for your knowledge. I twisted the power lead around each side of the fuse then wrapped heat proof tape over each side individually then put insulation taper over that leaving no metal/wire exposed. Obviously not the most professional method but I thought it shouldn't make much difference? (I do have a soldering iron to use). As for the battery I think this small Chinese made unit is likely made with very cheap low power cells and has had more cycles than I'd care to imagine. I think I'll just keep things how they are, bike seems to go well enough in modes 1,2,3, battery upgrade is needed next. I have changed the the display to show me real time voltage so when I next take it for a ride a can watch what voltage sagg the bike has, this will be an indicator. Thanks again for the advice

 

[E-HP]

If that is a required step to repower the system, then the battery's BMS probably shutdown to protect the cells due to a problem with the cells and/or interconnects *inside* the battery, and the whole battery would require testing (and possibly disassembly to do that).

It doesn’t seem like the BMS has shutdown, if the display stays on, which is why it appears to be downstream of the battery, and a fault internal to the controller (the controller is receiving power if the display is on).

 

[amberwolf]

I twisted the power lead around each side of the fuse then wrapped heat proof tape over each side individually then put insulation taper over that leaving no metal/wire exposed.

I'm not sure I understand (I hope I don't).

Do you mean there is no fuse holder, and no mechanical connection of the wires other than basically just "laying" on the surface of the metal fuse contacts, held in place with tape?

That sounds like a fire waiting to happen, when the tape adhesive ages and begins to loosen, and the wire moves away from the surface enough have enough resistance to cause enough heat to melt the tape, which then lets the wires come fully away from the surface to allow arcing, which then can set the tape, wire insulation, and other flammable materials aroudn it on fire, which then sets the bike on fire.

Even without a fire risk, it's a high resistance "connection" that is going to drop voltage across it, so your controller will see a drop in battery voltage that increases the more current you draw. (it will also cause a drop in voltage the other way when using regen). Aside from the power and performance loss negatively affecting the motor response to your inputs, it also means all that lost power at the connection is heating up the fuse and wires connected to it.


I'd strongly recommend using a fuse holder made for that specific fuse.

If it's a bolt on fuse, there should be ring terminals crimped (with the right crimper) or solder-type ring terminals properly soldered (with a high wattage large-size chisel tip iron) to the cables, and then the fuse bolted properly to the ring terminals with all-metal hardware with locking washers. (no nylock nuts; the nylon could deform from heat and loosen the connection). Or use a fuseholder designed for that specific bolt-on fuse, with whatever connections to the holder that are appropriate for the wires you have to connect it to.

If it's some other type of fuse, then what you'll need for a holder depends on it's design. I prefer bolt-ons because they ahve the lowest resistance connections and are the most secure.

 

[narakutaishi]

Thanks very much for that advice I will be ordering a fuse + holder combo

 

[narakutaishi]

UPDATE!

Changed DC current, boost current and rated phase current to 120A MQCON app and runs like a dream, hasn't cut out since.

Swapped the 80A fuse for a 120A fuse

54MPH on private land, very happy

 

[Dui ni shuo de dui]

The difference is that a properly designed controller (especially an FOC type) ought to be able to detect the excessive phase current (especially vs motor speed, decreasing or zero) and shut that down, so it doesn't cook the motor or blow the FETs.

Even if it doesn't have phase current sensors (most non-FOC types won't), it should see a sustained battery current spike vs actual motor speed (decreasing or zero) and shutdown.

At minimum it ought to have some form of temperature monitoring of the FETs and shutdown when those get too hot so they don't blow up.

Most controllers dont' actually have any of those protections, though, and I don't recall any that have all of them. So...effectively it's the "same" as what you're describing.

The Sabvotons controllers have all these protections built in.
I've done lots of rear wheel lock using the brakes on these controllers (no brake wire connected to the controller as well) and never had this issue.

No idea what happened there, its likely that the controller malfunctionned but it is not typical behavior from these.

 

[Dui ni shuo de dui]

Buuut...under extreme conditions near the spec limits for FETs especially, FETs tend to fail shorted, leaving the BMS stuck on, so it can't actually protect against overcurrents. It's a silent failure, so unless you test specifically for this, you won't even know that this problem exists until you encounter a situation where the battery should have turned off but didn't (like overdischarging it because a cell went too low), and probably already damaged the cells in the process.

This contradicts my experience, at least with the ANT BMS
I actually never had one fail shorted, I had one who couldn't activate its mosfets anymore, but they remained closed (Edit: Opened, not closed).
And the thing is able to detect that something is wrong with the mosfets, it will then beep at you continuously to tell you that something's not working as it should, it's not silent anymore

But there are many different brands of BMS and not all are created equal