damaged sabvoton 7280 SVMC7280 (M)

Gkhodeli

100 µW
Joined
Aug 14, 2022
Messages
8
Hi all. I have purchased a kit form NB Power for my Sondors Madmods Scrambler. The kit is this Sabvoton 80a controller (SVMC7280) paired iwth NP power's own motor, which they call fatmotor 3-5k. I asked for 4T winding to get more torque than speed (have lots of hills). I have a 72v batttery with 80a daly bms. The pack is built with 160pcs of p421 Molicells (21700) in a 20s8p configuration amounting to 72v 33.6 ah. I installed the setup and the bike was running beatiful for 20 minutes. I was a happy camper. Parked and the next day I tried the bike the wheel had constant drag and vibration (usually damaged hall sensor). Tried both hall sensors, same issue. Contacted NB power and their technician would not believe there was a probelm with the motor, despite the permanent drag and vibration (plus noise during operation, again, typical of hall sensor damage). The NB power technician kept asking to change some parameters and test again, changed ofset andgles several times, nothing helped. He eventually said the bike was working properly (despite drag and noise and vibration) and asked me to test ride it. During the testing immediately as I gave it throttle the display shut off and wouldn't turn back on. I had to connect and reconnect from the battery (I am using QS8 antispark connectors), but it gave a spark. Tried second time, and it gave a spark again, so I disconnected everything and tested my battery first with a multimeter. Battery tests ok, and the voltage matches what the daly bms displays through the bluetooth app. So now I tessted the controller and the positive and negative wires are short circuited. Also by running basic tests I can see the short circuit on the blue and green phase wires of the controller. NB power did not honor their warranty, though it has been on ly 5 months since purchase (just took me a while to build the battery pack, the kit actually worked only for 20 minutes). So I might be stuck with having to replace the ahll sensors (whatever the NB Power uses, will have to figure that out somehow) and repairing the sabvoton controller, hoiping I only need to replace the mosfets. Could anyone please tell me what mosfets are installed on the 80a sabvoton controller? Many thanks in advance.
 
Unfortunatley when FETs fail, they usually take out the gate drivers, too. Sometimes it even takes the MCU out. :(

For which FETs to use, since you'll probably need to replace them all anyway, use FETs that are suitable for more than the phase current you need, and as much more voltage than you need as possible. If you prefer to use the same as what is in there, you'll have to open it up and check the markings, since different batches (and different models) may use different FETs.

For halls, if they are failed (you can test them with the procedure at ebikes.ca on their learn - troubleshooting page), you can use honeywell ss41 or ss411 to replace them with.


The FETs were probably blown up by whatever the original failure was. If there is a hall signal problem, then the wrong current timing caused by this could easily kill the FETs.

But something had to cause the original problem, and it'd be good to figure that out now, so it can be fixed / prevented from causing future failures.

If it worked perfectly before parking it, and not after, is it possible someone else messed with the bike in any way, intentionally or not? (this kind of thing has happened to others and turned out to be the cause of their problems).

If not, then perhaps there was already a problem, such as with the phase/hall wiring order between motor and controller (which could happen even with a "kit" and even with prewired cabling), so that the timing of the current to the motor is wrong. With some timing problems (some combinations), the motor can appear to operate correctly, even faster than usual sometimes, but get very hot. This heat, especially as it soaks into the halls and stays there a while, can damage them. If this happened, it could work for long enough to pass the test ride, then succumb to the heat while sitting there, and then not work correctly when next used.

A poor connection between motor and controller could cause the hall signals to be unreadable enough to cause the failure, too, so if the cable already had a problem or was bumped or pulled while sitting after the ride that worked....

ESD damage to the controller itself, during manufacturing, etc., could also have caused a problem with any part of it's electronics, that didn't manifest until after it had been used a while, perhaps even at power-on the next day after the first ride.

Even just wrong settings programmed into the controller, or one that changed by itself due to corrupted flash memory (from ESD or even cosmic rays, unlikely as those are they do cause this kind of stuff), can cause all sorts of issues.


Gkhodeli said:
Hi all. I have purchased a kit form NB Power for my Sondors Madmods Scrambler. The kit is this Sabvoton 80a controller (SVMC7280) paired iwth NP power's own motor, which they call fatmotor 3-5k. I asked for 4T winding to get more torque than speed (have lots of hills). I have a 72v batttery with 80a daly bms. The pack is built with 160pcs of p421 Molicells (21700) in a 20s8p configuration amounting to 72v 33.6 ah. I installed the setup and the bike was running beatiful for 20 minutes. I was a happy camper. Parked and the next day I tried the bike the wheel had constant drag and vibration (usually damaged hall sensor). Tried both hall sensors, same issue. Contacted NB power and their technician would not believe there was a probelm with the motor, despite the permanent drag and vibration (plus noise during operation, again, typical of hall sensor damage). The NB power technician kept asking to change some parameters and test again, changed ofset andgles several times, nothing helped. He eventually said the bike was working properly (despite drag and noise and vibration) and asked me to test ride it. During the testing immediately as I gave it throttle the display shut off and wouldn't turn back on. I had to connect and reconnect from the battery (I am using QS8 antispark connectors), but it gave a spark. Tried second time, and it gave a spark again, so I disconnected everything and tested my battery first with a multimeter. Battery tests ok, and the voltage matches what the daly bms displays through the bluetooth app. So now I tessted the controller and the positive and negative wires are short circuited. Also by running basic tests I can see the short circuit on the blue and green phase wires of the controller. NB power did not honor their warranty, though it has been on ly 5 months since purchase (just took me a while to build the battery pack, the kit actually worked only for 20 minutes). So I might be stuck with having to replace the ahll sensors (whatever the NB Power uses, will have to figure that out somehow) and repairing the sabvoton controller, hoiping I only need to replace the mosfets. Could anyone please tell me what mosfets are installed on the 80a sabvoton controller? Many thanks in advance.
 
amberwolf, thank you so much for the information. YOu have provided very detailed and realistic scenarios, out of which I believe the following is what happened: "With some timing problems (some combinations), the motor can appear to operate correctly, even faster than usual sometimes, but get very hot. This heat, especially as it soaks into the halls and stays there a while, can damage them. If this happened, it could work for long enough to pass the test ride, then succumb to the heat while sitting there, and then not work correctly when next used." And once the hall sensors got damaged and the motor aquired constant drag and vibration when turning the technician wouldn't believe the motor was faulty and kept asking for further tests under load. So from what I understand, what I need is to replace hall sensors with better quality ones if possible (higher heat resistance) and add statorade tot he motor when tested as reparied. Are the ones you suggested same or better than what is usually installed on these motors in China? Is there one best make of the halls I can use?
Regarding the mosfets, I found very little literature but so far what I have inquired is that the sabvoton 7280 uses IRFB4115PBF (TO-220 150V 104A). Here the voltage is more than double I need (150v as opposed to 72v) and amperage is 104a (as opposed tot he 80a controller). I did think to install stronger fets, but was afraid that in case of future issues something else, more problematic and expensive to replace would blow if the fets let the current though. BUt you are clearly way more qualified (my level is 0.1) so if this is ok to do by all means I would prefer to beef things up. When searching online i see a wide range of price difference. These mosfets cots about 4usd each on RS-components (which would amount to the total cost of new controller if I was to replace all mostfets) which is a european site, but the same can be purchased for 50cents per pc on alibaba/aliexpresss. There is a high likelihood that the euro or US sellers that have these mosfets come from the same exact chinese factories. I would greatly apprciate if you have a link for quality mosfets with realistic prices. Thank you very much for the very informative response. Cheers!
 
Gkhodeli said:
And once the hall sensors got damaged and the motor aquired constant drag and vibration when turning the technician wouldn't believe the motor was faulty and kept asking for further tests under load. So from what I understand, what I need is to replace hall sensors with better quality ones if possible (higher heat resistance) and add statorade tot he motor when tested as reparied. Are the ones you suggested same or better than what is usually installed on these motors in China? Is there one best make of the halls I can use?
Even the halls that came in it can take a lot of heat; you won't find parts that are much better at that part. If the motor gets hot enough to cause a problem with them, it's too hot because the system is not working right, or the motor is being used too hard for it's capabilities.

The problem causing the heat needs to be found and fixed to prevent it from causing other, even worse damage (burning motor windings, demagnetizing magnets, etc).

If the scenario of wrong phase/hall combo is the cause, it's easy to fix, once the controller is working or replaced. Right now, don't worry about that becuase you can't test for it. :/

If you get genuine SS41 or SS411 halls from a reputable supplier (farnell, digikey, mouser, etc), they'll work fine (assuming ESD precautions are used during all handling and repair of all parts involved).

But before ordering / replacing any sensors, test them. They might not be the problem, and it would kinda suck to have opened the motor and done all that work to not have fixed anything. ;)


Regarding the mosfets, I found very little literature but so far what I have inquired is that the sabvoton 7280 uses IRFB4115PBF (TO-220 150V 104A).
To be certain of what's in there, you'd have to actually open your controller to find out what FETs it really has--it's very common for all sorts of devices to not have what they say they do, either because the manufacturer cheaped out and went with whatever was available that cost less, or because the parts they bought were counterfeit (marked as something they aren't; it's also a really common problem and a good reason to only buy from well-known reputable suppliers, vs ebay, aliexpress, etc--you just can't know what you are going to get otherwise).

If you're replacing them all, it doesn't really matter what's in there, though. ;)


Here the voltage is more than double I need (150v as opposed to 72v) and amperage is 104a (as opposed tot he 80a controller). I did think to install stronger fets, but was afraid that in case of future issues something else, more problematic and expensive to replace would blow if the fets let the current though.
I definitely recommend putting FETs at least as good as what should be in there. The reason for higher voltage rating is because there can be spikes of voltage back from the motor during some usages and events, and if they are not able to handle that voltage, they fail. :(

Also, system voltages are more complicated than they seem: your system is not really "72v". If you have a "72v" 20s (20 cell groups in series) lithium battery, then it is about 84v fully charged, and about 60-66v empty, depending on the BMS LVC.

The current is also more complicated--80A is the battery current rating; this is what most controllers are rated by. (A very few, like Kelly, rate by peak maximum motor phase current, everybody else rates by continuous max battery current, because that is much easier to determine and measure under various conditions, even though the more advanced of these does actually measure phase current internally to do what it needs to do).

The FETs don't see battery current, they see motor phase current, which because of the way the controller switches voltage and current to the motor to drive it, can be much higher, so the FETs need to be able to handle that without heating up too much. The more there are in parallel, the easier that is (which is why you see 6FET, 12FET, 18FET, etc for bigger and bigger controllers), but they still have to have a sufficient per-FET rating. Safer to go higher than lower, and if there is a rating already known, go with at least that much. Might be overkill...but that's better than the ragged edge that could fail at the worst possible moment.


When searching online i see a wide range of price difference. These mosfets cots about 4usd each on RS-components (which would amount to the total cost of new controller if I was to replace all mostfets) which is a european site, but the same can be purchased for 50cents per pc on alibaba/aliexpresss. There is a high likelihood that the euro or US sellers that have these mosfets come from the same exact chinese factories. I would greatly apprciate if you have a link for quality mosfets with realistic prices.
As noted above, counterfeit parts (that cant' do the job they claim, if they even work at all) are common on the cheap sites, even if they are not priced cheaply. The big parts companies need to vet their suppliers and buy from the actual manufacturers, because if they contaminate their supply chain with counterfeit parts, they open themselves up for serious problems with the large companies they sell to. So you're much more likely to get parts that will do the job you need them to from them. ;)

If you look around at some recently-posted-to threads about repairing devices, there's posts by I think "whatever" that talk about the bad parts received from aliexpress/etc. that didn't work, and there are others around over the years. Sometimes you get lucky and the parts you get work well enough for the purpose, so you can try it if it's cheap enough for you to risk that money and time. I think one set of the parts he got did fix his charger; that's luck. ;)


It is unfortunate but common that a new device will be cheaper than fixing a broken one; this is partly because of bulk pricing vs single-part pricing, etc.

Even if you buy the more expensive FETs, it's easily possible the controller still wont' work because of blown gate drivers (which you might be able to fix if you do enough electronics troubleshooting) or even a damaged MCU (which you can't fix). :(


If you do have to replace the controller, you might consider a sensorless-capable (that can still use hall sensors if the motor has them) controller so that you can use it whether or not your motor's sensors work, and save the trouble of opening the motor up for that part.
 
Again, this is sea of information, and all most useful. Thank you very much. The I ordered genuine honeywell SS411P sensors (10pcs, sohuld only need six). I tested the sensors and I believe all are cooked. All are showiwng zero volts contantly, with the wheel stoppped or in motion. That replacement doesn't scare me, and I believe that is what happened. The bike is not light (about 130-140lb with all mods and I am around 190, so a total of 320lb. The motor is rated 3000w to 500Ok, but as you said, I was runningit much higher and going uphill when it got hot. I was controlling the BMS temp sensor (dave a daly 80a bluetooth bms) and it was always very low, but by the time I stopped to test the motor it was hot enough that I could not hold my hand on it for more thana touch. I was pushng about 80vx80a so about 6400W uphill. The motor was pulling beautifully, there was no voltage lag or drag before it overheated, I hoped there would be some protection for overheating. I did set temp parameters for low and high temp on my bms, and my battery is still working fine, but mess with any factory parameters until the motor started having problems and I contacted NB Power, at which pint the technician had me troubleshoot under load. I believe this is what happened: I asked them to test the kit before shipping, and repare it for plug and play so I wouldn't have to do anything. They clam to have done the hall test before shipping, but they didn't shoot and send a video, so they may have tested my motor with another "test" controller, and not the exact one they sent me. In that case the parameters could have been mismatched or not ideal for the motor, and while it was spinning fine and the halls were working the motor overheated easy, to the extent that all sensors got damaged. Otherwwise why and how would all the sensors die all together, how could I have overheated the motor in 20 minutes so much, I first did about a 5 mile downhill, and then came back up. The whole time the motor was pulling well and I was at about 40-50km on the uphills. The gradient is not that high, maybe from 7 to 14 degrees, it's just a constant ascent of 5 miles, but still, should this have happened if the controller was matched and properly set up with the motor? Thanks again
 
Gkhodeli said:
The I ordered genuine honeywell SS411P sensors (10pcs, sohuld only need six). I tested the sensors and I believe all are cooked. All are showiwng zero volts contantly, with the wheel stoppped or in motion.
How was the test performed, exactly? To get a valid signal, you need a pullup resistor to 5v from the hall signal output (this is built into the controller, but you don't have that available right now with a dead controller), and 5v power supplied to the power wire of the sensors, and ground from the 5v power source to the ground of the sensors. Motor hall sensors don't output anything, they only ground their outputs when a magnet passes. The latching type usually used stays at the last output state until a magnet of opposite polarity passes.

The bike is not light (about 130-140lb with all mods and I am around 190, so a total of 320lb. The motor is rated 3000w to 500Ok, but as you said, I was runningit much higher and going uphill when it got hot. I was controlling the BMS temp sensor (dave a daly 80a bluetooth bms) and it was always very low, but by the time I stopped to test the motor it was hot enough that I could not hold my hand on it for more thana touch. I was pushng about 80vx80a so about 6400W uphill. The motor was pulling beautifully, there was no voltage lag or drag before it overheated, I hoped there would be some protection for overheating.
Unfortunately most systems don't have anything like that. Some do, but it would be something you'd usually need to set limits for based on your system and usage, and verify it is reading correctly within the controller to be sure it works before pushing it too hard.

Under the right conditions, you can run a motor for a short time at significantly higher than it's design limit--but when running the motor at high load but low RPM like usually happens climbing hills quickly overheats things and damages not only the sensors but the windings themselves, especially if this is done for very long. If the motor was able to stay near it's typical max speed while at that high wattage, then it wouldn't overheat as quickly, but if it is slower, especially a lot slower, it can melt down fairly rapidly. Heat takes time to do damage, so it may not fail immediately, and only actually fail (shorted windings and/or failed sensors).

What is the max speed on flat ground that your motor was able to do? Or, what is the max speed unloaded, off ground, that your motor was able to do? (before the problems began)

If there is a significant difference between that speed and the speed you were climbing the hills with, the motor could draw more current than it was capable of handling and overheat. How fast that happens and how severe it is depends on a bunch of things, so different sitautions and systems react differently. The http://ebikes.ca/tools/simulator.html can help you see how this can happen.

Sometimes it's even just the soft insulation on the cable that passes thru the axle, not melting until the axle itself heats up enough to do that...then the wires inside that cable (phases and halls) can short to each other, which can blow up the controller immediately, or if the short is not a "hard short" between phases, just a high-resistance short between a phase and a hall, or two halls, etc., it could just alter teh signals enough to cause the controller to not drive the motor correctly, causing it to heat more and create further shorting, eventually blowing the controller.

But I would bet on the windings of the two phases with blown FETs being shorted together, either at the stator windings themselves, or the phase wires inside the axle.

So I don't think you have any miswire or hall/phase combo problems, just unintentional / unknowing severe over-powering of the motor resulting in motor overheating and damage, and subsequent controller failure. :(

If you disconnect the motor from the controller, does it spin freely? Or does it feel different than it used to? If it does, there may be winding damage, and *that* could be what destroyed the controller.

If the motor windings/etc are damaged, you will probably want a bigger motor, like one of the QS205 or preferably much bigger motors from QSMotor, with the highest number of turns available (they'll make whatever you want; more turns is slower than less turns all other conditions the same), with a thermal sensor built into the windings and a controller capable of shutting down when it gets too hot, and then setting that limit as low as practically possible to prevent another failure.
 
Also, regarding "If you do have to replace the controller, you might consider a sensorless-capable (that can still use hall sensors if the motor has them) controller so that you can use it whether or not your motor's sensors work, and save the trouble of opening the motor up for that part." - my motor has constant drag. even when unplugged from everythiing, so the problem is not with the controller only. The motor should roll freely, but it has constant drag and this drag intensifies in equal intervals when I roll the wheel. Not even sure the hall sensors will fix the issue. The motor has some sort of constant magnetic resistance whech gets weaker and stronger as your rotate the wheel, as if passing some sort of magnetic threshhold. Could one of the phaze wires be shorted?
I have detailed videos that I shot while showing the problems to the NB power technician. I have an ongoing claim with NB power through paypal and I uploaded all evidence documents as well as the transcripts of my communication with NB Power. here is the google drive link if anybody is interested in the experience I've had with NB power (not good): https://drive.google.com/drive/folders/1oMWYg3WR1YpjLVWwa5_kAMGVsnwtNgQ4?usp=sharing

Cheers!
 
very good questions inddeed. The motor was able to do max 70-75km on straigth with me on it. It did not bog down when going up hills, of course it lost speed bu I was still going between 40 and 50 km. As I said the hills are not very steep, it's just a constand incline for about 5 miles. The motor was going well the whole time. It still overheated. By the way, good thing I remembered, the motor got very hot when I was trying to do hall teest. When motor started having problems I contacted NB Power and tried to do hall test but it would not complete, but when. I put my hand on the motor and it was hot. I also told this to the technician and he kept saying the moto was fine. There are detailed videos of what tests he asked for on the following link: https://drive.google.com/drive/folders/1oMWYg3WR1YpjLVWwa5_kAMGVsnwtNgQ4?usp=sharing
 
The last part of my post above covers the motor part. Unfortunately it sounds like you'll probably need a new one. :( You can still open it up to check the wiring and windings, but I'd bet on a cooked level of "overdone". :(

Gkhodeli said:
Also, regarding "If you do have to replace the controller, you might consider a sensorless-capable (that can still use hall sensors if the motor has them) controller so that you can use it whether or not your motor's sensors work, and save the trouble of opening the motor up for that part." - my motor has constant drag. even when unplugged from everythiing, so the problem is not with the controller only. The motor should roll freely, but it has constant drag and this drag intensifies in equal intervals when I roll the wheel. Not even sure the hall sensors will fix the issue. The motor has some sort of constant magnetic resistance whech gets weaker and stronger as your rotate the wheel, as if passing some sort of magnetic threshhold. Could one of the phaze wires be shorted?
I have detailed videos that I shot while showing the problems to the NB power technician. I have an ongoing claim with NB power through paypal and I uploaded all evidence documents as well as the transcripts of my communication with NB Power. here is the google drive link if anybody is interested in the experience I've had with NB power (not good): https://drive.google.com/drive/folders/1oMWYg3WR1YpjLVWwa5_kAMGVsnwtNgQ4?usp=sharing

Cheers!
 
Only being able to go half to 3/4 of the full speed, but still at the same loading / power usage means quite a lot of current in it, and so quite a lot of heat.

Part of the problem is that the difference in the power needed to go the speed you were actually going, vs the power actually being used, ends up as waste heat in the motor. The simulator can show you some of how this works, with some experimentation with it.





Gkhodeli said:
very good questions inddeed. The motor was able to do max 70-75km on straigth with me on it. It did not bog down when going up hills, of course it lost speed bu I was still going between 40 and 50 km. As I said the hills are not very steep, it's just a constand incline for about 5 miles. The motor was going well the whole time. It still overheated. By the way, good thing I remembered, the motor got very hot when I was trying to do hall teest. When motor started having problems I contacted NB Power and tried to do hall test but it would not complete, but when. I put my hand on the motor and it was hot. I also told this to the technician and he kept saying the moto was fine. There are detailed videos of what tests he asked for on the following link: https://drive.google.com/drive/folders/1oMWYg3WR1YpjLVWwa5_kAMGVsnwtNgQ4?usp=sharing
 
Only being able to go half to 3/4 of the full speed, but still at the same loading / power usage means quite a lot of current in it, and so quite a lot of heat.

Part of the problem is that the difference in the power needed to go the speed you were actually going, vs the power actually being used, ends up as waste heat in the motor. The simulator can show you some of how this works, with some experimentation with it.

FWIW, a motor (or any other device) shouldn't have any heat at all just sitting there doing no work (like during a hall test). Heat is only generated doing work, or from a fault (short).


Gkhodeli said:
very good questions inddeed. The motor was able to do max 70-75km on straigth with me on it. It did not bog down when going up hills, of course it lost speed bu I was still going between 40 and 50 km. As I said the hills are not very steep, it's just a constand incline for about 5 miles. The motor was going well the whole time. It still overheated. By the way, good thing I remembered, the motor got very hot when I was trying to do hall teest. When motor started having problems I contacted NB Power and tried to do hall test but it would not complete, but when. I put my hand on the motor and it was hot. I also told this to the technician and he kept saying the moto was fine. There are detailed videos of what tests he asked for on the following link: https://drive.google.com/drive/folders/1oMWYg3WR1YpjLVWwa5_kAMGVsnwtNgQ4?usp=sharing
 
Thank you very much again. I guess I'll have to wait for Paypal's response untill I open up anythng because I might have to send it back depending on what aliexpress decides. But thansk for all the useful insight, your really helped.
 
I have been looking for a digram for the 3000w 5t winding. I am thining of rewinding the motor myself, I can easily get the copper wire, but don't know what gauge I would need and how much of it, what the winding scheme is, etc. My current motor is supposed to be 3000w 4t winding, I could replicate, but don't know how the 5T winding would change the scheme, do I use same gauge wire, just different number of turns, etc. Any tips would be much appreciated. Cheers
 
If you are careful during the unwinding, you can note down how many turns of wire are on each tooth and how they are interconnected, so you can rewind it with the same pattern but more or less turns as you wish.

More turns gives higher torque but less speed, all other conditions the same, and less turns higher speed but less torque.

You can also get higher speed by increasing system voltage (higher voltage battery), and higher torque by increasing phase current (bigger controller), without changing anything in the motor itself.

You can also get higher speed without changing system or motor by using a larger diameter tire, but this reduces torque. Alternately, using a smaller diameter tire increases torque but reduces speed.


You can probably find out what wire to use in the various motor rewinding threads around the forum, or the DIY motor design / build threads; I don't know offhand what would be best for your usage.


Gkhodeli said:
I have been looking for a digram for the 3000w 5t winding. I am thining of rewinding the motor myself, I can easily get the copper wire, but don't know what gauge I would need and how much of it, what the winding scheme is, etc. My current motor is supposed to be 3000w 4t winding, I could replicate, but don't know how the 5T winding would change the scheme, do I use same gauge wire, just different number of turns, etc. Any tips would be much appreciated. Cheers
 
Check this video out: https://www.youtube.com/watch?v=k4V8zBjfYt0

NB Power sent me a 3T wound motor (not even consistently 3T, there are several instances where it was only 2T) instead of the 4T I ordered. The winding was inconsitent and all over the place, so I couldn't replicate that.
 
Back
Top