Help troubleshooting electric mini bike that intermittently blows fuses

[Balmorhea]

Well shoot, I had no idea caps had a positive and negative

Not all of them do! But all of them that matter to us, do.

 

[bigedgar]

The hole in the case thing only works easily if the top comes off the case, not if the board has to slide into it from the end.

If it goes in from the end, you'd need a long slot as wide as the caps.

The top of the case "flips" open, so fortunately I can drill the holes and flip it back into place.

 

[bigedgar]

A relatively cheap solution if you can find one is a Kelly KDS48050E 24-48V 50A programmable controller, and be sure to get the programming cord if you can't make one up yourself, or already have what's needed.

If you want to go the ultra-cheap route that requires some learning about electronics, and you are willing to put the effort into tuning it to your system, you could DIY a controller using something like this as your base https://www.amazon.com/Controller-1...=psdc_306530011_t1_B07D31MGFF#customerReviews

Thanks for the thoughts, John. I hadn't had any luck sourcing what appeared to be a reliable controller that looked like it could handle a 36v / 1000w system - lots of options for 36v / 500w, or 48v / 1000w, but nothing right in my wheelhouse. The Kelly controller looks like a good potential alternative if my current efforts to re-vamp the stock controllers don't pan out. And the bare controller board is yet another option that I'd love to have in my quiver, so to speak, I'll keep that in mind. Thank you.

 

[John in CR]

And the bare controller board is yet another option that I'd love to have in my quiver, so to speak, I'll keep that in mind. Thank you.

For just $16 it's something I'd try. Since all of the mosfets have individual heat sinks, you could mount that board in your own box with a fan, and it should work fine based on customer comments I saw. The seller seems supportive too and may be able to direct you on how to limit current. I'd find out if there's a way to hard wire or program an LVC (Low Voltage Cutoff) and for safety an Ebrake cutoff too, especially since brushed motor controllers can fail in WOT (Wide Open Throttle) state.

 

[amberwolf]

and for safety an Ebrake cutoff too, especially since brushed motor controllers can fail in WOT (Wide Open Throttle) state.

An ebrake is not useful for brushed controller WOT failure, because the FETs are actually shorted, making an effectively-direct connection between motor and batteries.

It requires a full-current-at-full-voltage capable electrical disconnect to do that. (if it cant' handle full current at full voltage it could weld together at the moment it's needed most). Contactors are one way,
kill switches are another, but a cheap and easy way is a ripcord or T-handle attached to a connector with a wire loop from one contact to the other, that plugs into a matching connector bolted to the frame, that's wired into the main battery positive (or negative) wire, breaking it.

 

[John in CR]

Very true AW. I had 2 thoughts run together while typing that sentence. Both are really needed. An ebrake is an essential safety feature of a bike with multiple users, which would best be done at the controller, though one could rig something where the ebrake cuts the throttle input. In addition, some kind of battery main interrupt is needed like AW pointed out.

 

[bigedgar]

Update: I was successfully able to get the capacitors re-soldered with the positive/negative in the right spots, and with the posts as close to the board as possible. The good news is that the motor spins without blowing fuses. The bad news, however, is that I seem to have lost throttle control. What I mean by that is that as soon as I connect the controller to the battery, motor, throttle control, etc., the motor kicks on full-bore. What happened with my pre-modded controller is that I connect everything up, and the motor doesn't spin until I a) turn the power button on the throttle control to "on", and b) turn the throttle.

I put a switch into the circuit so that I could have some more control over turning things on and off (it's just a regular household AC light switch - I don't think that's part of the problem, but I could be wrong). You can see in the gif below that as soon as I complete the circuit by turning on the switch, the motor kicks on full throttle.

https://imgur.com/a/6OdBair

I replaced the caps on two separate controllers and both had the same result (full throttle once all connections were made). My guess is that something is wrong with my modded controllers. One of the modded controllers was effectively brand new, the other was one that had already blown the prior set of caps (and perhaps sustained some damage to the mosfets in the process).

A suspicion that I have is that something is amiss with my soldering job for the capacitors. I'm not super experienced with soldering, but this board has a setup I haven't seen before where the solder appears to be on the underside of the board, and there's some kind of foil tape over all the solder joints. You can see what I mean in this picture:



In order to de-solder the old caps, I had to heat the foil tape, which then heated up the solder underneath the tape and allowed me to pull the old caps off. I did the same process to solder the new caps. I'm wondering if by heating up the foil tape, I may have unintentionally heated some solder for a different component and caused a short. Does that seem plausible? Or is there, perhaps, a much simpler explanation for my loss of throttle control?

As always, I'm grateful for any thoughts on this. Thanks!

 

[amberwolf]

The bad news, however, is that I seem to have lost throttle control. What I mean by that is that as soon as I connect the controller to the battery, motor, throttle control, etc., the motor kicks on full-bore. What happened with my pre-modded controller is that I connect everything up, and the motor doesn't spin until I a) turn the power button on the throttle control to "on", and b) turn the throttle.

That usually means the FETs are blown, shorted on, for a brushed motor. Or it can simply mean the ground wire to the throttle is broken.

You can see in the gif below that as soon as I complete the circuit by turning on the switch, the motor kicks on full throttle.
https://imgur.com/a/6OdBair

I can't see the image because the external site is blocking me, but if you upload it to the forum with teh attachements tab then anyone that can see the post can see the pic.

replaced the caps on two separate controllers and both had the same result (full throttle once all connections were made). My guess is that something is wrong with my modded controllers. One of the modded controllers was effectively brand new, the other was one that had already blown the prior set of caps (and perhaps sustained some damage to the mosfets in the process).

There is no reason that replacing the caps would cause this. Even if they were connected backwards, all that would happen is they would explode shortly after power was connected, if the fuse didn't blow first.

So something else is different now than it used to be...most likely the connector, contacts, or wires, in the throttle itself.

there's some kind of foil tape over all the solder joints.

It is not tape. It is solder, over the entire surface of the PCB traces. It's common in cheaper controllers to do this to "thicken" the traces to carry more current than they could with just the bare copper. (better controllers use busbars or thicker copper traces, etc., to do this job).

I may have unintentionally heated some solder for a different component and caused a short.

While that is possible, it isn't very likely taht it would cause full throtlte operation--it would more likely prevent it form working at all. If there were a short big enough to cause full throttle operation, you'd see it across the wide soldered traces (the ones that look taped), shorting between them, either on this side of the board or the other.

 

[bigedgar]

Thanks for the thoughts, as always. I appreciate the explanation on the solder tape/glob/whatever it is - I had never seen that before. I don't like the fact that I can't see where things are connected :x

I had to put the GIF on an external site b/c it's too large for the forum (it's an animated GIF). It just shows me turning the switch on and the motor going full bore, with no throttle control.

I considered that I might have a short in the throttle wiring, but I haven't touched the throttle except to connect the tamiya connector to the controller to test out my modded controllers, so I thought it was unlikely that it was the cause of the problem. I have a new, un-modded controller that I can use. I'll try that one out tomorrow and if I get the same results (open throttle), that will inform me that the issue is in the throttle connection and not on the controller.

Thanks!

 

[amberwolf]

I don't like the fact that I can't see where things are connected :x

But you can.

Everything that is on the same soldered trace is connected to everything else on the same soldered trace, just as if it were on one that was bare copper covered with the green (or whatever color) coating. Also, you can see all the part leads sticking up thru the solder, at least as little bumps, and usually as complete cutoff lead tips.

I considered that I might have a short in the throttle wiring, but I haven't touched the throttle except to connect the tamiya connector to the controller to test out my modded controllers, so I thought it was unlikely that it was the cause of the problem.

As noted before, it's probably not a short, but rather an open circuit, broken wire or pushed out contact, etc. This can happen from plugging and unplugging stuff, which you've had to do a fair bit with this system.

But it could also be FETs blown in the common way that leaves them permanently on.

 

[bigedgar]

Ah, so everything under one of those "tape" sections is effectively connected together?

I connected everything up to a stock, unmodified controller today and everything works as expected. I assume from this that the throttle wiring is all intact and there is something wrong with the controller. I suppose it is possible that I blew up the MOSFETs when I was soldering in the new capacitors. I know it's not easy to test the MOSFETs, it may be time to point on this controller and find another option.

 

[amberwolf]

Ah, so everything under one of those "tape" sections is effectively connected together?

Yes, because it is *not* tape.

It is solder (which like other metals used in electronics is conductive, if a bit less so than some), coating all of the trace and component legs it is touching, and connecting them all. Solder does exactly the same thing on all the other connections on the board--it connects the legs to the traces. Just that in this case, solder covers all of the trace, too.

I connected everything up to a stock, unmodified controller today and everything works as expected. I assume from this that the throttle wiring is all intact and there is something wrong with the controller.

This is the most liekly thing..but if it's a contact that is backing out of the ocnnector as you plug it in, on either the throttle *or* the controller side, you could still see this result.

If you *gently*!! pull on the wires into the throttle ground connection, on both controller and motor, on each controller and on the throttle, if any of them are loose, they may pull out of the housing and show you where the problem is. Don't pull hard, or you will cause a probem that didn't exist before.

I suppose it is possible that I blew up the MOSFETs when I was soldering in the new capacitors. I know it's not easy to test the MOSFETs, it may be time to point on this controller and find another option.

It's unlikely that you could've damaged the FETs durign this process, unless the entire board around them was heated so hot you couldn't touch it without actually burning yourself, or you built up a lot of static electricity (like wool socks carpet and doorknob will show) and zapped the FETs. Neither one of those is likely.


It is actually very easy to test the FETs. http://ebike.ca has a Learn tab, Troubleshooting section, that shows one way to do this.

However...another test is simply to disconnect the throttle, and power on the controller without one. If it still starts running the motor when power is connected, it's not a throttle connection issue, and is probably (but not guaranteed) to be the FETs.

 

[bigedgar]

However...another test is simply to disconnect the throttle, and power on the controller without one. If it still starts running the motor when power is connected, it's not a throttle connection issue, and is probably (but not guaranteed) to be the FETs.

Great idea! I connected one of my modified controllers to just the battery and the motor and as soon as I completed the circuit the motor kicked on full bore. That would seem to narrow down the issue to the controller. I will look into testing the MOSFETs. If nothing else I'll learn something else cool. Thank you again!

 

[bigedgar]

Well, I'm giving up on rehabbing my existing controllers. Testing the MOSFETs has turned out to be a real challenge - I can't seem to get consistent readings across any of my blown or known-good controllers. Time to cut bait and move on.

I'm now going to look for a better controller that suits this build. Does anyone have any recommendations? Here' are the specs I'm looking for:

• 36v
• 1000w brushless motor
• Must be able to handle at least to 50 amps
• Must have low-voltage cutoff
• Ideally it can integrate easily with my existing components (throttle, battery charger, brake) - here's a link to the existing, under-powered controller: https://megamoto.co/products/70-10003-00-controller-36v-1000w.html. If not I'm going to need to figure out how to connect things up (example - the Kelly KDS48050E looks like a solid controller, but its connections don't look anything like my existing connections).

John recommended the Kelly controller, which looks quality, but looks like it'll be some tinkering to get it programmed and connected.

Are there other controllers out there that fit my specs that you would recommend?

Thanks for any thoughts.

 

[Balmorhea]

I'm now going to look for a better controller that suits this build. Does anyone have any recommendations? Here' are the specs I'm looking for:

• 36v
• 1000w brushless motor
• Must be able to handle at least to 50 amps
• Must have low-voltage cutoff
• Ideally it can integrate easily with my existing components (throttle, battery charger, brake) - here's a link to the existing, under-powered controller: https://megamoto.co/products/70-10003-00-controller-36v-1000w.html. If not I'm going to need to figure out how to connect things up (example - the Kelly KDS48050E looks like a solid controller, but its connections don't look anything like my existing connections).

Finding a 36V controller that operates at 50A or more puts you in a whole 'nother price range from what you have been working with. I've found 45A controllers for cheap, though.

https://www.ebay.com/itm/36V-48V-1500W-45A-Brushless-Motor-Controller-For-E-bike-Scooter-Electric-Bicycle/174204584859

If I were you, I'd be getting rid of that lead battery pack ASAP.

 

[amberwolf]

[*]1000w brushless motor

This means you must be changing out the motor, too, because the link you give is for a *brushed* controller, which means you have a brushed motor, which a *brushless* controller cannot run.

If you are only changing the controller, you must make sure it is a *brushed* controller, if your existing motor is *brushed*.

Also, if you are looking for 1000w at 36v, you're looking for at most a 30A controller. A 50A controller will allow the motor to pull almost twice that amount of power, which could damage your motor if it's not well-cooled. And it's going to be pretty hard on your batteries, which will sag in voltage more (probably a lot more) than at the lower current rating. (where they probably already sag a lot).

[*]Ideally it can integrate easily with my existing components (throttle, battery charger, brake) - here's a link to the existing, under-powered controller: https://megamoto.co/products/70-10003-00-controller-36v-1000w.html. If not I'm going to need to figure out how to connect things up (example - the Kelly KDS48050E looks like a solid controller, but its connections don't look anything like my existing connections).

Pretty much every controller has different wiring. There's no standards. Most of them hook up to the same types of stuff, but the actual connectors, wire colors, and even what features they have wires for, may be very different from one to the next.

So no matter what you get, you're likely to have to deal with that.

Also, note what kind of throttle you have--if it is a potentiometer throttle, you will want to be sure the new controller can deal with that, or you may have a big deadband at top and botth of throttle, and less usable rotation to control motor with (can be fixed with adding some resistors, or changing the throtlte to a hall type, but that's one more thing to do). If it's a hall throttle, then the ocntroller needs to handle that type, as if it's meant for a pot you won't be able to go down to zero throttle, and the motor will always be running when powered on.

 

[bigedgar]

Shoot, my mistake - it's a brushed motor (two wires), not brushless.

Amberwolf - It seems that my motor occasionally demands more than 30A, which is why I was looking for a controller that could handle more than 30A. My current controller brand is supposedly rated for 30A, but as you know it keeps on blowing up. Is your suggestion to find a better, higher quality 30A controller?

And thanks for the feedback on the connections being different across many controllers - that seems like something I'll be able to figure out. As for the throttle, I have is this one: https://www.monsterscooterparts.com...-the-monster-moto-classic-1000w-mm-e1000.html. I can't tell if it's potentiometer based or hall.

Balmorhea - thanks for the link. Is eBay really the place where most people get their controllers? If that's the case, then so be it. My only hesitations with eBay are quality and support - I want to make sure I get a controller that is well built, and if there are any issues, I want the merchant to stand behind their products. Maybe that's the case with controllers off eBay...

Thanks for the thoughts!

 

[amberwolf]

Amberwolf - It seems that my motor occasionally demands more than 30A, which is why I was looking for a controller that could handle more than 30A. My current controller brand is supposedly rated for 30A, but as you know it keeps on blowing up. Is your suggestion to find a better, higher quality 30A controller?

If the motor/controller is drawing a lot more than 30A (I forget how you measured that) from the battery constantly, then the controller isn't doing it's job right. Most controllers can handle and allow short bursts of higher current than their limit, sometimes up to twice as much for a second or two before it limits, if their components are capable of handling that, and it doesn't see that kind of burst repeatedly frequently.

As an aside, if it does use a lot more than 30A a lot then if you do replace it with one that actually restricts it to that, the torque of the motor will be less, reducing performance of the system by the difference.


The question is still..."why" is it blowing up? If it doesn't have fast enough current limiting (or doesn't really have good limiting), then that could be the cause, if the motor suddenly spikes current draw.

A bad connection can arc, generating RF, or brush noise (which *is* arcing, and thus generates RF) can also feed back into the FETs and cause failures, regardless of current capability, if there's nothing in the controller design to filter it out. (like if the caps fail, or there arent' enough of the right type and value).

So, if all the connections are good, then do the controllers blow up only after being used a while? If so, it might be that the caps are failing *first*, and then allowing RF from brushes in there, or simply allowing too great a current flow back and forth (which the caps are there to mitigate) causing voltage spikes and dips on the the FET supply bus, possibly high enough in voltage to go beyond their limits, and then they blow.


A well-engineered controller with crappy caps will still fail, and a poorly engineered controller with good caps may not. Better FETs helps, too, but better design, gate drivers, layout, etc., will do a lot to make a better controller under high (for it) loads.

Generally a better controller will cost more, sometimes a lot more. But cheap controllers can last a really long time just fine, unless parts (like caps) in them are crappy and fail, or there is a system problem causing the failures.

WHich one...I don't know, but bad caps are a REALLY COMMON problem, ever since the Capacitor Plague a couple of decades ago, because manufacturers use wrong formulations for the elecrolyte, so it boils, and blows up the cap, etc. Counterfeit parts are also a really common problem, so a part (like a FET) may be marked as some good name brand, but actually be some other part wtih much lower specs, or even defective, or even worse may not even be that *kind* of part (though that last two would result in the system it's in not operating at all).


Sorry for all the rambling. Not sure what you should try next, but a better brand *should* last longer.

https://www.monsterscooterparts.com...-the-monster-moto-classic-1000w-mm-e1000.html. I can't tell if it's potentiometer based or hall.

Almost certainly hall based, just based on it's looks, but you can test the voltage output when hooked to a controller. If the signal output goes 0v-5v over hte rotation range, it's a pot. If it is around 1v-4v ish, it's a hall.

 

[bigedgar]

Thanks for the thoughts as always, Amberwolf. I believe it's highly likely that the caps are part of the problem (if not THE problem) with my old controllers - two of my three blown controllers had the caps blow, and the third has very swollen caps (and although the third one didn't "blow", per se, the wiring connection to the battery melted). And even though I replaced the caps with higher capacity and better quality caps, now I can't control the motor - when I connect everything up, the motor goes full WOT. My attempts to troubleshoot the FETS as a possible adjacent issue didn't bear fruit. I just don't know what else to try on my existing controllers, so I'd like to cut bait and look for a better quality controller.

Do you know of any controllers for a 36v / 1000w system that you feel have quality components?

 

[amberwolf]

I don't know these days which ones are good. Kelly brushless stuff hasn't the greatest reputation around here, but have worked well for many people; their brushed stuff should be less complicated and work better.

4QD made good stuff back when I was doing brushed a decade ago. Not sure what they have now that might do what you need.

Curtis makes good stuff but it's not usually cheap, unless you find old used golfcart stuff--most of them are going to be much higher power than you need, so you'd either have to be careful with your usage, maybe keep a wattmeter on there so you dont' exceed limits, or find a way to modify it to supply less current.

I can't remember the other brands.

 

[bigedgar]

Thanks Amberwolf - I'll check those out!

 

[joe.ipsen@gmail.com]

bigedgar, did you end up getting this thing running?

I've got one in my shed with the same issue and my son has asked me to revisit my own troubleshooting of the issue. I also saw blown capacitors and was considering replacing them with a higher quality and capacity cap.

Looks like they still have controllers for sale (now at $43), but I experienced the same issue the first time I started working on this and blew the replacement controller that I purchased for ~$20.

https://www.monsterscooterparts.com/36-volt-1000-watt-controller-for-the-monster-moto-classic-1000w-mm-e1000.html

I also found that they are selling a universal controller with higher voltage and watts for cheaper ($40).

https://www.monsterscooterparts.com/48vo12unscco.html