Modify a Magic Controller Ver. 1

[Dave Watson]

I installed a new controller (BAC-0282P) and 1000W 48V motor (MW16B)recently into my chinese e-bike (stock G$O bike with a supposed 350/500 W motor).
The motor ran well but fell short of my expectations of at least 1/3 more power.  Of course I suspected the controller.
So now I am posting pics of the controller to get advice for MODDING. 

So here it is

[Leslie]

Your shunt.

Nice pic of both sides.  Makes life easier.



Right its hard for me to tell but how thick is that shunt wire you have there?

[Leslie]

Looks good for 20 amps! Maybe we try for 25.

I may suggest first just checking first that there is plenty of solder on the B- and B+ connection points and both shunt solder points and beef them up with a good think blob each, the B- and B+ connections look feeble

And beefing up those blue green yellow connectors would help a little too.

After you have done this.

If you're still not happy read on.

There are a few ways to go about this.

Shorten the shunt.
By pass the shunt.
solder the shunt.



If you want more amps then just add a little solder to one side of the shunt from its anchor hole along the length of the shunt, in essence shortening its resistivity. 

Try only going 4 mm from the anchor hole then leave the rest of it as it was.  Shortening can make more heat in the shunt due to its gauge thickness. 

If you want more which I think you be pretty happy with just beefing the connectors up with solder you will need to get another shunt, find something that can serve as a shunt or I may be able to send you one from my old controller if its a last resort.

If you have any old broken controllers around the maybe you can use on of those.

The shunt you have there looks short and stumpy,  Good for about 20 cont 25 amps max.  So I do suspect its you solder points.

[Leslie]

There looks like a bad solder point where I have labeld Whats this.  Plenty of beef on the flip side here but a little feeble where I labled more beef here.

If this is what I think it is, (the bad solder point), there! is where your magic controller power going down the drain.



 

[Leslie]

And I found another point that concerns me.

One fet looks like its not got enough to its drain leg.



A Mosfet.  (FET)^^

Check to see the solder is all the way up to all the 2nd and 3rd legs of the fets.  The first leg is the gate you don't need beef on this connection.






IMO don't modify the shunt until you beef up the PCB tracks and try the bike.  We may not need to modify this controller just solder up a few spots.

No need to go over board just make the power circuit robust.

Good luck  

[Leslie]

And just watch you work

Look at you pictures and recheck the board and make sure you havent messed up any tracks with short circuits.  It can be easy to do when you apply a lot of solder around the board.

[Dave Watson]

Right then,  I'll post a close up of the problem areas before and after.   Will have to practice a little on some junk boards (I'll have to brush up with floating large amounts of solder to a board).  The rest should be a piece of cake and I'll be sure to get data results (thermal and current).

[Leslie]

Good to see you are cautious and preparing.  It is the easiest mod and the easiest mod to mess up.  Because it is so easy to solder a bit here and there people often go to much and they have a controller just waiting to fail.

Try to aim for the ratings for your motor.

The MP is approx rated at 1350 watts 26 amps max.

And 1000 watts 19 amps constant.  20 amps is fine.

This should give you a pretty wide range of performance with some good head room for overheating heat.  Unless you live in the Sahara desert and or weigh 120 kgs or carry passengers you will enjoy a long life of your motor.




Before you post the pics could you let us know if those areas I pointed out are of concern?

Its pretty hard to tell for certain from pictures as they don't show the 3d nature of solder blobs.

Thanks.

[o00scorpion00o]

I noticed a few dodgy spots of solder on my controllers too, I just re soldered them, while I was beefing up the tracks.

It will be interesting Dave, to see how your modifications go as mine all went up in smoke! LOL

I had 1 set to 35 amps, It lasted a few weeks. I think the heat sink is absolutely not suitable for more than 20 amps. When I was pulling 35 amps from it it really roasted, I mean it was like touching the oven door! It could have been one of the reasons it died, However when it died I was not putting it under any stress at all and it was running cool, maybe the earlier heating up was the initial cause. I was going slow and the controller was not getting a lot of air. But I seriously think these controllers should be installed in another case with a good large heatsink. Maybe one of the reasons the more expensive controllers have 12-18 fet's ?

One other thing. If you modify the shunt, You will need to turn the amps down via the software cable. I had it set to around 8 amps continuous and 10 max to give me the 35 amps! Also do you have any device for monitoring your current consumption while riding your bike so you can monitor it?

[Leslie]

I noticed a few dodgy spots of solder on my controllers too, I just re soldered them, while I was beefing up the tracks.

It will be interesting Dave, to see how your modifications go as mine all went up in smoke! LOL

It it really roasted, I mean it was like touching the oven door! It could have been one of the reasons it died,

 

LOL you crack me up sometimes.

The reason why it broke I think you heat up the fet so hot the silicon can fracture or the p region and n regions or the depletion region between them crack or something. when cools or cools too fast.  So the next time you ride it may fail.  It might take a few hits on the throttle and bam the controller dies.



One reason you don't hose your car motor down when its overheated.

Id love to go through a few dozen fets and examine them with some high magnification one day.

Really who knows?  The above is just an assumption but it makes sense..

[o00scorpion00o]

I noticed a few dodgy spots of solder on my controllers too, I just re soldered them, while I was beefing up the tracks.

It will be interesting Dave, to see how your modifications go as mine all went up in smoke! LOL

It it really roasted, I mean it was like touching the oven door! It could have been one of the reasons it died,

 

LOL you crack me up sometimes.

The reason why it broke I think you heat up the fet so hot the silicon can fracture or the p region and n regions or the depletion region between them crack or something. when cools or cools too fast.  So the next time you ride it may fail.  It might take a few hits on the throttle and bam the controller dies.



One reason you don't hose your car motor down when its overheated.

Id love to go through a few dozen fets and examine them with some high magnification one day.

Really who knows?  The above is just an assumption but it makes sense..

Yes total sense Leslie, and one of the reasons I suggested any modding should NOT be done using the existing controllers casing. It's extremely unsuitable for more than 20-25 amps in my opinion.

Remember I thought about installing a large heatsink on my battery box? I was going to remove the circuit board and bolt the fet's to the heatsing on my battery box! I didn't do it because Only riding in the woods and pulling 35 amps for long periods was toasting it, the controllers surface area was not large enough to catch air while going slow. A large heatsink with large finns would have worked well. I didn't get the chance in the end. Controller R.I.P 

[Leslie]

Oooh I have to check that one out.

Some fets to-220 package metal is the same as the drain leg meaning you cant hook them up to the same heat sink unless you use thermal transfer insulator and plastic isolation washers. That light blue stuff you see under the fets on the old controllers.


Yes I thought about this.  I think the fets expand under high heat and everything bend about more easily.

When they cool the steel and silicon warp differently as the steel cools faster than inside medium causing any bondings to fracture.

You can even get intermittent behavior just before they crap out.  Fets get warm they may work, like the xbox360 and towel trick a different thing happening but explains what I am trying to translate here..  there are a few things that can happen.

But with the high power 3 phase controllers once failure has occurred I think the other fets lance the weaker fet in the heart and it all over.

[Dave Watson]

Hell again, upon closer inspection prior to shunt and solder beef ups, I have noticed the following and took pics of the areas LESLIE pointed out earlier.

So checking out the FET leg, it seems well soldered on the opposite side to a bus bar?


There appears to be missing solder on one leg to some type of transformer?


The area LESLIE labeled WHAT'S THIS is the soldered side of the SHUNT shown in between the capacitors near the negative battery lead.  Looks like a weak solder.


The controller has since malfunctioned and gives 5 beeps on startup, the wheel cogs when spun by hand.  I was moving the bike when it started to beep suddenly and this is when the controller stopped responding.  The usb interface still functions so I am going to do the above repairs first and report from there.

[Dave Watson]

Solder beefed up and connections fixed.

[Leslie]

Sorry for my slow response,


Yea that thing has been pointed out before on the switch mode that's normal.


Have you tried it yet I think you soldered that point where I said its normal.  I have no idea what that's is for.

I would put the controller the way it was and repair the shunt and the joint I posted.