Controller Upgrade KLS6030H to KLS7250N with Field Weakening

[EpsilonZero]

A few weeks back, a new member (soilsamplerbill) was messaging me regarding motor controllers for his build. When he was talking to Fany at Kelly, he was asked if he wanted field weakening on his KLS-H controller. Now, it turned out the KLS-H line doesn't support it but the KLS-N line secretly does. If you ask for it in the notes, you can get a controller that, along with special software, supports field weakening. So, I decided to order a new controller because I already bought two to three versions of everything else on the kart, so why not upgrade the controller too?

First, disassembling...







I ordered it on a backplate, thinking I might be able to just affix the pre-assembled unit and save some time... Nope! The orientation was not good and components too spread out, making the whole thing too big.



Oh well... I don't have a bunch of tools for nothing. I mocked up a cardboard sub-assembly for the controller, battery switch, and charging port (you can see it under the white dustpan, with the started steel sub-assembly lying on the gloves. I left a huge open area to provide airflow to the back and sides of the controller.



Lots of cutting, welding, and drilling later (before I realized I forgot the charging port tab)... I like welding in direct sunlight or no sunlight with an LED light on top of my work. The latter maybe more because my eyes don't need to adjust with my cheapo mask.



I made mounting points from angle iron and bolted them to the sub-assembly. I used my extra hands (a thrashed tire set and a c-clamp) to hold the assembly in place while I welded the mounting points to the kart frame.


After I painted the sub-assembly, I realized I forgot the charging port and promptly cut the wrong sized hole while making the tab. (correct size hole not pictured...)



Here is the back of my completed sub-assembly.



And the front... Note the giant aluminum plate it came on and the old controller assembly.



<continued...>

 

[EpsilonZero]

Sub-assembly attached... I only used two mounting points, but it is extremely solid.




Battery re-installed.


I am so far very impressed with the performance. The KLS6030H couldn't use above 60v and would turn the motor at 5000 rpm no matter how far above 60v I fed it. The new controller doesn't have that issue plus it has field weakening, which can spin the motor to the full rated 5500rpm and take it above 55mph. Surprisingly, the kart feels stronger now geared with 12t/44t that the previous configuration with 12t/58t. I will be doing some more testing and be making videos soon.

 

[Functional Artist]

Nice work!
How does the air (flow) get to/thru the cut-out, that you put in the back plate, with the seat right there?
Have you noticed much heating of the SC?

 

[EpsilonZero]

Nice work!
How does the air (flow) get to/thru the cut-out, that you put in the back plate, with the seat right there?
Have you noticed much heating of the SC?

I noticed a while back that behind the seat, the air flows strongly in reverse (sort of like you see in pickup trucks). I put a little flag on the seat and it actually flies hard backwards as I drive. So, I spaced out the sub-assembly about 1 1/2" with the angle iron and the natural air flow does the rest. If I added a fan, I think it would even slow down the airflow at speed. There is no noticeable heat coming from the controller even when running it hard, which is a big difference from the KLS6030H configuration.

 

[EpsilonZero]

Here's my first test video with the new setup. The only issue I am having is that when I cut throttle in the field weakening range, the motor suddenly becomes a generator and a strong regenerative current ensues, resulting in 100% regenerative braking until below FW range. It is strong enough that teeth skip. I'm not sure if there is a way around this issue, so I may turn off the field weakening or put it on the low/high speed switch based on speed.

 

[EpsilonZero]

Regarding the unwanted FW braking, Fany from Kelly ssuggested setting "RLS_TPS Brk %" to "1" and test. After that change, I didn't get such violent braking (no more teeth skip), but the tires still skid when letting off throttle at top speed. I also noticed getting no braking on the first few runs even before that change, so it could be that as voltage drops from max the problem presents itself. More testing will follow.

Another interesting thing I found was that field weakening top speed was drastically reduced at 20kHz PWM (~52mph). 14kHz and 16kHz were able to do about 58mph.

 

[soilsamplerbill]

I also noticed getting no braking on the first few runs even before that change, so it could be that as voltage drops from max the problem presents itself. More testing will follow.

I had an epiphany. If the battery has high enough voltage, the controller will not allow it to be charged further, therefore no regen??? When regen does activate in field weakening state is it less violent when you ease off the throttle slowly?

 

[EpsilonZero]

I had an epiphany. If the battery has high enough voltage, the controller will not allow it to be charged further, therefore no regen??? When regen does activate in field weakening state is it less violent when you ease off the throttle slowly?

I haven't tried setting the over-voltage lower. I don't know how the controller will handle the error exactly, so I would need to test it out. If it simply reduced or eliminated regen momentarily that would be okay, but I wouldn't like the controller needing to be cycled if that's what it does. Easing off doesn't seem to make a real difference nor does "Accel Rls Time" ("time of TPS Torque from max to 0").

How is your kart coming along?

 

[soilsamplerbill]

Trying to work on it with my grandson so that slows things down some but he's loving it. I found out the kart was supposed to include an engine mount, so I was able to get that to make mounting the motor easier. Also had my nephew cut a motor mount on his plasma. It has holes that match up with the engine mount. Tomorrow, he going to have it powder coated for me. I'm going to start a build thread with photos soon. I'm actually still waiting for the axle sprocket to show up.

On another note, did your regen always work for you before employing field weakening?

 

[EpsilonZero]

Cool! I'm looking forward to the build thread. Yes, regen has worked fine on both controllers without and below field weakened speeds. I anticipated this could be an issue because I have heard of this sort of problem before (not necessarily Kelly specific). It could be that all the kinks aren't worked out yet with Kelly's deployment of field weakening, so they aren't advertising the feature. Regardless, seat of the pants tells me this controller is getting more from the motor and it doesn't get as hot and reduce power. I was actually running the phase current a little lower to drop the temperature on the old controller. Plus, it gave me an excuse to design and build a whole new mounting assembly that blows the aluminum mono-plate, hose clamps, and zip ties quick setup out of the water.

 

[soilsamplerbill]

That's a very nice job. Looks much cleaner.

I was at a friend's house Friday night. He worked for a electric lift truck manufacturer and was telling me about how they are set up. He said that field weakening was an popular option that could be ordered. If he was still there I probably could have had him get my parts.

 

[EpsilonZero]

That's a very nice job. Looks much cleaner.

Thanks. It took way longer than I originally anticipated. If I was going to do this any more often, I'd probably need to get a plasma cutting table like your nephew.

Fany at Kelly says the controller will shut off and need to be cycled if the overvolt protection is tripped. So, I won't be using that to counteract the field weakening induced regenerative braking.

 

[soilsamplerbill]

Fany at Kelly says the controller will shut off and need to be cycled if the overvolt protection is tripped. So, I won't be using that to counteract the field weakening induced regenerative braking

Well, I wasn't really suggesting that. I was referring to your lack of braking at the first of the run. If, let's say your battery is fully charged and you start off on a long steep downhill run, would the regenerative braking work? All that energy has to go somewhere and if the battery is fully charged I don't think the controller will allow regenerative drag to be turned on. Therefore no brakes (regen).

 

[EpsilonZero]

Well, I wasn't really suggesting that. I was referring to your lack of braking at the first of the run. If, let's say your battery is fully charged and you start off on a long steep downhill run, would the regenerative braking work? All that energy has to go somewhere and if the battery is fully charged I don't think the controller will allow regenerative drag to be turned on. Therefore no brakes (regen).

Regenerative braking is only like 15% efficient. Unless I roll down from the top of a hill without using the throttle (can't happen here), I won't be putting in more energy than I took out.

 

[soilsamplerbill]

So you don't think that could cause what you experienced ( lack of regenerative braking). I guess I was thinking that you charge your battery to a higher voltage than what the controller was set to limit regen drag charging. I'm like you though, rather give up a little speed and have it all functioning properly. Kind of neat to follow your experiments with it. I won't get to do much experimenting, but my oldest son and one of my employees are small enough to drive it and give me feed back. I hope.

 

[EpsilonZero]

When I said "getting no braking", I meant not getting unwanted braking without hitting the brake.

 

[soilsamplerbill]

Communication, lol