Controller shunt mod questions etc.

lukejbarry

10 µW
Joined
Nov 8, 2020
Messages
6
Location
Dublin, Ireland
Hi guys,

This is a sort of theory based question as I don't have any real experience in fiddling with electrics.

I have a 48v 1000w rear hub motor using an MTX33 rim with a 52v EM3ev battery with 25r Samsung cells. The controller I am using is a KT, 12-Mosfet controller with the serial number KT48ZWSRLT-SJT02L.

The controller has a rated amperage of 15A with a maximum of 30±1A. My question is, what upgrades could I make to the internals that might give me an increased level of performance/amp draw whilst maintaining safety and longevity in the controller? (for example changing to higher quality capacitors, mosfets or doing a shunt mod)

Any and all replies are appreciated, this thinking is coming from the fact that I am yet to build the bike, waiting on one part due in in the next 2-3 days. I would be really interested in logging performance before and after a possible upgrade of the controller, as I understand it is a vital element in the performance of the bike.

PS.
The reason for doing such an upgrade and not buying a new, high performance controller is mainly due to the cost.
 
First off, the obvious, don't do any modd'ing til you build and ride the bike. It may accelerate fast enough to satisfy your need for speed (That what we are talking about here, Amps. Volts is top speed).
That controller has two shunts and the common dirt floor mod is to squeeze them together and "silver" them w/ solder about 2/3's of their length. That will get 2 or 3 more Amps @ max rating, not that much.
If you really want to smoke the tires, lose the sine-wave and get an Infineon. Lyen.com can set-up some hard-hitting units and they are reasonable. You probably should get a Cycle Analyst to go w/ with any 40 A or more controller so the ramp-up can be tuned. That adds $$$ though.
The next tier would be where you get into the high-end stuff.
You don't specify, but I'm assuming the motor is direct drive, like a 9C clone. Those are better to throw Volts at, not so much Amps.
Stoplite to stoplite racers usually opt for a larger geared hubbie. like a Mac.
There's an old saying in the World of performance cars and motorcycles; "If you want to play, you gotta pay." Applies to ebikes too.
 
First, build it and ride it. See if it does what you want.

If it doesn't, check out the threads by Casainho and Stancecoke about the Open Source Firmware for the KT series controllers. That may let you modify the current limit (and other controller behavior) without doing potentially destructive (and irreversible without parts replacement) hardware modifications to the controller.

FWIW, changing FETs and caps and such out is usually more expensive than just using a better controller in the first place, because if you're not going to significantly better ones (taht cost a lot more than the "just a little better" ones), it's not usualy worth doing at all.

The two usual reasons for changing FETs is to reduce controller heating when using very high currents (for more torque), and to survive higher voltage (for more speed).

For the former, you'd need a lower RDSon (a faster switching time helps too, but this usually requires controller redesign or you may start blowing FETs from shoot-thru; there's a bunch of good threads about designing and building controllers if you want ot get into that).

For the latter, you'd need a higher VDS voltage capability; I'd recommend 25-50% higher voltage than you're going to actually charge the battery to, so you don't blow the FETs with spikes during unusual events, or with regen, etc.

Higher voltage also means replacing the caps with higher voltage ones. Replacing them with higher temperature range ones is also sometimes done, when the ocntroller is geting hot and blowing them up. Sometimes just replacing with higher quality than the common noname stuff that comes in them helps with that sort of thing anyway. Sometimes using larger capacitance (more uF per cap) helps with behavioral problems, if you have any, with higher currents, if bigger ones will fit.

Higher voltage *also* may mean replacing or modifying the LVPS (low-voltage power supply) in the controller that converts battery voltage to the 12v and 5v that everything in there but the FETs runs off of. The LVPS has a limit beyond which it will be damaged, and sometimes it fails in a way that passes battery voltage to low-voltage parts and then the controller is a brick. :( What that limit is depends on the design, but may be close to what the controller was sold as rated for (or may be much higher).

So...it's usually cheaper and easier to replace the controller with one that already does what you want. ;)
 
I would like to do this shunt mod

I don't care if my controller burns down.

What do you think.

The resistance is 0.5R.
With copper I think below that value.

Regards Bob
e296519d49739c4253ba920c854b875e.jpg


Gesendet von meinem CPH1911 mit Tapatalk

 
FWIW, if "0.5R" means half an ohm, I guarantee you that's a whole lot less than that. A regular multimeter cant' measure resistances low enough to tell even the shunt's resistance, much less that of a short piece of copper wire. ;) Most can barely measure down to 1ohm, and typically not reliably at that low range. You'd need something that is designed to measure less than a thousandth of that to begin reliably measuring a shunt's resistance, and even less than that for the wire.


However, if you don't care what happens to the stuff, then you can just put a piece of 8G or 6G wire, or a copper bar, in place of or in addition to the shunt (as soon as you put copper wire in there, it's so low a resistance you might as well take the shunt out, it doesn't really matter anymore). This will remove all current limiting from the controller, as it won't have any way to know what the current flowing is, and so it won't have any way to protect anything from overcurrent spikes.

It also cant' protect the battery against huge current spikes, so you can also risk damaging or destroying that, even causing a fire if it's bad enough, since you also want to disable the BMS's ability to protect against overcurrent, according to another post of yours in a different thread.
 
What a great question with answers, including '" build 'em & break ' em '" , in this short discussion!
 
amberwolf said:
First, build it and ride it. See if it does what you want.

If it doesn't, check out the threads by Casainho and Stancecoke about the Open Source Firmware for the KT series controllers. That may let you modify the current limit (and other controller behavior) without doing potentially destructive (and irreversible without parts replacement) hardware modifications to the controller.

FWIW, changing FETs and caps and such out is usually more expensive than just using a better controller in the first place, because if you're not going to significantly better ones (taht cost a lot more than the "just a little better" ones), it's not usualy worth doing at all.

The two usual reasons for changing FETs is to reduce controller heating when using very high currents (for more torque), and to survive higher voltage (for more speed).

For the former, you'd need a lower RDSon (a faster switching time helps too, but this usually requires controller redesign or you may start blowing FETs from shoot-thru; there's a bunch of good threads about designing and building controllers if you want ot get into that).

For the latter, you'd need a higher VDS voltage capability; I'd recommend 25-50% higher voltage than you're going to actually charge the battery to, so you don't blow the FETs with spikes during unusual events, or with regen, etc.

Higher voltage also means replacing the caps with higher voltage ones. Replacing them with higher temperature range ones is also sometimes done, when the ocntroller is geting hot and blowing them up. Sometimes just replacing with higher quality than the common noname stuff that comes in them helps with that sort of thing anyway. Sometimes using larger capacitance (more uF per cap) helps with behavioral problems, if you have any, with higher currents, if bigger ones will fit.

Higher voltage *also* may mean replacing or modifying the LVPS (low-voltage power supply) in the controller that converts battery voltage to the 12v and 5v that everything in there but the FETs runs off of. The LVPS has a limit beyond which it will be damaged, and sometimes it fails in a way that passes battery voltage to low-voltage parts and then the controller is a brick. :( What that limit is depends on the design, but may be close to what the controller was sold as rated for (or may be much higher).

So...it's usually cheaper and easier to replace the controller with one that already does what you want. ;)

I've got another question, also thanks for your great reply as well as everyone else that has contributed it has been a great help!

What Infineon controller on Lyen.com or elsewhere would suit the 1000w 48v direct drive and the 52v 12.3ah (25r) battery?

Thanks!

What
 
Why don't you contact Lyen directly.
It's been at least 5 years since he set-up a 22Amp 6-FET Mini Monster for me and he was very helpful. Real personlized service. I haven't seen any of his products on here for a while, but since his site is still up, I'd assume he is still in business.
He set-up it w/ a LVC customized to the LiPoly Chem pack I use and I remember the Mini Monster packed a punch for a cig. pack-sized controller. I felt about the same as a Grineon 25 Amp unit. Both would spin the tire on my frt. mounted MXUS mini being run on 14S LiPoly (Around 52 Volts).
Tell him what you (will) have and what you want to do. Manbe the 9-FET will suffice and it would be a lot smaller than the KT 12-FET. And it goes up to 72 Volts, which is what those who want to have fun with the 9C clones like to do.
And take a look at his add-ons to see if any of those interests you.
 
My Shunt is deactivated,
I will hard fuse my current.

But my controller is becoming an active cooling.

I have drilling a 3.9mm pothole in the Mosfet cooling body and added a temperature sensor. The sensor is controlled by the thermostat control unit. This unit runs on 12V and will activate the ventilation.

I can program the value for temperature.

Greetings Bob



Gesendet von meinem CPH1911 mit Tapatalk


View attachment 1
 
Back
Top