Controller shunt mods

Yeah I think I found it, but its going to the chip that controls the Motor, maybe that chip has that capabilty built in as well. If it was going to an op amp, that means the amps across the shunt are at the micro level?
 
On newer designs, the op amp is built into the main MCU chip so all you see is maybe a RC filter between the shunt and the MCU.

The shunt resistors are generally in the neighborhood of 1 milliohm. Making an adjustable divider on the signal is pretty easy but most people just lower the shunt resistance and adjust the current limit in software.
 
thanks very much, I will hack the solder bridge shunt off a damaged controller and try that, for the voltage divider, I am lost trying to find it. i see a trace coming from the shunt towards a gate driver, but seems to be going under the driver to where, difficult to tell
 
My controller has 6 mosfets, so if I increase the amp capacity of the shunt, I think i must add more mosfets. If I added 6 more mosfets, could I add them in parallel, or does this mod require a new board and controller?
 
newbiehere said:
My controller has 6 mosfets, so if I increase the amp capacity of the shunt, I think i must add more mosfets. If I added 6 more mosfets, could I add them in parallel, or does this mod require a new board and controller?
It depends on the FETs and how much current you end up with. Many times you can get away with just cooling the outside of the controller adequately (mounted in the open where there is some air flow).

Can you read the numbers off the existing FETs?

Adding more in parallel makes it hard to cool the added ones. It's usually better to stay with 6 but maybe upgrade them to better ones.
 
IRF 3205

Yes I have some heat sinks out of old PC, was thinking about putting the 6 mosfets on it, but dont know if the leads to the gates etc will be too long. maybe use the PTC or NTC sensor with the arduino or nano to sense when the temp gets too high and power the fan on the heat sink, its out of an old desktop
 
I see the drain current for the IRF 3805 is 98 amps so I think Im good as you said, no need to parallel any more mosfets just heat sink them or open air it
 
Those are rated for 55V and have an on resistance of 8 milliohms. For sure you can find FETs with much lower resistance. Less resistance will produce less heating. The legs will melt off the FETs around 70A, so the silicon rating is not the limiting factor.

I'd just try the stock FETs and see if it gets too hot. If they do, then replace them. Any kind of additional heat sinking or air flow will greatly reduce heating also.
 
for example, on a display panel,where I have my arduino nano display, could a variable pot be used instead of a solder shunt? I guess the resistor values would have to be very small for this to work
 
newbiehere said:
for example, on a display panel,where I have my arduino nano display, could a variable pot be used instead of a solder shunt? I guess the resistor values would have to be very small for this to work

Not really. You won't find a variable resistor that can handle that kind of current. You can place a small variable resistor between the shunt and the line going to the MCU. I've done this on several controllers with good results. You have to find the right trace on the board to cut into. If you post good enough pictures of your board, especially the area around the shunt, I can probably find it.
 
The picture is not the best, but the shunt is there, hard to see the traces though
 

Attachments

  • IMG_6986.JPG
    IMG_6986.JPG
    50.5 KB · Views: 3,959
fechter said:
Not really. You won't find a variable resistor that can handle that kind of current.
Sure you can. But it will probably be bigger than teh whole controller; possibly bigger than a typical DD hubmotor. ;)

It's not really what you had in mind when you said that, I'm sure, but:

Luxtrol stage light dimmer, effectively a giant wirewound potentiometer. Pic of one next to the stator of an MXUS / 9C typical hubmotor (the motor is lighter and smaller!).
file.php

https://www.endless-sphere.com/forums/viewtopic.php?f=7&t=61839&p=984742&hilit=luxtrol#p984742
 
The pictures are not so good. Where the main black battery wire attaches to the board, it goes to one side of the shunt. The other side of the shunt goes to a bunch of the FET legs. The side that goes to the FET legs will have a skinny trace coming off it somewhere that heads to one of the IC chips. It could be on either side of the board. Other than the FET legs, this sensing trace should be about the only thing connected to the FET rail. Ignore any capacitors. Look closely at the board, and if you see something like that, try to get a picture of it. It helps to have lots of light. With some cameras, it might work better to back off a little and zoom in to keep more stuff in focus.

I like the stage light dimmer. That's more like a variable transformer, but the size isn't too far off.
 
using a cont. tester, I traced on the opposite side of the board and the trace from the shunt opposite of the black wire, goes under a 2101 gate driver onto a pin of a LM 339, will post a pic later
 
I have outlined the trace in black, it seems to be on the opposite side of the board, this is coming from the underside of the shunt, opposite of the black wire side; seems to be going under the gate driver, or to a pin on it, then onwards to the comparator? then to the MCU
 
so i dremel tool off some of the trace, then put a pot in between? if its a 3 terminal pot, how do I connect the terminals?
 
so I put the trim pot across the shunt then the wiper is connected past the trace cut? as shown here?

 
Hello,

I'd like to increase a controller from 25A to ~34A.
I'll use the shunt modification, but like to be a bit precise.
The controller, currently use 2x 5 mohm shunts.

So, please tel me if I'm right :

2x5mohms= 2.5 mohms

2.5 mohms -> 25A
??? mohms -> 34A

From 25A to 34A, it's +36%
So, 2.5 mohms - 36% = 1.6 mohms

3x 5 mohms =~1.67 mohms

I finally just need to add a 3rd 5 milliohms shunt on my controller to reach my goal ?
Thank's, happy new year.
 
newbiehere said:
so i dremel tool off some of the trace, then put a pot in between? if its a 3 terminal pot, how do I connect the terminals?

Sorry for the delay, holidays are interfering with the advancement of technology again.

The way you did it will work but you if you turn the pot too far, you can totally defeat the current limiting and possibly blow something up. I've used that method before and it will be OK if you just dial it in carefully and leave it alone after that.

There are many other ways to do this. Below is just one. By using a pot and a fixed resistor, you can limit the maximum increase. In the circuit below, the pot will adjust the current limit from stock level (100%) to 2x (200%). The pot and the fixed resistor are the same value, which gives the 2:1 range. With a different fixed resistor, you could make the range whatever you want.

2x variable shunt divider.jpg

I'd like to increase a controller from 25A to ~34A.
I'll use the shunt modification, but like to be a bit precise.
The controller, currently use 2x 5 mohm shunts.

So, please tel me if I'm right :

2x5mohms= 2.5 mohms

2.5 mohms -> 25A
??? mohms -> 34A

From 25A to 34A, it's +36%
So, 2.5 mohms - 36% = 1.6 mohms

3x 5 mohms =~1.67 mohms

I finally just need to add a 3rd 5 milliohms shunt on my controller to reach my goal ?

To get 34A, the shunt would be about 1.84 milliohm. (2.5 x 25/34)

If you add one 5 milliohm to the existing, your limit would be about 37.5A (that's probably close enough)
 
yes i figured it was the holidays :D

the shunt resistor goes right across the board solder shunt? then the pot in parallel to the shunt resistor?

so the stock shunt is basically providing a reference for a comparator to limit current? How does that circuit work, if I may ask?
 
The "shunt resistor" in my diagram is the one soldered into your board. You are just dividing that voltage before it goes to the MCU.

As current passes through the shunt resistor on the board, a voltage is developed across it. Typical shunt resistors are around 1 milliohm, so you would get 1 mv for every amp. This gets amplified and sent to an A/D converter inside the MCU and software takes it from there.
 
oh ok cool got it about the shunt resistor, but I still need to break the trace for the wiper connect, correct?

Ive seen complicated work arounds for measuring current with an arduino, seems this type of shunt could do the same thing?
 
Back
Top