shunt mod via a voltage divider

nebriancent

100 W
Joined
Aug 9, 2011
Messages
252
Location
surrey BC CA
just wanted to run this past you guys to see if there is anything actually wrong with this idea
the main reasons i wanted to mod the shunt via a voltage divider is so that my DPCA wont need a bulky stand alone shunt to read battery current and i thought it might be nice to have it adjustable so if i get a (oops that's running a little warm) i can dial it down easily unlike (with destructive shunt mods) soldering them up this shouldn't change the shunts value by a easily measurable amount

all voltages and shunt values are based on my controller but the resistor and pot values for voltage divider should allow any controller currently spiting out 40 amps to go to 99 amps and or multiply battery current on any controller by a factor of 2.5r
for a smaller gain of 2x battery current replace resistor 1 with 1k
and yes i know that the resistance value for resistor one as not standard .... i guess one could us another 1k pot for that aswell and dial it in for that resistance
View attachment New Bitmap Image.bmp
 
btw the high level of the shunt current sensing path on the pcb will be cut just upstream of the ca current connection and will be rerouted to the voltage divider (center of the resistors) and that should be the only mod needed on the pcb

and just for the sake of saying yes if i get a go ahead i will post pictures on this thread

and yes i know that drawing 99-100 amps through the existing shunt will force it to dissipate 22-23watts and may alter its value due to the materials temperature coefficient but for burst current it should be fine and... the resistance will climb on the shunt as it heats up causing the current to drop a bit lower then expected instead of it having a negative temperature coefficient and going into a awesome based thermal run away
 
note
if replacing r1 with a pot i would advise putting a 500ohm resistor in serries with a 500 ohm pot to replace r1 otherwise you could have a huge oops moment and have nearly no current limit inside the controller and make a neat smoke show out of your controller and or wiring /hub/battery
 
another question
in thease controlers dose the regen current get sensed through the same shunt..... if so would this not also boost regen current helping one stop a little faster
 
I have done this routinely over the years in my power electronics design during development and for prototypes. I would recommend lowering the values of your voltage divider to lower the source impedance to the circuits that follow it on the board. This will also increase the frequency response (bandwidth) of the sense circuit. Personally, if you wanted to only double the current. Use 10 ohms in series with a 10 ohm 10 turn pot. Glad you brought this up.
 
it just came to me yesterday

like a light went off in my head

use a simple circuit to fake out the current sensing in anything that is using a shunt to read current

since it is in fact a divider once set what the controller is seeing and what it is rely consuming should be very linear
but since the offset to the original resistance of the shunt is so small (i think about .2% using 10ohm resistors) it should still be fine to use the original for the cycle annalist /watts up /amp meter and not suffer a huge discrepancy in what it is metering meaning that you can have a precise control of how you modify your shunt (as opposed to my first controller i modded with the hole i think that's enough solder..... lets see) wile at the same time not need a stand alone shunt (can get a little large and wont fit in the controller) for any real time metering you are doing on the road

if my math is right i should even be able to get away with 1/8w resistors and still be miles over what they need to dissipate for heat
 
I can backup what BM has said, since I saw him mention it in another thread It just seemed the logical way to go and I havent looked back ( or had to reprogram my controllers :D ) but as BM has said use lower resistances for your PD.
 
depending on time this weekend i may get this built and installed
 
works awesome....

i used 10 ohm resistors instead of 1k as suggested so one 10 ohm and one 22turn 10 ohm trim pot
cycle annalist reads stock shunt for values and wattage per kmh seems the same
my controller must have been 40 constant 50 amp peak because after a modest 100% increase in current i read over 103 amps peak but since its only for a sec i think itl be fine

i didn't take pics off my controller during the mod but will take pics during my buddy's mod in the next copple of days (same controller same mod)

decided to add a temp sensor to the controller

ca pics

and yes i know i have a dead pixil line and its dark outside
2012-02-24 22.41.15.jpg
2012-02-24 22.41.28.jpg
2012-02-24 22.41.40.jpg
View attachment 3

oh and by the way i can confirm that at least my infineon controller dose calculate Regen current through the shunt ..
my mod has also boosted my regen current
 

Attachments

  • 2012-02-24 22.41.50.jpg
    2012-02-24 22.41.50.jpg
    75.1 KB · Views: 2,777
unfortunately i wont have those pics till next weekend as my buddy wants to see how my controller holds up till then

i may have to dial it down a tad
today i had a peak of 107a battery current it felt grate but i am slightly worried about the fets as it is only a 12 fet using 4110s
good news is all i have to do to crank back the current is turn a small screw inside the controller located right behind the switch plate

controller seems cold but i haven't taken it out too far yet
ill take it a little easy until i figure out a mounting/waterproofing solution to the temp receivers
 
I'm thinking of doing this mod so that when I do a shunt mod so it dissipates less heat I can re-calibrate everything so that when I program the controller with the software I don't need to figure in a correction factor. Right now my battery amps are 1:3.15. If I program my controller for 20A I actually get about 63 actual battery amps. Would be much simpler than trying to remember to correct it all the time. All it's going to take is one tiny slip up forgetting which controller I'm programming and next thing you know I'll have my 18 FET 4115 controller trying to pull almost 200A from the battery (hopefully the fuse will save it... it has before when I made the mistake of selecting the wrong controller model and got over 100A instead of 66A).
 
well the nice thing about this mod is it doesn't actually modify the original shunt

that is to say you will not need to remeasure the shunt for you ca reading afterward

i have played with the trim pot and fount i can change the battery current (Regen included) from stock to 2x with the turn of a screw

so for me there is no reason to reprogram the controller as all i have to do if it is running a little hot is pull over take one cover off and turn a small screw counter clockwise then i win


one 10 ohm resistor hooks up to the neg in end of the shunt then the pot on one end will hook up to the other end of the shunt
hook the free end of the resistor to the wiper of the pot and this becomes your output to the current sensing line (runs from fet side of the shunt to a small chip-set) you cut the path and patch in downstream

another thing i did wile i was in there is replace the power lines with 8 gauge wire and beef up the power traces on the pcb with solid 10 gauge wire

ill post pics at some point to show you guys where the lead is and were you patch back in
 
sorry I got busy with some prototyping for a 4 amp balanced
here are some pics as promised
high side Shunt to logic trace cut.jpg
voltage devider output to logic mcu.jpg
TRACE UPGRADE 1.jpg
trace upgrade2.jpg
 
I thought about doing this mod myself until I calculated that my 0.95 mOhm shunt would dissipate 38W at 200A :(

Need to parallel a piece of copper with it to hopefully get it down to the 0.33 mOhm range so I only have to deal with 13W.
 
Zombies,

You need to get some of those high accuracy .25mohm shunts Methods uses that aren't the wire type. Then for programming you need to name each controller and save up to date files for each controller's settings, since the programmed vs actual limits can be so variable between controllers. For program changes just retrieve the last saved setting as your starting point for changes.

What I need help with from you electronics guys is I want to take the switching for this mod up to the handlebars. After tuning using a pot I'd go with resistors to give me 3 settings (low for mountains or low traction conditions, mid for normal high power riding, and a WOW setting so high it can only be used on flat land and a few hard launches without risking my controller too much for showing off at stop lights). I don't want to bring pack voltage up to my handlebars for safety reasons, so what do you guys think is the best/easiest way to remotely switch between the different sets of resistor circuits?

John
 
A digital pot is really easy. They make a few different kinds. Mine are made by Microchip and use the SPI protocol to set the wiper range from 0-255 and come in several different values. There are some that are voltage driven I've seen before as well.

I'm actually using the SPI version in my throttle control interface and have used them before in other projects. Work great, but require knowledge of programming. You will be looking for a voltage 0-5V digital pot in a very low value.
 
I have tried that with my 18 mosfet extreme Doc edition controller.

The problem was that the wire of the pot that i installed to replace one of the resistor of teh voltage divider become a big antenna and the controller became really unconstant and had irregular glitch making it difficult to use.
I removed the pot and everyting came back to normal.

I tried to replace one of the two resistor instead of installing a pot.. and i got the same problem.

probably that i should have lowered the value of the resistor... to increase the S/N ratio...

But i think to beef up the shunt is better because just playing with the resistor and not the shunt will make the original shunt to get higher temp and change in value...
Well it become a kind of protection because when it heat up the current limit is decreasing but anyway....


Doc
 
John in CR said:
... I want to take the switching for this mod up to the handlebars. After tuning using a pot I'd go with resistors to give me 3 settings I don't want to bring pack voltage up to my handlebars for safety reasons, so what do you guys think is the best/easiest way to remotely switch between the different sets of resistor circuits?
Here I'm guessing that you don't want to get a CA involved?
 
Seth here is my pot


ill have to check out what bigmoose has to say about it

and yes doc at up to 20 w dissapation ivwould imagine it would get a bit warm but after a month of operation I found it to not be a problem as my vs is still reading within 1% and after examining the shunt found no desoldering or the like
like I said it is 100 amp peak not constant ... I decided to make this mod to address two issues
issue 1 .. stand alone shunts are large and cannot be placed inside the controller
issue 2 .. I don't have to hard mod the shunt to turn it down afterwards if I find the controller heating up too much
and a nice side bonus is as the. shunt starts to heat its resistance will climb causing the controller to limmit anyways

I have found that after one month of hard riding that the shunts resistance has not changed at all
resistance test done by reading a current through the shunt and measuring voltage drop ..psee pics for black wire on back of pcb it surges two perposes
1 it is on the high side of the shunt so annum accessorys running to this ground will be meaterd on the direct plug in ca
2 I can test the shunts resistance externally without opening the controller

my commute on this bike is 30 km round trip with a 6 km accent up a 20 degree hill .. so I do get it warm even at a modest 44 volt nominal I am still burning 3200-3500 watts up the hill and peaking over 100 amps on take off
so I would call this a sustainable mod for a controller
and one thing I would like to add is the increase in regen is nice
20 amp regen peak instead of 10th amp makes a huge differance
and if I do need to turn it down I can always iether turn down via ca or open the top and turn the pot down
 
ill get a better pic of the vd later
 
I didn't realize big moose had already covered this type of mod
... sorry if I steped on anyone's toes here it just made sence to me that this would be a logical alternitive to hard nodding a shunt that is also reversable
 
Following this with interest. I want to mod a controller to acheive less current consumption or at least make it adjustable up or down, i guess the divider network would do this?
What actually happens when a controller hits current limit? does it cut the current to a lower lever or cut it completely, i think i had this happen to me but i'm not sure if it was a sync problem or current limit, i had to close the throttle and open it again slowly. Holding the throttle open after cut out did nothing. I swapped to a higher spec controller soon afterwards and i have not experienced it again. Just wondering.....
 
Back
Top