24 Mosfet Controller XJC8b116 microchip Information Thread

[GreenKnight]

Thanks Steveo, Knuckles and Stator for the thread pointer.

Wow this really is a dedicated chip.

So much to read and even more to understand.

Beats watching the soaps.

 

[steveo]

Hey Everyone,

I just wanted help on determining which reistor i should use on the shunt to limit current. I belive methods used a .2500mohms shunt with the 18 mosfet controlles, and if i'm not mistaken i belive he was able to hit peaks over 140 amps.. I would like to peak at 100 amps max at 144v, do you guys think .5000mohms should do it? is there a way to calculate the value, the shunt on the 24 mosfet board is pertty much identical to that of the 18 mosfet controller isn't it?

Anyone know the watt size of the resistor to use also .. i'm guessing its like a 5w resistor ..

-steveo

 

[amberwolf]

You are wanting to increase the resistance of the shunt, right? Are you trying to cut the current in half? If so, just take out two of the shunt wires. If you want to cut it by 1/4, take out only one shunt wire.

It'll do the same thing as adding a resistor in series with it, cost less, and be easier to do.

It's easily possible I am misunderstanding what you want to do, so correct me if so!

 

[tostino]

Steveo, why would you want to increase the shunt resistance to lower the current limit? You can do that in software, and you have less wasted heat through the shunt.

I probably just misunderstood what you want to do though.

BTW, on my 12 fet, I lowered my shunt to 1.245mohm, which is still much higher than yours, but still allows me to hit 100+amps when accelerating with software limit at 38a I think. I can't use 100a continuous on my bike though, because of my limited voltage at "only" 51v

 

[steveo]

Steveo, why would you want to increase the shunt resistance to lower the current limit? You can do that in software, and you have less wasted heat through the shunt.

I probably just misunderstood what you want to do though.

BTW, on my 12 fet, I lowered my shunt to 1.245mohm, which is still much higher than yours, but still allows me to hit 100+amps when accelerating with software limit at 38a I think. I can't use 100a continuous on my bike though, because of my limited voltage at "only" 51v

You are wanting to increase the resistance of the shunt, right? Are you trying to cut the current in half? If so, just take out two of the shunt wires. If you want to cut it by 1/4, take out only one shunt wire.

It'll do the same thing as adding a resistor in series with it, cost less, and be easier to do.

It's easily possible I am misunderstanding what you want to do, so correct me if so!

Hey

no i'm looking to reduce the resistance of the shunt to increase the power output! Has anyone seen methods resistor on the shunt increase the current instead of the traditional; load the shunt with solder mod!

-steveo

 

[amberwolf]

no i'm looking to reduce the resistance of the shunt to increase the power output! Has anyone seen methods resistor on the shunt increase the current instead of the traditional; load the shunt with solder mod!

Any wire or resistor you put in parallel with the shunt will decrease the total shunt resistance, because you are making more a wider path for current to flow thru. How *much* wider depends on what the original shunt resistance was (which you can calculate by measuring the voltage drop across it at a known current), and what the addon resistance is.

Something I don't quite understand, though, is that if the current limit is programmed in the MCU's firmware, why not just program a different limit, instead of forcing the hardware to lie to the MCU about what's really flowing thru it?

Or is the problem actually that the shunt resistance is so high that too much power is dissipated at that point?

I have a feeling that for controllers running at these current levels, a hall-current-sensor on the B+ input line and/or each phase output wire would be a much better idea, with no wasted internal-to-the-case power dissipation from the shunt resistance. I've no direct experience with them, though, so I could be wrong.

 

[Knuckles]

Steveo,

Chances are the 4 shunts on the 24-fet controller are about 1.9 mOhm…
4 shunts = 7.5 mOHM / 4 = 1.9 mOhm.

At 100 amps … W = 100 x 100 x 0.0019 = 19 Watts (that’s a lot of heat).

For a 0.5 mOhm shunt … W = 100 x 100 x 0.0005 = 5 Watts
For a 0.25 mOhm shunt … W = 100 x 100 x 0.00025 = 2.5 Watts

The software goes up to 100 amps (probably based on the 1.9 mOhm shunt value).

So if you did replace the shunts with say the 0.25 mOhm shunt (digikey).
Then you would program in 100 x .00025 / .0019 = 13 amps in the software
To get a true limit of 100 amps.

Of course you would need to calibrate the shunt to get a truly accurate current limit.
(It's been reported that the current limit shunts can be way off ... so check them).

 

[steveo]

Hey Everyone,

I Need Everyone that has placed an order to please pm me if you would like a 100v unit of the 24 mosfet controller? or a 150v unit of the controller?

fully assembled units
the 100v units will come with 1000uf 100v capasitors & irfb 4110
the 150v units will come with 470uf 200v capasitors & irfb 4115

kits
the 100v units will come with 1000uf 100v capasitors & NO MOSFETS
the 150v units will come with 470uf 200v capasitors & NO MOSFETS

The kits will come with
-resistors needed to set the pcb for 100v use or 150v
-insulator strip that goes between the mosfets & heat sink
-All Screws to assemble everything
-on/off switch
-Pre-charge resistor (if you are looking to do 150v use)
-full casing/bus bar for the mosfets
-phase wires (may or may not include with kit, working on this)
-the pcb itself with all components installed except for mosfets
-i will try to get some pig tails for programming the pcb (just like the connections methods used!

It has come to my attention that i will not be getting the silicon wires for these kits as i had hoped for. This will help again to lower the pricing, however depending on everone that orders the kits, i may go out and buy wire in bulk for these controllers to sell with these kits!

I'm in the midest of getting prices sorted out; so everyone bare with me, i've had several pms on pricing at this time, I need alot of buyers to get the price as low as possible to benefit everyone!

-steve

 

[Knuckles]

A lifetime supply of shunt wire ...

http://cgi.ebay.com/1xM-CONSTANTAN-...al_Components_Supplies_ET?hash=item19b9589497

This shunt wire is 0.2355 ohms / m (1.6 mm dia.)

A 2-cm length of shunt wire would be 0.2355 /50 = 4.71 mOhm
Four of these shunts would be 4.71 / 4 = 1.18 mOhm

At 100 amps … W = 100 x 100 x 0.00118 = 12 Watts

The circuit board hole spacing for the shunts is about 12 mm (holes are about 1.7 mm dia.)

 

[Knuckles]

Steveo,

Got news from Keywin today about the shunts in your controller.

Each shunt wire is 4 mOhm each so the total shunt value is 1 mOhm (10 Watts at 100 amps).

Keywin says the accuracy is good to +/- 3% soooo ...

NO NEED TO MOD THE SHUNTS AT ALL.

Cheers
-K

 

[steveo]

Steveo,

Got news from Keywin today about the shunts in your controller.

Each shunt wire is 4 mOhm each so the total shunt value is 1 mOhm (10 Watts at 100 amps).

Keywin says the accuracy is good to +/- 3% soooo ...

NO NEED TO MOD THE SHUNTS AT ALL.

Cheers
-K

Knuckles You are Awsome!!!

I was meaning to ask keywin, however i kept forgetting!!!

i love how everything in this controller was built to our needs!!!

Alot of mosfets (LOTS OF POWER CAPABILITY)
Factory 100 amps shunt!!!
the option to install 100v 1000uf caps or 200v 470uf caps!!!
flipen awsome!

-steveo

 

[etard]

So what is the price on these? How many have committed so far? I am only interested in the 100 volt model, unless the price difference is negligable, in which case I am sure you will have more orderes for the more powerful option then.

 

[steveo]

So what is the price on these? How many have committed so far? I am only interested in the 100 volt model, unless the price difference is negligable, in which case I am sure you will have more orderes for the more powerful option then.

I will be posting shortly all the details

-steveo

 

[steveo]

Hey Everyone,

I've added a copy of the software to the first post!

-steveo

 

[tostino]

Any update on the price yet?

 

[andreym]

Hi! Is there any way to ship one of kits to Russia?? ))
Going to buy 408/4011 dual speed - what do you think, is it worth it?
And what is the final price??

 

[steveo]

PRICES HAVE BEEN ADDED GUYS!

go here:

http://endless-sphere.com/forums/viewtopic.php?f=9&t=15742

 

[Mike1]

Steveo,

Got news from Keywin today about the shunts in your controller.

Each shunt wire is 4 mOhm each so the total shunt value is 1 mOhm (10 Watts at 100 amps).

Keywin says the accuracy is good to +/- 3% soooo ...

NO NEED TO MOD THE SHUNTS AT ALL.

Cheers
-K

1 mOhm sounds a bit high to me, I believe the 18 fet board expected about half that value. You might need to set a lower board type to get the actual limits to match the software settings.

 

[steveo]

Steveo,

Got news from Keywin today about the shunts in your controller.

Each shunt wire is 4 mOhm each so the total shunt value is 1 mOhm (10 Watts at 100 amps).

Keywin says the accuracy is good to +/- 3% soooo ...

NO NEED TO MOD THE SHUNTS AT ALL.

Cheers
-K

1 mOhm sounds a bit high to me, I believe the 18 fet board expected about half that value. You might need to set a lower board type to get the actual limits to match the software settings.

Hey,

You have a good point, this is what i thought also, i will run it through keywin when i have the chance, but this is very simple to modify the shunt .. worst case scenario, methods had to do this with the 18 mosfet boards with a small resistor.

-steveo

 

[Mike1]

Yes, I changed my shunt for the 250u Digikey one, that gives me about twice the current limits set in the software when programmed as EB218. Gives an outrageously high current limit programmed as an EB206. My guess would be that the software will be accurate for that the 24 fet board and 1mOhm if you set it to EB212.

 

[steveo]

Yes, I changed my shunt for the 250u Digikey one, that gives me about twice the current limits set in the software when programmed as EB218. Gives an outrageously high current limit programmed as an EB206. My guess would be that the software will be accurate for that the 24 fet board and 1mOhm if you set it to EB212.

I would'nt want to try programing it with a different microchip in the software, it could risk messing up the controller!

-steveo

 

[Mike1]

Same microchip. The board setting value just tells the microchip what size shunt to expect so it can scale the signal it receives accordingly.
I ran my 218 programmed as a 212 for a few weeks while I waited for the Digikey shunt to arrive.

 

[Doctorbass]

Steveo.. I found someone that seem to follow our Infineon thread and that is making a sumarry on his webpage.. is it a E-S member?
( I was searching for the IRF4115 spec sheet on google when i discovereds that link)

here is the INfoneon 6 to 18 fets sumarry webpage:

http://www.ebike.biz/infineon.htm

Doc

 

[amberwolf]

Isn't that SolarBBQ's site?

 

[tostino]

Yes it is.