Infineon XC846 72V 45 Amp Controller

[fechter]

Seeing the discussion on the shenzhen thread... Here in Italy they can get very anal with law sometimes (a guy got a ~3000$ fine for a not-standard ebike ) it would be nice to implement a quick modification so you can limit the power with a switch.. Here law puts a 15mph limit and 250w max for the motor.. and you must have a PAS so the motor doesn't run if you don't pedal.

I think something like that could be done. The PAS part would be more difficult and you'd need some kind of sensor. You could place a resistor in series with the throttle to seriously reduce the maximum. The resistor could be switched in and out. The switch could be disguised, hidden or magnetic.

[gip_mad]

I was actually thinking about decreasing the current limit, so the bike won't start by herself... maybe I can hack the code to use the 60/180° input for the PAS sensor (that I have already installed, but is not connected atm). I still have to think about the switch, maybe a piece of paper that when removed makes the contact, so the police won't notice a piece of paper on the floor...

[fechter]

Lowering the current limit would be fine. You could change the gain of the current sensing amplifier.

If you really want PAS, you could use a sensorless setup like a few guys are working on.

You could possibly hide the switch in plain sight. Label it "odometer reset" or something.
A reed switch like the ones used on speedometers along with a small magnet make for a nice hidden switch. Just move the magnet away to turn it off.

[geoff57]

hi all
I think I have the wiring of the CA worked out, I cross refrencenced with how fechter wired the first china controller also using a CA DP x'tlyte controller I have, with the pictures on justins web site.

this is for use with a puma geared hub,the pinout works for all geared hubs.
So the pinout
1: Vbatt use a polyswitch and attach to VK2+ or Vcc-L near the orange wire
2: Digital Gnd thick black
3: Analog Gnd thick black
4: Isense shunt controller side NOT thick blackwire side
5: Hall external cycle computer reed switch then to a GND pad
6: Ebrake EBS-
see below

shunt closeup.jpg (95.98 KiB) Viewed 2618 times

the picture below should help with pin 5 and pin 6

brake closeup.jpg (129.55 KiB) Viewed 2625 times

pin 6 goes to EBS- and the hall sensor pin 5 goes to a reed switch then into the controller and is soldered to a GND pad like the one I have shown.


pins 1,2,3,5 I am sure of,pin 4,6 I am 95% sure of
for pin 4 I used information from how fechter attached a CA to the first china controller see post below
http://endless-sphere.com/forums/viewtopic.php?f=2&t=4282&st=0&sk=t&sd=a#p63528
I have put the Isense wire in the same sort of place after following the components to make sure the place matched.
for pin 6 the ebrake I have got it correct if the CA shorts to GND when the Ebrake is applyed, from the 20A controller rewire picture I have got this correct.
hope this helps you guys out there Fechter will you check my pin 4 location is correct.

[keywin]

Hi all,
the picture is hall sensor connection.

[geoff57]

Hi all,
the picture is hall sensor connection.

Great you fixed the problem of no hall pickup point to the DP CA for the infineon not sure if I can use that with a geared hub but others can.
In your picture there is a 10 k resistor my controller did not come with one what wattage is it so I can fit my own.
The controller is a 72v 45a model

[geoff57]

Hi all
I have just taken the side of my Puma and checked on the following number of poles answer 32. gear ratio 5:1.

If I wanted to use the hall sensor to mesure speed with a DP CA I would have to set the number of poles in the CA to 5 X 32=160 poles per turn of the wheel the max setting on the CA is 99 so this would not work .

while writing this post I found a way to get round this problem speed is mesured in a cycle computer by multiplying the circumfrence by the rpm of the wheel, well if we set the CA pole number to 80 half of the actual number and leave the circumfrence normal then the CA will think you are going twice as fast as you are, so half the circumfrence and the speed is reading correctly again although the numbers are not correct half the circumfrence and two and a half times the number of poles.
the results will be correct and that is what we all want.
I hope this helps .

[Knuckles]

HA HA! YOU SO CRAZY!

Good Hack!

Fried Hampster award!

[geoff57]

hi all
going tho use to se if I can fool a cycle computer into giving rpm readings not speed I will use the idea in the CA web page on how to show rpm instead of speed. many thanks to justin and the team for that.
the settings are as follows set the wheel circumfrence to 1666mm and use kmh, whatever the computer reads multiply by 10 to get the rpm, so if the lcd reads 34.6 then the rpm =346rpm.
Its not my knowledge or maths though I could of worked if I had hat to but justin already has, a local superstore does offers evry so often cycle bits including cycle computers for £5 as a result I have 5 in stock so setting up 2 or 3 for tests is no worrys.

[geoff57]

any word on the throttle situation hall resistive or both

[abhimanyu]

i found a circuit in bldc motor control pcb...for overcurrent detection...im unable to figure out what exactly is happining in it...so please give your valuable suggestions...the microcontroller used is XC846..( if any one could post any document regarding the pin layout and the modules of this microcontroller that will most helpful)...below is the circuit


im using cypress microcontroller for bldc motor control in my ebike model...

everything else is sorted out except the overcurrent detection part...for which i traced the circuit which i have shown above from a pcb of some japanese ebike...it has infineon XC846..for which im unable to find any datasheet or pin config details(if u have any material related to them please post)..and i dont have the code...im trying to figureout the logic and hope to implement in my board...

any suggestions are most welcome...

[gip_mad]

The opamp on the top is a comparator, and switches an "alarm" on overcurrent. The opamp on the bottom is to amplify and measure the current. Datasheets for InfineOn motor control ICs are on the infineon website (google for it, or you can find the links some post above)

[abhimanyu]

infineon website didnt have even a single document on XC846. its not available still in the thread...

can u extend the point you mentioned...i mean regarding how the XC846 ...handles the alarm and the amplified voltage...

thanks

abhimanyu

[ZapPat]

XC866 - Same core, also has same motor control PWM module

As for gip_mad's explaination of the two op-amps, it just means the first op amp filters and amplifies the analog signal from the current sense resistor, and the second op-amp works in comparator mode to produce a digital-like (On/Off) signal from the first op-amp's output. This digital signal is probably fed directly to one of the micro's inputs that is interrupt-ready for rapid response to the overcurrent event.

[abhimanyu]

""the second op-amp works in comparator mode to produce a digital-like (On/Off) signal from the first op-amp's output."//

neither of the outputs are goin to the other opamps..?

can u explain that point

[gip_mad]

Since you're trying to make your own controller, I strongly suggest a good course on OpAmps basics first!
ZapPat was probably meaning that the 2 op amps share the same input signal, they take the voltage from the shunt resistor and one outputs a logic high when voltage goes above 0,65V. The other simply put out a voltage proportional to the current measured by the shunt, multiplied by a factor thet you can calculate with standard formulas (google "opamp basics" please), so it can be easily read by the ADC of the controller.

Wait, did you just post the same questions on the Shenzhen controller topic?

[abhimanyu]

srry...im new here..from now on ill discuss in the "shenzhen controller" topic....i wont post here further...and regarding the opamp basics...i know before posting here itself that they are comparator and amplifier...i was expecting suggestions regarding why both are used...and how the MCU is handling them....
thnks to fetcher ...i got an idea over those points...this will be my last post regarding this' in this topic

[ZapPat]

Sorry abhimanyu, I didn't look very well at the circuit, I just assumed one op-amp was a low level amplifier and the other a comparator using the first one's output. Others have answered correctly, sorry for jumping to conclusions!

I wonder if the new infinion also uses dual current signals like this (both digital and analog)? I don't have mine yet to check it out...

And wouldn't some amount of RC filtering using small caps be necessary in this type of setup? I would think that PWM switching noise might be a problem without them, no?

[gip_mad]

And wouldn't some amount of RC filtering using small caps be necessary in this type of setup? I would think that PWM switching noise might be a problem without them, no?

I'm pretty confident that there is an RC filter, maybe he just left it out of the schematics..

[mikereidis]

Can I buy one of these controllers yet ?

My XLyte 72v, 20a controller is working fine at 84v, but I'm getting tired of pedal first on this heavy recumbent.

[pwbset]

My XLyte 72v, 20a controller is working fine at 84v, but I'm getting tired of pedal first on this heavy recumbent.

If you decide to sell that controller for a nice low used cost PM me. Pedal first is the only way to go for me. Looking to get another one.

EDIT: Oops.. that was supposed to be a private message for Mr. Reid. Sorry.

[ZapPat]

Hi everyone,

I just got an infinion controller, but am having a hard time getting it to work well.

I first had to use the provided variable resistor to adjust the Low Voltage Cutoff down because I have a 48V Ping battery, not a 72V one. I had to turn it clockwise, but I have no idea what the LVC is now, exept that it is lower than before, and now works at least at 55V.

After doing this the motor at least reacted when I press the throttle - but not well at all. So that means finding the right hall wire/phase wire combo. Despite my feeling half way through trying all 36 hall/phase wire combinations that something else was probably wrong, I did the whole lot just to be sure anyways.

Now the situation is that the motor turns using the best combination of hall/phase wires, but it also consumes about 10 times more current than this same motor runing with its previous controller - NOT normal!

So then I started to think that maybe it is the motor's hall sensors that are not configured in the same way as the controller expects them to be, so I checked my old controller and found that it was set for 120o hall sensors. Now the old ecrazyman had a wire to change between 60o and 120o, but I have none like this on the new controller. Which leads to my questions:

What is the new infinion controller's hall sensor inputs set to by default - 60o or 120o?
What pad on the PCB might be this 60o/120o hall selection? Maybe it has the same markings as on the old ecarzyman - anyone know what that is?

I can't wait to get this thing on the road, so I appreciate anyone's help with this!

PS: My motor is an oldish golden motors 500W hub (2~3 years old) in case this might help...

[John in CR]

I asked about 60° and 120° phasing and was told the controller is supposed to sense which the motor is and send the correct signal accordingly.

John

[Microbatman]

Has anyone used this controller on a Crystalyte X5?

What is the wire configuration for an X5?

[ZapPat]

I asked about 60° and 120° phasing and was told the controller is supposed to sense which the motor is and send the correct signal accordingly.

John

Thanks John, I just found this out yesterday from Knuckles too, so that rules out this possibility.

The motor is still runing like a tank at it's best hall/phase combo, taking much more current than it should. As soon as I drop the throttle it just coasts silently to a stop.

I'm thinking now that maybe my recent motor repair job didn't fix the motor 100%, although it seems to look OK. It outputs nice sine-waves on the scope when turned, and all phases look like they are outputing pretty much the same amplitude signal too. Wouldn't this confirm that there is no shorts in the motor windings if the BEMF's are all fairly equal? Also, the resistance checks out OK, at least as much as I can check it with a regular multi-meter.

Hall outputs look OK too.

I also tried running at both 50V and 72V, still no difference.

Has anyone ever experienced a "grumbling" behavior with a motor like this? Any ideas?