SOLVED! Infineon Cutting out Question

matt_in_mtl

100 mW
Joined
Aug 13, 2008
Messages
49
Hi,
I have a Turnigy 63-74 that I added hall sensors to and I have hooked it up to a 6-fet infineon controller. The issue that I am having is that I seem to have found a phase/hall wiring combo that works. I can roll the throttle on gently and the motor starts to run with a low no-load current. Unfortunately if I ramp up the throttle, the controller stops driving the motor. I then have to close the throttle before it will start again. It is as if I am tripping LVC or some fault. I am running it of 12S1P 5000mAh lipo, and it is the 48V version of the controller.

Any suggestions on where I should look? Thanks in advance.

-Matt

*Edit*
Spoiler Alert:
The 6-fet, 12fet, and 18fet 116 controllers have 2 types of current limit. 1 is a software limit which reads the voltage across the shunt, and uses it to control the power to the motor. There is a second hard cutoff which uses the voltage across the shunt to switch a transistor which causes the controller to fault. For RC type motors, it is very easy to trip this second shunt from very short high current pulses.

There are 3 ways to Change this hard cutoff:
(from ZapPat on page 5, red are additions by me)

- So I guess the easiest fix for the cutout is the programming technique:
Selecting "EB212" instead of "EB206", and then doubling both the current limits (as compared to the usually used limits for the "EB206" you need to double your desired currents because the 12-fet has 2 parallel shunts, where the 6-fet only has 1, so the 12-fet will give twice the current for a given setting).
Or again... Selecting "EB218" instead of "EB206", and then tripling both the programmed current limits with the "EB218" compared to the usually used limits for the "EB206".

- There is the method of adding a resistor across surface mount capacitor "C20" which acts as a voltage divider for the overcurrent signal. A 1kohm resistor here should double your overcurrent cutoff, a 500ohm one tripple it, and a 333ohm quadruple it.

- There is also the good old shunt soldering method that would also help reduce the cutoffs of course.

- Adding some extra capacitance in parallel with "C20" may also help reduce sporadic overcurrent cutoffs by filtering out the spikes.
C20 is unmarked, so we don't know what value it is. My solution was to increase R43, it is 1k standard, I have increased it to 47K, which has effectively filtered out all the current spikes, but if the current is high long enough it will still trip the cutoff.
 
Check the solder joints on the LM317 voltage regulator. A friend had exactly the same problem recently and it was a dry joint on the voltage regulator.
 
matt_in_mtl said:
Hi,
I have a Turnigy 63-74 that I added hall sensors to and I have hooked it up to a 6-fet infineon controller. The issue that I am having is that I seem to have found a phase/hall wiring combo that works. I can roll the throttle on gently and the motor starts to run with a low no-load current. Unfortunately if I ramp up the throttle, the controller stops driving the motor. I then have to close the throttle before it will start again. It is as if I am tripping LVC or some fault. I am running it of 12S1P 5000mAh lipo, and it is the 48V version of the controller.

Any suggestions on where I should look? Thanks in advance.

-Matt

This sounds like the very same problem I was having, It was resolved by a slightly different spacing on the hall sensors. What is your current spacing for your halls ?. It only happend on my 2 newer 6fet controllers.
 
Have you programmed your LVC on the Infineon or just assumed?

12S1P @ 5AH = 44.4v nominal

Normally the 48v controller would have a LVC of 41.5v (the 36v have 31.5) which is a bit too high since your lVC on a 12s pack should be around 39.5v... still what current are you drawing from the pack when the controller tripps out?

Have you done any reprogramming of the controller?

-Mike
 
gwhy! said:
This sounds like the very same problem I was having, It was resolved by a slightly different spacing on the hall sensors. What is your current spacing for your halls ?. It only happend on my 2 newer 6fet controllers.
gwhy! I was wondering if this was the same as you were experiencing. I temporarily fitted the halls inside the motor, in the gaps where the wire is wound, spaced equally (120deg mechanical). It should provide corect timing I thought? I seem to remember something about diffferent hall placement for delta vs. star. I know other sensored BLDC outrunners that had the halls placed I did were star winding and the Turnigy is Delta... maybe that is causing a problem? I am planning on doing external mounting, but wanted to get up and running for proof of concept before spending the time. If this is the same as what you saw, then I should probably make my external hall bracket now and stop wasting time.

mwkeefer said:
Have you programmed your LVC on the Infineon or just assumed?

12S1P @ 5AH = 44.4v nominal

Normally the 48v controller would have a LVC of 41.5v (the 36v have 31.5) which is a bit too high since your lVC on a 12s pack should be around 39.5v... still what current are you drawing from the pack when the controller tripps out?

Have you done any reprogramming of the controller?

-Mike

Mike,
the controller says LVC is 41.5V. I have not successfully programmed the controller (can't get my controller to work properly, cant program my controller... I'm starting to look like a real noob :oops: ) I thought I would try it as is now that I have wiring for series connecting my 2 6s1p lip packs, before screwing around with the controller. I can tell you that according to my cell monitor and turnigy watt meter, the pack voltage didn't drop below 48V.

Have you guys connected LEDs to your controllers? I connected one to the pwr/diagnostic 1 output, and it lights when the controller is on, but has never blinked.

Thanks again for the suggestions, you guys rock.

-Matt
 
The different hall wiring for delta vs/star would be just flipping the middle hall upside down or adding an upside down hall (I forget why) but it doesn't seem needed with the infineon on hubs.

With regards to dropping below LVC - it may be marked 41.5 but until I program them myself I never count on them to be exactly what I expected.

The hall sensors could be out of place but if it starts from a dead stop, then I doubt that's the issue.

What if you slam the throttle on to WOT... does it spin up or do nothing?

-Mike
 
matt_in_mtl said:
gwhy! said:
This sounds like the very same problem I was having, It was resolved by a slightly different spacing on the hall sensors. What is your current spacing for your halls ?. It only happend on my 2 newer 6fet controllers.
gwhy! I was wondering if this was the same as you were experiencing. I temporarily fitted the halls inside the motor, in the gaps where the wire is wound, spaced equally (120deg mechanical). It should provide corect timing I thought? I seem to remember something about diffferent hall placement for delta vs. star. I know other sensored BLDC outrunners that had the halls placed I did were star winding and the Turnigy is Delta... maybe that is causing a problem? I am planning on doing external mounting, but wanted to get up and running for proof of concept before spending the time. If this is the same as what you saw, then I should probably make my external hall bracket now and stop wasting time.

mwkeefer said:
Have you programmed your LVC on the Infineon or just assumed?

12S1P @ 5AH = 44.4v nominal

Normally the 48v controller would have a LVC of 41.5v (the 36v have 31.5) which is a bit too high since your lVC on a 12s pack should be around 39.5v... still what current are you drawing from the pack when the controller tripps out?

Have you done any reprogramming of the controller?

-Mike

Mike,
the controller says LVC is 41.5V. I have not successfully programmed the controller (can't get my controller to work properly, cant program my controller... I'm starting to look like a real noob :oops: ) I thought I would try it as is now that I have wiring for series connecting my 2 6s1p lip packs, before screwing around with the controller. I can tell you that according to my cell monitor and turnigy watt meter, the pack voltage didn't drop below 48V.

Have you guys connected LEDs to your controllers? I connected one to the pwr/diagnostic 1 output, and it lights when the controller is on, but has never blinked.

Thanks again for the suggestions, you guys rock.

-Matt

Hi Matt,
My older 6fet controller works really well ( 806 ) and my 12fet controller (212 ) also works well on 3 different motors without no problems, When I bought another 6 fet it was the newer (206) and before i even tried it I modded it with fets and caps and soldered the shunt but when I tried on a motor it showed the same symptoms as you describe, turn throt very slowly and you can get it to run right though the throt range but if the motor stopped you would have to zero the throt and start again or if you spun the motor up with a drill once it was over a certain speed it worked reliably it was very strange. I carried out loads of test and found nothing ( thought it might be something to do with me revamping the controller ) so I bought another (206) to test out of the box and it done exactly the same :cry: . ( I did notice that after I bought the second controller from E-crazyman that you needed to state what type you wanted, something about a peddle first controller or instant start ) this started me to look at the putting a pulse signal into the peddle sensor input on the controllers ( I got some very strange effects 8) but still could not get it to work. When I re-done my halls to 17ish degrees I retried the 2 newer 6fet controllers and they now worked a lot better ( when the motor had no load ) but when loaded up still this symptom still appears , so at the moment Im still putting it down to hall spacing ( I think ) not being quite right But it do seem very odd that my other controllers just work, so it still may be that they are peddle first controllers ( but this can be turnt on/off in the programming software :?: ) so I am still not 100% what the problem is with them. I know Jeremy Harris has used the little 6fet controllers on his converted outrunners (I wonder what controllers he has used 806 or 206) and I wonder if he has encountered any problems like this one .

Edit: typo on controller type
 
mwkeefer said:
The different hall wiring for delta vs/star would be just flipping the middle hall upside down or adding an upside down hall (I forget why) but it doesn't seem needed with the infineon on hubs.

With regards to dropping below LVC - it may be marked 41.5 but until I program them myself I never count on them to be exactly what I expected.

The hall sensors could be out of place but if it starts from a dead stop, then I doubt that's the issue.

What if you slam the throttle on to WOT... does it spin up or do nothing?

-Mike

If it is the same problem I was having I reprogrammed lvc down to the lowest setting ( didnt make any difference ) if you crack the throt open the motor will just do a quick half a second spin then stop and not move again until the throt is put back to zero.
 
Hi Matt,
I have just re-connected one of my problem controllers back upto my bike ( just to recap my memory... its not what it used to be ) and it works perfectly now even with a big 60A load placed on the motor it still spins up So the problem was defo to do with the hall spacing initially ( thats if it is the same problem that you are having ).
 
gwhy! said:
Hi Matt,
I have just re-connected one of my problem controllers back upto my bike ( just to recap my memory... its not what it used to be ) and it works perfectly now even with a big 60A load placed on the motor it still spins up So the problem was defo to do with the hall spacing initially ( thats if it is the same problem that you are having ).

Thanks, I'm happy to hear that. What you were describing sounds exactly like what i am seeing. Funny little detail, it seems like when I set it up for 1 direction it works better (I have been able to slowly ramp up to full throttle a couple times), than the other direction (never been able to get close to full throttle. Maybe 2x as bad.). I will definitely make up a hall bracket. I think the math agreed on was 17.14 degrees? (120/(14 magnets/2))=120/7=17.14

-Matt
 
matt_in_mtl said:
I think the math agreed on was 17.14 degrees? (120/(14 magnets/2))=120/7=17.14

-Matt

Yes this is my current hall spacing, You get very similar symptoms if the timing isn't quite right also but the sweet spot is there you just need to find it .
 
matt_in_mtl said:
Hi,
I have a Turnigy 63-74 that I added hall sensors to and I have hooked it up to a 6-fet infineon controller. The issue that I am having is that I seem to have found a phase/hall wiring combo that works. I can roll the throttle on gently and the motor starts to run with a low no-load current. Unfortunately if I ramp up the throttle, the controller stops driving the motor. I then have to close the throttle before it will start again. It is as if I am tripping LVC or some fault. I am running it of 12S1P 5000mAh lipo, and it is the 48V version of the controller.

Any suggestions on where I should look? Thanks in advance.

-Matt

The MCU (Motor Control Unit) was trying to protect itself from overload. Please verify your motor hall spacing. You may also try to increase the current output and/or overload limiter.
 
RESULTS!!! :D
So, I managed to reprogram the controller to up the current and lower the LVC. This didn't help. I also made a bracket to mount the halls externally. This was cool, but also didn't help. With the lowered to LVC (to 20V??) I was able to run the controller off of one of my 6S packs, and.... it worked!!! FINE!! not problems at all! So, the question is, what is wrong with the controller at 48V? Does the capacitor need to be enlarged?
 
Matt have you had any joy ?. I tried one of my problematic 6fet controllers on my other setup with what I thought 17.14deg ( but it appears that this setup is slightly off )and this setup was still causing problems. It would appear that the spacing needs to be precise to work well. I have ordered some more hall sensors and I am going to make up some pcb's so the hall sensors can be mounted exactly at the correct spacing and orientation so this should confirm oneway or another if its a hall spacing problem.
 
matt_in_mtl said:
RESULTS!!! :D
So, I managed to reprogram the controller to up the current and lower the LVC. This didn't help. I also made a bracket to mount the halls externally. This was cool, but also didn't help. With the lowered to LVC (to 20V??) I was able to run the controller off of one of my 6S packs, and.... it worked!!! FINE!! not problems at all! So, the question is, what is wrong with the controller at 48V? Does the capacitor need to be enlarged?

This almost sounds like your controller's logic power supply doesn't have enough resistors to run at 12S (50V). Add in enough resistance to drop an extra 15V or so to see if this might help. You can just add the extra resistance externally to the logic supply input (only!), but be aware that this also makes the controller think it's working on a lower voltage (so probably just leave the programming settings that work with your 6S setup in this case).

To figure out how much resistance to add, divide the desired voltage drop (15V-20V) by the current draw of the logic section of the controller (~60mA). 0.060A X 15V = 250ohms (or up to 300ohms would be OK too).

Hope this might help you!

Pat
 
Sorry for the lack of responses, I haven't gotten much done on the bike this week, I bought a new metal lathe to live at ENTS (Edmonton New Technology Society ENTS.ca) like this:
Denis_Martineau_Workshop4.jpg


so I have been spending my evenings welding up a stand for it. Hopefully this weekend I will get some more work done on the Cady.

Thanks for the replies.

-Matt
 
gwhy! said:
Matt have you had any joy ?. I tried one of my problematic 6fet controllers on my other setup with what I thought 17.14deg ( but it appears that this setup is slightly off )and this setup was still causing problems. It would appear that the spacing needs to be precise to work well. I have ordered some more hall sensors and I am going to make up some pcb's so the hall sensors can be mounted exactly at the correct spacing and orientation so this should confirm oneway or another if its a hall spacing problem.

Ok Made some pcbs with slightly different spacings up to precisely setup the halls sensors spacing and orientation and didnt make the slightest difference to this problem :evil: So it back to square one.
Lyen said:
The MCU (Motor Control Unit) was trying to protect itself from overload. Please verify your motor hall spacing. You may also try to increase the current output and/or overload limiter.

Lyen,
what do you mean by overload limiter ( phase current :?: )
 
ZapPat said:
This almost sounds like your controller's logic power supply doesn't have enough resistors to run at 12S (50V). Add in enough resistance to drop an extra 15V or so to see if this might help. You can just add the extra resistance externally to the logic supply input (only!), but be aware that this also makes the controller think it's working on a lower voltage (so probably just leave the programming settings that work with your 6S setup in this case).

To figure out how much resistance to add, divide the desired voltage drop (15V-20V) by the current draw of the logic section of the controller (~60mA). 0.060A X 15V = 250ohms (or up to 300ohms would be OK too).

Hope this might help you!

Pat

Hi Pat,
I jumped all over this this weekend . I resulted to putting a seperate ignition voltage into the controller 24v ( which works as Matt has already said ) and a power voltage of 48v and still the problem persists. There are still a few things I want to try but I ran out of time :cry: . The best I could get it was reliably running to full speed but with a slow throttle movement for the first third of the throttle once past this spot you could do what you liked with the throttle and the motor just responded as it should as long as the speed of the motor didn't drop below about a third of the max speed. When I tested mine when I said they worked after re-adjusting the spacing of the hall sensors I must have only tested them with 24v. I will re-visit the lv power lines with a scope when I get the chance. I dont think this is just related to outrunners with halls fitted, In the "monster 6 fet" thread I don't think stator (or was it Ben) fully resolved the problem either ( think he was he using a hub motor ). All I know is that its dammed annoying.
 
gwhy - about your cutout problems --> have you tried lowering the programmed current limits in your controller? The phase current setting might be too high (and/or the battery current setting too), and there seems to be some kind of current overload going on in the controller that is independant of the programmed values. I had one of these 206's cutting out at low RPM and high throttle and had to lower the value(s), although I do not recall now which one(s) fixed the problem.

*EDIT* : Is this part of the controller circuit what Lyan calls the "overload limiter"? Lyan, might you know what parts on the PCB are involved?
 
ZapPat said:
gwhy - about your cutout problems --> have you tried lowering the programmed current limits in your controller? The phase current setting might be too high (and/or the battery current setting too), and there seems to be some kind of current overload going on in the controller that is independant of the programmed values. I had one of these 206's cutting out at low RPM and high throttle and had to lower the value(s), although I do not recall now which one(s) fixed the problem.

*EDIT* : Is this part of the controller circuit what Lyan calls the "overload limiter"? Lyan, might you know what parts on the PCB are involved?

Who's Lyan? LOL
Well, this is his cousin Lyen by the way. LOL

The 6 FET controller was originally designed and intended for the European market which limited the allowable power of 200-250W. Therefore, anything over that limit can easily tripped the overload circuitry.

The best way to troubleshoot is to narrow down one thing at a time.

1. Borrow another controller and test with your motor & throttle
2. Use another throttle to test both controller & motor
3. Use another motor to test the throttle & controller

Let me know if you have any other questions. :)

Your Truly,
Lyen
 
ZapPat said:
have you tried lowering the programmed current limits in your controller? The phase current setting might be too high (and/or the battery current setting too), and there seems to be some kind of current overload going on in the controller that is independant of the programmed values.

I will give this a go, thanks.

Lyen said:
The 6 FET controller was originally designed and intended for the European market which limited the allowable power of 200-250W. Therefore, anything over that limit can easily tripped the overload circuitry.

The best way to troubleshoot is to narrow down one thing at a time.

1. Borrow another controller and test with your motor & throttle
2. Use another throttle to test both controller & motor
3. Use another motor to test the throttle & controller

Let me know if you have any other questions. :)

Your Truly,
Lyen

Thanks Lyen,
This may well be by design then,
I have tried a repacement unmodded/out of the box controller with 2 motors and 2 throttles, with no joy.
This only appears to happen with the newer (206)6fet controllers, I have a older (806) one that works perfectly.
I will try lowering all the current setting down as ZapPat has suggested to see if this helps. Do you happen to know the associated circuitry to do with this overload condition and can it be got around.

Thank you both for your help, its appreciated.
 
gwhy,

I think this may be as Lyen has said, an over-current cut out.

As far as I've been able to work out, the EB206 controllers regulate phase current by inference, rather than direct measurement. The controller measures average (no peak) supply current via the voltage across the shunt and then calculates the maximum allowable "on" pulse width for any particular motor rpm based on a combination of the throttle setting and the phase current setting. The guidance seems to be to keep the phase current setting to 2.5X the supply current limit setting, but this presupposes that the FETs are standard.

With an uprated controller, then it should be possible to increase the phase current setting, albeit at the risk of blowing FETs, to a higher ratio. The problem you run into with the 6 FET controller is that the shunt is too high a value to allow much of an increase, so ideally it needs to be reduced a bit (the bigger controllers can be fooled easily enough by programming them as an EB206, which makes them think they have a higher shunt resistance than they really have).

The test for this would be to deliberately programme a low phase current ratio and then see whether or not the problem is worse (apologies if you've already tried this). If changing the phase current doesn't make a difference then it's likely to be something else, like a momentary voltage drop under high peak current demand conditions causing the controller to glitch, as has already been mentioned. It might be worth lowering the LVC to as low a voltage as possible, in case voltage drops are the problem (again, apologies if you've already tried this).

I guess the only sure-fire way to pin this fault down might be to do a bench test, with a 'scope. This would need some form of brake or dyno to load the motor though, so wouldn't be an easy undertaking.

Jeremy
 
Jeremy is right about the 206 using the battery shunt to indirectly calculate the motor current, but I'm quite sure even the older 'infinion MCU' based controllers used the same technique (which is the 806/infinion, right?), as there are no phase shunts either. However, it is true that it sounds like it has something to do with the phase current limit rather than the battery current limit, because it only happens at low speed when the phase currents are considerably higher than the battery current.

I don't really follow that myth about the phase current value to be set at 2.5X the battery current, I have tried many ratios within reason with no directly related issues. In fact, playing with these can be pretty usefull, like on a 2WD setup where you would often want to reduce the phase current limit closer to the battery current limit, so that your front wheel doesn't spin nealy as much while accerating hard. Just remember that it's the phase current limit that's real for the FETs, so don't set this too close to what your individual or paralleled FET(s) can take.

gwhy:
--> What values are programmed into the controller right now?
--> What shunt value / how much solder was added to the shunt?

We would think that the overload limiter cicuitry works off the shunt too, so lowering the shunt value would raise the overload limiter value as well as the regular programmed current limits. It does seem to ring a bell come to think of it, as I have three of these EB206's controllers myself: 2 on my 2WD 26" bike, and 1 more on my RWD mountain bike. And yes the RWD one does cut out, but mostly only when it is cold and at of course always at low RPM. But it is also set to a pretty high current limit for a 6 FET controller. I'll have to try that bike bike out to refresh my memory (it's been many months)... after it stops snowing again that is! :|

Pat
PS: Sorry Lyen!
 
ZapPat said:
Jeremy is right about the 206 using the battery shunt to indirectly calculate the motor current, but I'm quite sure even the older 'infinion MCU' based controllers used the same technique (which is the 806/infinion, right?), as there are no phase shunts either. However, it is true that it sounds like it has something to do with the phase current limit rather than the battery current limit, because it only happens at low speed when the phase currents are considerably higher than the battery current.

I don't really follow that myth about the phase current value to be set at 2.5X the battery current, I have tried many ratios within reason with no directly related issues. In fact, playing with these can be pretty usefull, like on a 2WD setup where you would often want to reduce the phase current limit closer to the battery current limit, so that your front wheel doesn't spin nealy as much while accerating hard. Just remember that it's the phase current limit that's real for the FETs, so don't set this too close to what your individual or paralleled FET(s) can take.

gwhy:
--> What values are programmed into the controller right now?
--> What shunt value / how much solder was added to the shunt?

We would think that the overload limiter cicuitry works off the shunt too, so lowering the shunt value would raise the overload limiter value as well as the regular programmed current limits. It does seem to ring a bell come to think of it, as I have three of these EB206's controllers myself: 2 on my 2WD 26" bike, and 1 more on my RWD mountain bike. And yes the RWD one does cut out, but mostly only when it is cold and at of course always at low RPM. But it is also set to a pretty high current limit for a 6 FET controller. I'll have to try that bike bike out to refresh my memory (it's been many months)... after it stops snowing again that is! :|

Pat

Yes my old 6fet is a 806 and has been modded and reprogrammed for 40A battery and 100A phase, with a soldered shunt that given me about 80/90A battery current max ( shown on a watt meter ) , this works well and have not given me any problems. My newer 6fet 206 is setup in a identical way and plays up. I also have a brand new 206 6fet with no mods also not reprogrammed its a 48v 300w type this also plays up on my 2 motors. I will remove the solder from the shunt and set bat current to 15A and try some different phase current ratios and see what happens. I also have a modded 12fet (212) setup the same as my old 6fet but this works as well as the old 6fet. So its beginning to look very much like some form of over current protection on the newer 6fet controller. I will do some reprogramming and report back the results.
 
Jeremy Harris said:
gwhy,

I think this may be as Lyen has said, an over-current cut out.

As far as I've been able to work out, the EB206 controllers regulate phase current by inference, rather than direct measurement. The controller measures average (no peak) supply current via the voltage across the shunt and then calculates the maximum allowable "on" pulse width for any particular motor rpm based on a combination of the throttle setting and the phase current setting. The guidance seems to be to keep the phase current setting to 2.5X the supply current limit setting, but this presupposes that the FETs are standard.

With an uprated controller, then it should be possible to increase the phase current setting, albeit at the risk of blowing FETs, to a higher ratio. The problem you run into with the 6 FET controller is that the shunt is too high a value to allow much of an increase, so ideally it needs to be reduced a bit (the bigger controllers can be fooled easily enough by programming them as an EB206, which makes them think they have a higher shunt resistance than they really have).

The test for this would be to deliberately programme a low phase current ratio and then see whether or not the problem is worse (apologies if you've already tried this). If changing the phase current doesn't make a difference then it's likely to be something else, like a momentary voltage drop under high peak current demand conditions causing the controller to glitch, as has already been mentioned. It might be worth lowering the LVC to as low a voltage as possible, in case voltage drops are the problem (again, apologies if you've already tried this).

I guess the only sure-fire way to pin this fault down might be to do a bench test, with a 'scope. This would need some form of brake or dyno to load the motor though, so wouldn't be an easy undertaking.

Jeremy

The bench test looks like it will have to be done either way, It shouldn't be to bad to do as this happens with no load on the motor. I have upped the phase current ratio a tad but have not tried lowering, I will remove the solder from the shunt and try some extreme ratios to see if this throws any light on the subject.
 
Back
Top