Modifing an ESC?

[DIYMark]

hi im new here and ive heard that there are very knowledgeable people here and even electrical engineers :) So i was after some help...

Ive been thinking that if you get a "low" power brushless esc that doesnt have a bank of mosfets (to make things easier), just the six required for typical drive of a BLDC, could you unsolder the factory mosfets and run the conrollers output wires onto your own 3 phase H bridge in an attempt to get more current and more voltage (the separte H bridge can be run of a higher voltage and current but shares the same ground connection with the contoller).

i see this as possible from what i think but then again im only 16 and probally wrong :'(

Could any one see if there is a reason why this wouldnt work?

Also i mentioned a low power esc becuase they normally only have 6 fets instead of the huge banks in high power ones (in parralell) and the 18 - 20 amp escs are the ones without this bank of fets so im hoping of trying it on my 18a esc - also it has two pcbs in it, a h bridge one and a controller one so this makes tapping into the output tracks easy.

Please tell me otherwise if this wont work before i open it up :D

[jeffs555]

It could possibly work to increase the current with some controllers, but raising the voltage would be tougher. First off, on the controllers you are talking about with only three wires to the motor, the motor wires are used as both outputs and inputs. At any time, the controller will only be driving two of the three wires. The controller will read the voltage on the third wire to determine when to switch which wires it is driving. You would have to continue to connect these wires to the motor along with the outputs from your new H-bridges. The input voltage is typically scaled down with resistive dividers to keep it in a useable range. You would probably have to change these dividers to increase the voltage rating. Since there are many ways they could be wired, you would have to trace the circuit and understand how it works to know how to change them. Also, the gate driver needs to be able to switch to the full supply voltage so it might not be able to handle the increased voltage. The input capacitor on the battery wires would also probably need to be replaced.


Using more or larger mosfets to increase the current rating may or may not work. Some esc's have over-current sensing and if so you would have to trace it and understand it to know how to increase current. Also, there may be insufficient drive for multiple or larger mosfets. Mosfets have input capacitance that must be charged or discharged by the gate driver each time it switches. Increasing the capacitance without increasing the gate drive current would increase the switching time which could cause problems.

[DIYMark]

Oh well thank you for pointing that out, looks like I'll have to re think ideas :rolleyes:

After building my own sensored BLDC driver i still had sensors in my mind and forgot that rc esc's are sensorless (majority of them) and i dont think sending higher voltage along the 3 outputs cables as back emf will be good :D

...put voltage regulators in the back emf circuit (smirks naively), nah that wont work lol!!!

Anyway so now i have 3 choices...

1. Rewind a 42 pole bldc motor (11 inches in diameter) for lower voltage use - bit time consuming...

2. Run it under voltage seeing as most rc esc's are only 20 ish volts - waste of a motor...

3. Make my own sensorless bldc driver - now thats a good idea....

Ive orderd ics before for my own bldc esc but it was sensored, would anyone know of a manufacturer that sells ic's that are for the commutation of NON sensored bl motors (sensorless)???

[AndyKunz]

Only to 20V? Have you ever seen one of these?

50V ESC (http://www.castlecreations.com/products/phoenix_hv-110.html)

There are lots of other vendors - Shulze and BK come to mind immediately. Anybody doing anything with any kind of performance market will have something.

Andy

[DIYMark]

hmm i was more after 100 volts :) btw this isnt exactly rc use

[KreAture]

At 100v you would think a bit extra during layout due to the high voltage and potential problems it entails. It would have to be a tad larger than the most compact models.

I suggest you see if you can modify one of the open designs from the BLDC thread. One of the concerns you will have is the BEC section. A 95v drop is quite large so you will have to address this. Probably use a switchmode regulator.
The voltage itself is no concern as you can get FET's that handle it. Be aware however that they may require drivers to allow proper switching due to the higher capacitance of the gates (as mentioned by jeffs555).

The bonus is ofcource I assume there won't be too high current in it with the high voltage? This makes it a tad easier again as high current can be quite pesky.20-30 amps is easy even at 100v, higher becomes more advanced.

[DIYMark]

Read next post; i accidently made a new one instead of modified and i cant delete this one :D

[DIYMark]

ive decided to scrap modifing an existing esc and just make my own...

yea the highest current in my situation is 7.5 amps so lets say 10A max.

the problem is this motor i have has hall sensors at 8.5 degrees and the sensored bldc controller i made a while back was to use 30 or 60 degree sensor spaces....

also becuase the sensors are on a flexible pcb that slots between 3 of the stator windings, one "crude" way to use my existing controller (140volts, 20 amps) would be to cut the sensor pcb, relocate them to 60 degree spacing then re wire them....

this would be a last choice thing though, id much rather make a sensorless bldc driver but i cant seem to find any ic manufactures that make sensorless bldc drivers that have such high ratings.

the only problem i see (If i can get a sensorless bldc driver) would be the high voltage back into the driver from BEMF (im sure this is a problem?) but could it be possible to accomplish with with the use of optocouplers?

also does anyone know of a few SENSORLESS bldc ic manufactuers?

and finally if i cant find anything i might do a data sheet search for the ic in one my rc sensorless esc and see if i can find a schematic of a suitable circuit then modify circuit and use the ic for a high voltage esc...

[AndyKunz]

You can find sensorless BLDC driver chips in the industrial electronics places and commercial white goods suppliers. I believe IR has one, a few others as well. They are used in air conditioning systems, washing machines, etc.

You would still do well to copy the concepts out of one of the BLDC threads. They will be more suitable. A 1000W inverter isn't such a big deal, you should be able to do it relatively easily.

Another option to hacking the motor for a sensored control would be to use an optical encoder.

Andy

[Bruce Abbott]

To do sensorless you could pass the bemf voltage through a voltage divider to drop it down to a safe voltage. However, sensored is much simpler to implement in the controller, and allows operation down to zero rpm.

[KreAture]

DIYMark
I think you are confusing the sensor placement.

If there are more than 2 poles in the rotor you have to divide down the distribution.
A 60 degree setup would mean you have 4 poles, 2 pairs. 120/2 = 60.
A 12 pole rotor would have 6 pairs making 60/6 = 10 degrees physical spacing, but still 60 degrees electromagnetic spacing.

The motor you have is set up with 8.5 degrees you say. This leads me to think it has 14 poles, typical in LRK systems. If it does, then the spacing is still 60 electromagnetic degrees and your controller should work. It is just a matter to matching the sensor connections and phases so the signal makes sence to the controller.

[jeffs555]

Like KreAture said, you are confusing physical degrees with electrical degrees. I would suspect that if the sensors came with the motor, they would be in the correct position for commutation.

[KreAture]

One thing though... If the sensors are sensing on the actual rotor magnets, they can be confused by the drive-field. We run theese motors to the point where the rotors magnets are cancelled out by the stators field.

[dleroi]

You might find this helpful:

[DIYMark]

Like KreAture said, you are confusing physical degrees with electrical degrees. I would suspect that if the sensors came with the motor, they would be in the correct position for commutation.

Nup, my motor has sensor spacing of 8.7 ~ degrees (mechanically) its 42 pole.

i already have that pdf btw thanks any how...

and ill check out ir's site secondly an optical encoder sounds really good but it may be difficlut to mount ot this outrunner; ill look into it more

[jeffs555]

How do you know that it is 42 poles? You said "the sensors are on a flexible pcb that slots between 3 of the stator windings", and they are spaced ~8.7 mechanical degrees. That would indicate that there are 42 slots on the stator. Are you sure you are not confusing slots or teeth on the stator with poles? They are rarely the same number, and could be greatly different. If this is a commercial motor, I can't think of any reason it would come with three sensors in a totally wrong position for commutation.

[dleroi]

DIYMark,

Could you post some pictures of this unusual motor, so we can all understand it better?

[KreAture]

You do not have to have the sensors spaced relative to the pole number. The important thing is that they are 120 degreese (or 60) spaced relative to the magnets angular placements. 1 set of magnets (n and s) is a 360 degree area in itself. With only 2 magnets, they cover 180 deg each. With 4 they cover 90 etc...
I am sure it is only a matter of linking the controller to the sensors and signals correctly.

[DIYMark]

ive been trying to avoid a huge post about the motor description because it can be considered as odd but here it does... :rolleyes:

its from a fisher and paykel smart drive washing machine and it is basically a huge out runner motor that is capable of turning a full drum of washing @ 1000 rpm during the spin cycle WITHOUT a gear box - in other words i chose this motor for its torque and power (~900watts).

However, being from a washing machine it needs the "agitate" the washing load in order to wash it...and that is why my sensors are really close together.

Because if they were @ 60 or 120 degrees it wouldnt really be any good at spining a half turn backwards and then a half turn forwards (to agitate the washing) so the manufactuer opted for a close sensor spacing so they could achive half (or less) turns so that can agitate the washing.

Heres a pic but it hasnt got the pcb that sit between 3 of the windings.

Picture of a Smart Drive Motor (http://www.watchtv.net/~rburmeister/Smart%20Drive/Smart_Motor.JPG)

NOW, if you look at the picture the motor it has an "exposed" star point thats next to the 3 blade connector...(its wound in "Wye") so i can change it to delta wound if thats any better but all that i need to do is to turn this damn thing with a controller lol - funny how we need an elaborate controller to remove a simple mechanical thing like the commutator.

Any how, Sensored would be nice but my sesnors are at wierd intervals so i could move them and have a bit of a spiders web wiring job OR make a sensorless controller.


Currently im looking at some ic manufactures of sensorless bldc chips but all have low voltages.

Lastly if you need a GOOD generator get one of these motors and hook it up to a 3 phase rectifier (6 diodes) and there - instant generator, good thing is that with a lazy turn of the rotor with my hand made ~6volts so imagine the potential power.

PS - in the pic the thing that looks like a small panetery gearbox is a bearing/seal assembly.

[DIYMark]

...I am sure it is only a matter of linking the controller to the sensors and signals correctly.

im positive it isnt because i tried it logically after inspecting positions of sensors etc then wired it up and turned my controller on and all the motor did is hum back and forth by 2 mm because the next sensor picked up the "signal" too quick (close spacing of sensors) and then the controller reversed the motor before it did 60 degree (like my controller is designed for).

[jeffs555]

That is definately not a 42 pole motor. From the picture, that would be a 14 pole, 42 slot motor, and ~8.6 degrees would be the proper physical spacing for the sensors. You probably just had the sensors wired to the wrong phases or maybe the sensor signals were inverted. You asked for advice from experts, and you have been getting good advice, but you are refusing to believe it.

[DIYMark]

slots...poles? im confused :confused:

so if my sensors are between four slots (3 sensors), side by side what is the degrees of it?

[DIYMark]

im not refusing to belive it but im wondering how sensors placed like this http://img160.imageshack.us/img160/1509/sensorplacementyi4.th.jpg (http://img160.imageshack.us/my.php?image=sensorplacementyi4.jpg) are at either 60 or 120 degrees spacing wether it be mechanical or electrical?

Heres more pictures...

Motor View
http://img296.imageshack.us/img296/7874/motorlg0.th.jpg (http://img296.imageshack.us/my.php?image=motorlg0.jpg)

[Bruce Abbott]

Degrees are irrelevant. Imagine cutting halfway through the stator and rotor, then flattening them out to make a linear motor. It would still work (linearly) with the sensors in the same positions, even though they are now spaced at an 'angle' of 0 degrees! Now imagine discarding all but 3 stator teeth and 2 rotor magnets, then bending the stator and rotor back into circular form to make a 2 pole motor. It would still work with the sensors in the same positions (now 120 degrees apart).

The sensors in your motor are exactly where they should be for a normal ABCABC winding scheme. For another example, take a look at this CDROM stator. Sensors are spaced 40 degrees apart, but note where they are relative to the stator teeth.

[Bruce Abbott]

im positive it isnt because i tried it logically after inspecting positions of sensors etc then wired it up and turned my controller on and all the motor did is hum back and forth by 2 mm because the next sensor picked up the "signal" too quick (close spacing of sensors) and then the controller reversed the motor before it did 60 degree (like my controller is designed for).The problem I have found with sensored motors is that the sensor signals must match the phase outputs, in both order and polarity. There are lots of wrong combinations but only two(?) that are correct. If the configuration is not correct the motor will either just jitter back and forth, or run rough and draw a lot of current.

[DIYMark]

now im really confused???

heres how i think it works (please correct me if im wrong)

My bldc driver operates on sensors at phasings of either 60 or 120 degrees. say if i set it to run on 120 degree phasings that means every 120 degrees of a turn (1/3 of 360) the driver expects feedback via the hall sensors and then it powers the next phase in the sequence after this signal.

Now if my sensors are at a stupid interval of 360/42's of a turn (decimal equivilant of ~8.6 degrees) and the controller assumes a 120 degree interval they will obvously get mixed up and not work hence the jittering because it powered the next phase WAY to early and reversed the motor.

So the context ive been using degrees in is as a fraction of a turn - ie. 120 degree phasing is 120/360 which is every thrid of a turn...now im not sure if this may be it but its likley that this has caused confusion.

[DIYMark]

Degrees are irrelevant...

sensors are the ONLY way my driver can sense when to power the next phase which has to happen in the proper order which only happens when the Degrees of the sensors match those of the driver.

[KreAture]

DIYMark
You are confusing how sensors work.
The sensors on any brushless motor is placed so that only one of the sensors change at any given time. There are 2 possible methods, direct and advanced. Direct means the sensors will change at the point when the power should be moved to the next stage in the drive cycle. (There are 6 commutation steps in any 3-phase brushless motor.) Advanced means the sensors change 30 degrees too soon, giving the controller a heads-up of theese 30 degrees. Theese are electrical degrees and are essentially the time between each commutation divided by 2. This is very uncmmon and even if you used a direct mode ESC with a advanced motor it would still work.

The sensors are spaced so that they create a code that will explain the actual rotor position. Since the magnets keep repeating around the rotor, this pattern also repeats. This means that for a 14 magnet-pole motor, there are 7 valid positions around the motor for each of the 3 sensors. There is no rule saying they have to be placed by the same magnet-pair.

Back to your motor...
You need to first verify that your sensors work and see what pattern they create.
You also have to verify what polarity they use. Look at the original circuit to see how they are powered. Then supply them with power and note the signals generated as the motor is moved 1/7 of a turn. 2/7 of a turn would make the pattern repeat once more.

You should see one bit change at a time.

The hard part is matching your drivers phases with the signal. Hopefully it is not inverted, and you just have to connect up the sensors and try all 6 combinations of phase-positions. If sensors are not inverted, one will work.

[KreAture]

I forgot to mention. The simple way to make a brushless controller using sensors is just to have a lookup table and a pwm generator.
The table works such that for each of the valid codes that can be received from the hall sensors, there is a new code consisting of the 6 FET outputs signals. 3 bits for the high side and 3 bits for the low side. The controller just looks up the word for the new code and outputs it any time one of the hall sensors change. The pwm controller is driven almost directly from the received throttle signal.

More advanced sensored brushless controllers work in exactly the same manner. They know what phases to drive based on the word. They do not just step through the cycle of words whenever they receive a pulse...

[KreAture]

Hmm, after looking closely at your pictures and some others in your folder I wonder...
- Did you rewind this motor ?
- Are there only 1 magnetic pole pr magnet segment in the rotor ?

[DIYMark]

no i havent rewound it - i was thinking of it at one time because i was going to use a hv rc esc but even then the resistance would never be low enough to get the current i wanted to have to same power as if i were running it on 120 volts.

hmm not sure about the magnets, all that i know is the motor is star wound with 15 ohms resistance between phases (2 coils each being 7.5 ohms) but the motor does work (should work) becuase its stock and i havent changed any thing.

As for my sensors i could check them by hooking up a 5 volt dc supply across them and an led from each sensors out put back to ground, then turn the motor by hand and see if the pattern/code/"binary" matches with my bldc driver ic data sheet and their hall sensor input logic table?

i see what you mean also with the code; my explaination would only work with a 3 phase 3 pole motor if they did exist :( oops.

[KreAture]

Start by checking those magnets.
Use a small screwdriver or similar and just drag it along the magnets noting where there are strong poles. I would expect you to find a high number of magnetic poles. Not 14. With 42 stator poles you would either have a lot less poles or a lot more. Maby something like 28 is likely it is wound ABCABCABC-> style. I cound 14 magnets on your first picture. I suspect theese are 2-pole magnets.

[jeffs555]

So the context ive been using degrees in is as a fraction of a turn - ie. 120 degree phasing is 120/360 which is every thrid of a turn...now im not sure if this may be it but its likley that this has caused confusion.

Again, you are confusing mechanical degrees with electrical degrees. The 60 or 120 degree specification for a 3 phase brushless controller refers to electrical degrees, not mechanical degrees.

Mechanical degrees refer to one complete mechanical revolution as 360 degrees, just like on a compass. Electrical degrees refer to one complete electrical cycle as 360 degrees and have nothing to do with mechanical position. For a 3 phase brushless motor, one complete cycle is usually 6 commutations, and each commutation is 60 electrical degrees apart. If the motor was designed to turn 1 mechanical revolution for each electrical cycle, then 60 electrical degrees would correspond to 60 mechanical degrees, and the sensors would be physically spaced 60 degrees apart. However, many brushless motors including the one you are talking about, require several electrical cycles to cause them to turn a complete revolution. The motor you are talking about appears to require 7 electrical cycles to cause it to turn one mechanical revolution, and each electrical degree corresponds to 1/7 of a mechanical degree. The sensors would be spaced 60/7 mechancial degrees apart, or approximately 8.6 mechanical degrees.



I am not an expert in brushless design, but I would suspect that what appears in the picture as 42 coils and 14 magnets is exactly that, 14 magnets with alternating polarity(ie 7 magnets with positive poles to the inside, and 7 magnets with negative poles to the inside). It could work with the phases star connected if each adjacent coil were wound or connected with opposite polarity. Calling the phases A,B,C and using + and - to indicate oppositely wound or connected coils, the coils would be A+, B-, C+, A-, B+,C-, A+, B-,C+, etc. This should commutate normally using your standard 60 degree controller as long as you get the correct sensors connected to the correct phase, and get the polarity of the sensors correct.

[DIYMark]

yes theres 14 magnets and when i did this last night i dragged a screw driver around and ther ewere somthing like 3? places the screw driver would stick on one magnet?

[KreAture]

Aha! 3 places pr magnet you say ? Were that 3 strong spots or just 2 strong spots and one weak one?

I still say this is set up right for running.

Oh, and you have to check your sensors. They may be analog hall sensors instead of digital. That will confuse your ESC a lot!

[KreAture]

jeffs555
We can clearly see from his pictures that the motor is wound ABCABCABC(repeating ofcource). This would indicate he should have either 28 or 56 magnet poles. If he has 3 pr magnet he would have 42 making the rotor and stator identical, that would not work.

[DIYMark]

i checked once again and the magents have 4 poles per magent so 56 is right :)

now the sensors, their on a pcb board and they are all SMD, there are 3 of them and each sensor has a resistor and capacitor next to it (i think their resistors and a cap - small black box and one bigger brown box; both being SMD) could this have been used to invert the signal?

If not, and my sensors are in the correct postion, then they must be broken which for this motors sensors is common due to corrosion. so what ill do is get a new sensor assembly for this motor (not cheap!!! they are way over priced at AU$40 a unit) and try my esc (once i build a second one - i am making a few upgrades) and if that dosent work (but it SHOULD) i might have to "make" the sensors with this Hall Effect Sensor (http://jaycar.com.au/productView.asp?ID=ZD1902&CATID=&keywords=hall+effect+sensor&SPECIAL=&form=KEYWORD&ProdCodeOnly=&Keyword1=&Keyword2=&pageNumber=&priceMin=&priceMax=&SUBCATID=) and wire them up to the motors existing board - however this is my last preferance but i might have to resort to it if A the existing sensors are inverted, or B they are not digital.

if my signal is inverted can i change this?

also KreAture said digital meaning "on off" apposed to analogue being "ramp on ramp off"?

[KreAture]

Don't get new sensors. It's a waste of time and money.
You need to check the ones you have by applying correct power and measuring their output. Analog sensors will give a signal that increases more and more as the magnet approaches, then reduces again and swings the other way for the other magnet polarity. Digital sensors will give a high signal out when one polarity is close, and a low signal out when the opposite is close. It will switch between theese two states directly, and never float around inbetween. Check your ESC's docs to see which one it needs.

[DIYMark]

All i could find about my sensors in the data sheet for the esc ic im using is

"The sensor inputs are designed to interface directly with open collector type Hall Effect switches or opto slotted couplers..."

im kind of leaning towards using those senors that i showed in my last posts hyperlink only becauae they should work and then i could rule out sensors as a problem?!

[KreAture]

Well, sensors are generally matched to the specific flux available in the motor.
Open collector type would mean a kind of transistor-driven sensor ergo one that does not deliver a analog signal. Your ESC thus requires digital sensors.
Now you need to check the ones in the motor to see if they are analog or digital. I bet you they are analog as those are cheaper...

[DIYMark]

Well ive got the sensor board and hooked up a + voltage supply (3 x 1.5 volt AA batteries - normally 5v? is used for sensors but 4.5 is close enough?) and i hooked up the 4.5 volts across the Vc and GND pins of the sensor (left and middle if your viewing it from the branded side) and then i put some leds across the pins labelled OUT and GND. I turned it on and guess what - ALL the leds lit up :(

(and to add to all my confusion this just had to happen :P)

I passed magents near the sensors and even re fitted them to the stator and spun the rotor but thay all stayed lit up no matter what.

So next i tried connecting the leds between Vc and output of the hall sensor and still nothing. As a last resort i got another sensor board and still the same results.

Furthermore, its extremly unlikley that all the sensor just "broke" at once so im guessing im connecting it wrong?

But connecting a voltage supply to the sensor then a led between output and ground and passing a magnet over it should work, shouldnt it?

[KreAture]

I think the sensors deliver a current regardless and this lights up the LED's.
Try not loading the output at all. Just use a DVM (digital volt meter) and measure the signal on the sensors output. Digital sensors will switch between 0 and 5v while analog usually deliver 2.5v center and swing a bit to each side as magnets pass by.

[DIYMark]

ill try that but a while ago (10 years) my dad had a device he made for testing this motors sensors and what it was is a box with a plug to accept the sensors wiring and 3 leds plus a nine volt battery.

It was hooked up as i did above (but he used a resistor to drop down nine volts to 5v) and he said that as the rotor was spun it would light the leds in a sequence (depending on the direction it was spun) and only one would light up (one bit changed at a time)

So if turned one way led 1 was on and 2 and 3 were off. Turn it one more "click' (the motor jolts into place if turn a little bit) and led 1 and 3 were off and 2 were on and finally once more and leds 1 and 2 were off and 3 was on - this pattern would repeat its self.

Anyway im off to test the sensors with a multimeter (Fingers crossed that it works :D)

[KreAture]

You have the sequence wrong.
The system "changes" one bit at a time. Not just lights one at a time.
If one at a time were lit, you can only show 3 different states. (1 2 or 3 on)
It displays a code, and changes only 1 bit at a time. This means that binary speaking it can be 101 and then change to 100, 111 or 001 but never directly to 010, 011 or 110. It's a special form of counting and it allows you to trigger on change on any of the 3 signals, and use it directly to look up your next commutation-state.

[KreAture]

If the hall sensor sequence seems unclear, look at page 24 in this document:
http://www.freescale.com/files/microcontrollers/doc/ref_manual/DRM046.pdf
It shows how each hall sensor code can be directly used to know how to drive the motor at any given time. Notice how each sensors signal is on half the time and off half the time and how they are 120 degrees offset from eachother

[DIYMark]

Okay, i got a DMM and checked the Hall sensors.

Same setup as last time but we used a voltmeter instead of a LED and passed a magnet next to the hall sensors (cobalt motor in this case - it what was around at the time) and we saw small changes in the reading (quater of a volt). So we checked the other 2 sensors and through this checking i realised that i didnt move sensors in the magnets field (kinda hard if your using a motor) so i went back and checked the rest of the sensors but this time moving them in the magnets field and they checked out okay.

BUT, what happend was they were switching between 0 and 5 volts so thats good (their digital) however, they were "latching". I'm not sure how to describe this but if you move the magnet near it say the sensor turns on, then move the magnet away and its still on, its only when you move it next to the sensor again that it turns off...

Now this is most likley a stupid question as i have never used hall sesnors before but is this the typical function of hall sensors?

[KreAture]

Yes.
They are latching to N and S. After seeing a certain N level they stay high till they see a certain S level. This makes sure you get nice and clean signals.
This means all you need to do is hook up your sensor wires correctly relative to your phases. If the phases and sensor outputs are marked, this should be easy. If not, Try to count your way around the stator to the position of the sensors. That should give you only 2 options as to sA sB sC vs phases A B C.

A B C A sA B sB C sC A B C

:)

[DIYMark]

Wow finally i might get this motor to turn!!!

So everything seems to check out and all i have to do is make a second sensored controller.

First one was a prototype with transistors used for the 3 phase H bridge - i couldnt find and fets and This might have caused it to "not work"...

But for my new one ive changed the Circuit board a bit and i will be using 150volt 45 amp mosfets so this thing could drive a motor up to 6Kw.

But for my use 120 volts 8 amps is the max ratings (ive used 45 amp fets because they were way cheaper than less current rated one and i might get away without a heatsink).

The sensors are named A B and C and the schematic i have labels the Hall Sensor inputs 1 2 and 3 so this time i can definatley know everythings hooked up right.

I'll keep everyone updated when i finally get this motor going - but ill be a while (im on a 2 week break).

Anyway, thanks for all the help everyone. It was really appreciated, i wouldnt have got this far if it wasnt for all the knowledagble people here :D Thanks alot guys!

[KreAture]

Glad to hear :)
My method of throwing information at a problem till it goes away always works ;)

[Ron van Sommeren]

BLDC demo's and tutorials on www.freescale.com (http://www.freescale.com/), e.g.
http://www.freescale.com/files/microcontrollers/doc/train_ref_material/MOTORBLDCTUT.html

diy brushless controller designs:
http://www.rcgroups.com/forums/showthread.php?t=140454

[Ron van Sommeren]

Excellent motor building articles by Brian Mulder, a must read, will prevent you from asking a lot of questions you even did not know you were going to ask ;) ;)
http://www.southernsoaringclub.org.za/
-> Articles by SouthEasterners.
-> Electric Motors - part 1-5

Do-it-yourself motor homepages, manuals/tutorials, checks and tests in this motor builders tips and tricks thread. The checks and tests may save you from frying your controller or motor. Thread is active, bookmark it for future reference and subscribe to it:
http://www.rcgroups.com/forums/showthread.php?t=240993 (sticky thread, at top of subforum)

Two instructive motor winding/assembly videos, 14 & 10Mbyte respectively.
http://www.utahflyers.org/
-> Motor Winding #1 & #2