Fligh High wrote:
If I am at the building stages will open a new topic here at motor-conversions "Fligh High's alternator motor conversion experiments"
Maybe people will find it interesting....As I said before think this will be a good year for Alternator-conversions , hopefully we will all have some good and fun results !Because fun is what it's all about.
Could you tell me something more about the micro-controller you are using (type) Atmel it was ? programmable in C ?(edit:already found Atmel AVRMega 8535
) The home-made controller with hall-sensors is THE way to make an alternator work like it should I think.
Also how did you make the space for the hall-sensors in the stator? drill at an angle , dremel ? Because if you touch the copper wire ...... that's really bad
I myself will probably use a different approach first with externally made pill-magnets exiter-disk at the output shaft because I have LOTS of alternators that I want to try out and ease of manipulating the hall-sensors a bit but will certainly make something internal in the alternator eventually .
Also have you tried running your alternator with your made controller and the car-type battery already and with full power/revs , and do you want to use 12 or 24 volts on the buggy ? Or is the Fet-bridge/driver not yet up to it for trying to go to the max testing.
I'm sure people will be interested in the tread, Please Do so, I'll help as far as I could.
Yes, I did used the 8535. Use INT0,1,2 to detect hal sensor level change, two of the interupt support level change but 1, only support the raising/failing edge, hence execute the CONFIG interrupt to change the toggling modes in every interupt to detect next edge. On every hal interrupt, the program will reads the hall values and use it up as an index for the commutation stage. Its a 8 bit data, which 3 bit MSB used for the high side control, and 3-bit LSB for the low side. If you will I could provides you with the commutation table. The low side are staticly drove by TTL, while the High side was PWM drived using timer1a&b, and timer2.
Its working ok, but with lack of high speed, hence need to enhance the logic such with sinusoidal PWM, but this mean I have to redisign it again as I was out of PWM driver already. For the future, I'm planning to use 74LS138 for decoding one PWM to the designated high side, and used the saved Timer to drove the Sinusoidal ramp. The 2101, works flawlessly, only accidental EMF spikes due to not properly connected cable to the brush had let one IC blow of and creating a shoot through in one of the phase, but the all N-channel FET arrangement works OK. I droved the sensored alter-motor with HK Car ESC in sensored mode and it works very well, hence confirmed that this is mostly due to not really sophisticated logic
, and I need to enhance it quite significantly.
My ebuggy used 12V, but my ultimates goal is to used an Alter-motor to drove my 280GE front differential, John500 is helping me to look for 24-100A alternator, which will be driven by 48V supply
Anyway, been quite busy this weekend
with my stator (please forgive my BB's picture quality) :
1. Went to the machining shop, but before that marked the area which need to be removed with marker pen. Then drill them out. Its a 2 dollar jobs
Please be really careful for the machine not to touch the stator wire, but 1-2 scratch wont kill you and easily insulated by thin CA.
2. the results then looks like this
Please note that the sensor will be positioned just before the respective phase, and as the each phase interleaved with each other, on bosch/ND this would mean that each sensor will be separated by 1 tooth.
3. Wired up the hall sensor, previously I'm using the honeywell SS40A open collector, but hate the mess with the pull up resistor, hence currently I'm trying the totem-pole output of the SS411. Did have a problem previously with the pin assignment but it is turn out to be +, - , output.
Again, sorry for the picture quality. But I found Blackberry is the fastest medium for capture and share this project.
Please assure that all sensor was working before progressing further, as you will need to glue it up firmly to the stator. Found one which was not working after it has been glued, has been proved to be very frustrating.
The SS40 & SS411 was quite small, but as the space really tight, I removed its sides using a dremel, also rounded the edge a bit. Based on my experience it was OK to removed the BACKSIDE on the sensor a bit, but NOT the front side.
So far only able to wired 1 sensor before going out to the office, so will be continue later on.