e-Buggy: Alternator Conversion

@tyler: Surely will shared it up and thank you

@liveforphysics: I'll be the "Java Man", well I am in fact a Javanese :D. Anyway, I'm planning to implement a two phase commit, If I may use the term, i.e. switch all Fet off first, perhaps put a nano second dead time, the switch to next switching arrangement. However, as I don't really have an discrete electronic background (except for computer engineering plus a postgrad in telco), don't really know if this will cater the parasitic capacitance in Fet's gate, so what I'm planning to do is to also implement a fix switch state on the low side and the PWM pulse on the High side which was said to mitigate passthrough at certain degree. Do you suppose a nano second dead time is sufficient ?

On the Fet, the main reason that I'm using the IRF3007 is because it was the biggest Amps & Vds rating available in the clearance sales :lol: . The IRFP4468PBF was quoted for SGD 54.38 each for 290A/100V while mine was SGD 2.24 for 80A/70V, still both Fet have a 4V Vgs threshold I supposed we could just swap it out when needed. Still you are right that it will runs cooler . Perhaps if later on I could get my self a those high amps American alternator might go for this one.
 
This is definitely in my top 5 threads right now, very cool project! Regarding using IGBTs for this; I have been looking forward to an IGBT based design for a while but since the voltage needs to be high before IGBTs are viable I haven't seen it yet. What voltage do you think the alternator will sustain?
 
Since we had removed the diode, I suppose the limit is how good the lamination are and the max amps rating, so perhaps first we could try it for says 80V and 80Amps ? :twisted:

Anyway, as this might like reinventing the wheel, had to confirmed that all stage are really completed, not just based on assumption, hence below is the sensor output (through LCD for CW & CCW rotation)


as it could be seen, there are 6 steps i.e. A--, AB-, -B-,-BC, --C, A-C and in reversed order in the contra rotation, the up most important thing is to assure that all sensor align with the phase (i.e. Right phase to the right cable) hence will have to put the label later on in the input cable.
 
IT Works..... :D :D :D :D :D



Notice the RPM is quite low due to the fact that its only using TIP transistor rated at 15A/60V, now progressing to Mosfet driver construction
 
Widodo

I believe you are posting on a RC forum, correct ???

If so, I have been trying to understand this electronic stuff, enough, to try to run my Brand NEW 24V 150 AMP Alternator I just had shipped down from the States. If it does as stated, it would make a good candidate for my E Motorcycle or reverse Trike.

I just can't get the Electronic stuff to sink into my head. :roll: :roll:
 
Harold in CR said:
Widodo

I believe you are posting on a RC forum, correct ???

If so, I have been trying to understand this electronic stuff, enough, to try to run my Brand NEW 24V 150 AMP Alternator I just had shipped down from the States. If it does as stated, it would make a good candidate for my E Motorcycle or reverse Trike.

I just can't get the Electronic stuff to sink into my head. :roll: :roll:

Yes Harlod you are right, I was logged in RC group as vytis rex, my youngest son name, aeromodelling is where its all began :D

Start my first exposure with the BLDC motor in my Airwolf-Trex heli
airwolf3.jpg


and my first micro-controlled helicopter in 2007
4ch5.jpg


But then I found that It must more fun being an aeromodelling mechanics :mrgreen:

Anyway, Congratulate on your new endeavor in alternator conversion :D :D , will try to help you anyway I can, but first could you be so kind to point me on how to getting those alternator ? :wink:

I assumed you already know how to convert the alternator into a brushless motor, but if not then this is the step that you need to take:
1. Remove the Diode plane from the alternator. In the BOSCH alternator, you need to unsoldered 6 wire in the Metal/Plastic diode plane in the back of the alternator, which then will leave you with 6 cable (2 cable each phase, use multi-tester to check which connected to which one), then you need to rewire into star or delta.

2. you need to use the existing brush to supply the rotor, hence you have to keep the integrated regulator and brush intact, but please assure that you cut both wire coming out from the regulator as it will intervene with the rotor supply.

Now you got you self a very powerful alter-motor unfortunately its a sensorless , hence you need to have a speed controller that also support sensorless commutation. In my case, I used a RC car that support both type of commutation. eHirobo got a XERUN controller in http://www.ehirobo.com/shop/product_info.php?cPath=354_364&products_id=8727, it capable on handling 150Amps, and 6S (around 28 Volts max)
or like my eBuggy, you could use http://www.hobbyking.com/hobbyking/store/uh_viewItem.asp?idProduct=10370, support 150A but only with 4S voltage handling.
With those controller, you will be able to immediately runs the motor in the sensorless mode, with the option to turn it off into a sensored when you're ready.

Or you could go directly into kelly's controller or other Sensored BLDC Controller by installing hall sensor like I did in this Nippon Denso 35A alternator.
IMG00191-20100321-1524.jpg


Please note that due to the Claw rotor arrangement you have to install the hall sensor exactly in the middle

On the controller side :
Bob diode from RC group gave me this link (after I manage to figure it out from Atmel application notes), it's provides a very visualization on how the BLDC commutation works:
http://users.tinyworld.co.uk/flecc/4-pole-bldc-motor031102.swf

As it could be seen, one "electrical rotation" (for 12 pole rotor this would be mean 1/12 mechanical rotation) consist of 6 steps, each steps was represent by different arrangement of which Phase should be connected to (+) positive supply and which phase should be connected to the (-) negative supply, in every step there should be exactly 1 phase connected to (+) and exactly 1 phase connected to (-). To regulate the the power, either (+) or (-) or both supply was not connected continuously but rather only a fraction of time (switched ON-OFF), The power to the motor was then governed by the period of duty cycle (the ON time) this is called PWM. In RC world typically we used 16KHz or 32Khz PWM for some reason.

All step was executed successively according to the direction of the rotation which was triggered by the state change in any of hall sensor.
Below is the typical driver (using both N & P channel mosfet) of BLDC
image001.jpg


Q1,3 & 5 was referred as the High Side, which upon ON state will connect any phase to the (+) supply
Q2, 4, 6 was referred as the Low Side, which upon ON state will connect any phase to the (-) supply.

Hence, what a controller do is exactly turning which mosfet ON and which mosfet OFF based on the input of Hall controller, while assuring that none of both High Side & Low Side mosfet in one phase having on ON together as this would means a short circuit :shock: :shock: which some times referred as pass-through.

Hope this help you in your new endeavor :D :D
 
Widodo
Thanks for all that info.

I bought the Prestolite-Leese-Neville TRUCK Alternator off ebay, BRAND NEW, for $100.00 Now, I see them listed for $125.00, Buy it Now.

There must be large trucks where you live ??? Or even Buses ??? That is where the large Alternators are hiding. :)

Like many that read these forums, I do not have money to just try stuff and end up burning things up and wasting money.

It is good to understand why things work and how they work. It may sound selfish or lazy, but, Bob Diodes posts get very confusing. IF SOMEONE were to post exactly how to get the correct components and exactly how to wire them up, and what to be very careful of, there would be much more activity on these types of Alt-motors.

I will reread your info and see if I can figure out how it works. Thanks again, Harold
 
Anytime Halord, Glad if this could help you out.

I do have also a budget constrain for this kind of hobby, in the matter of fact this is my main key driver for pursuing the budget alternatives.

Perhaps, you could share your step-by-step in this thread (or perhaps opens a new thread for this) and I will try to help you out so both of us could be that "Someone" :D

Anyway, thanks for the info on the truck alternator, will try the salvage yard first
 
widodo said:
Anytime Halord, Glad if this could help you out.

I do have also a budget constrain for this kind of hobby, in the matter of fact this is my main key driver for pursuing the budget alternatives.

Perhaps, you could share your step-by-step in this thread (or perhaps opens a new thread for this) and I will try to help you out so both of us could be that "Someone" :D

Anyway, thanks for the info on the truck alternator, will try the salvage yard first

Also look for marine salvage for alternator. I once found new military surplus Leece Neville 28V 150 Amp alternator for $28. Orphans that did not fit anything.
 
Thanks John, though it might be a little hard to find marine salvage around here. So far my best finding only the 120Amps Nissan Alternator for around US30 without its regulator.

Anyway, this weekend result update are:

The IR2101 half bridge driver work flawlessly, hence all budget N-Channel Mosfet driver is really viable. Unfortunatelly, one of the IR2101 was failing on me which immediately creating shoot through in Phase B so there goes 2 mosfet out. After removing the problematic IC and cutting both Positive & Negative supply for the shorted mosfet which means that now there are only 2 Phase available to droves the alter-motor, Surprisingly it stills runs although you could really notice that its missing one steps. Due to one step missing, I need to gives a little kick start to have it runs. I suppose it is still working though with only 2-phase due to big rotor inertia/momentum.

Lesson learned during this week-end that
1. IRF design tips clearly explain that we had to used at least 4.7mF capacitor for the IR2101, 6.8mF Tantalum capacitor works ok in my circuit.
2. Having Binary codded Hall sensor output as and index and used it to look up the commutation table, reverse logic are easily implemented by using the formula of (7- Binary Coded Hall Output) and using the result to look it up in the same commutation table.
3. Never, ever plugs any IC while connecting your controller to computer for program download, this had cost me the IR2101 and 2 burned mosfet
4. One IR2101 only designed to droves 2 mosfet, hence to go for 150Amps with IRF3007 need additional 3

Below is the video of the Alter-motor with the Home made controller running only with 2 phase
 
widodo,

Yes, thanks for sharing the how-to's in a way that the electrically challenged, like myself, can clearly understand.

If I wanted to go the route of eliminating the field coil, eg by buying one of those PMA, Permanent Magnet Alternators, like thishttp://cgi.ebay.com/PC18-PMA-Wind-T...emQQptZLH_DefaultDomain_0?hash=item45f1e4f197 on Ebay, how much would things change for using it as a motor instead?

Also, in your sensored version it looks like the hall sensors are glued on the inside face of the stator. Is clearance in the air gap a concern? If not, then how are the high efficiencies claimed by BobDiode possible with a big airgap? If significantly large airgap, significant space between magnets, can be possible in a motor with over 90% efficiency, then there's definitely some not-very-well-understood black magic involved in motor design and optimization, and I find that very intriguing, especially when such cheap but hard to build components like alternator stators are so readily available for so cheap. I realize I should just jump in and give it a go, but I need a few more gentle shoves before taking the plunge.

John
 
John, you really make my day :D , really glad that it could help you out

Anyway, If I may, I would like to take you back to the reasoning why alternator (as it is) really interesting for me as an "electronically and mechanically challenged" person before going out into PMA.
  • 1. Torque and Rpm: Confirmed in practice that the bigger the torque will give you the lower the RPM will be for a given Voltage. Looking up at the google uni: While the motor running, the fact that the stator is in fact a coil, and the rotor provides dynamically changing magnetic fields means that the stator will produce current to the opposite of the supply current, what is limiting the RPM is in fact the BEMF produced which is linear to the motor RPM. Hence, Hi-torque BLDC will have low KV, while Hi-KV motor will gives low torque on low rpm. Wouldn't be nice if some how we could decrease the rotor magnetic field for you could have high torque, but then you could have higher RPM by reducing the Rotor magnetic field. It was really like an electronic CVT if I must say. The Possibility of this capability is the main reason why I had spend 6 month of my weekend. Because It would mean I could go higher in RPM without the need of adding up another battery. The Fact that it has Iron rotor claw, also a plus as based on what i've read, Iron Flux saturation level is higher compare to even rare earth permanent magnet .
  • 2. Small Heli BLDC could runs easily to 30K rpm, but try to runs Large in-runner motor, had to remember that there are centrifugal force always try to pull the rotor out. As I'm trying to go high in RPM band, had concern is the rotor integrity. On the other side, car alternator with it's 2:1 pulley size could runs up to 12K rpm (given max Car RPM of 6,000), not mentioning that the rotor was held by very large sealed bearing so I will have less to worried about.
  • 3. Ruggedness, My eBuggy used BOSCH 80A which was salvaged from my own car(1991 Merc 280GE). My mechanic told me that It was its original alternator, which means it was quite old, yet after those rugged service (my car known to have overheating issue), all the bearing and lamination was not affected at all. I always look alternator as piece of the art product due to its quality and clearance, yet it was very affordable. The Fact that it was very rugged was also contributed because it uses electromagnet which was still ok after overheating, unlike a permanent magnet one

Challenge are:
It was clearly a black-box and unlike normal BLDC as I know of, it has overlapping coil for their 3-phase, further, the rotor claw was shaped trapezoidal, don't know how this going to affect. Will it works with non-sinusoidal commutation control ?, will the rotor field coils could be controlled by PWM ?, Hence, just try everything up with step-by-step methodology to confirm all assumption and here I am.

Now lets go back to your question:
  • 1. Removing rotor field coils and go for Permanent Magnet rotor, for me that would means to loose all the interesting things. Because you will ends up with Fix RPM, Fix KV and Fix Torque for a given Battery Voltage. This would also mean that we convert the alternator to a plain BLDC yet this is in-runner one, I would go for out-runner for its bigger torque. Not mentioning, on the price side as Turnigy biggest outrunner known to run a go-cart only cost half the price of the PM rotor.

  • 2. For the Sensored Version, I believe it was as hard as putting sensor to the normal BLDC, All you need is bench-drill to remove some stator tooth, just enough to put the sensor in it, Sand Paper just in case it rubs the rotor, and few drop of CA. Comparing the air gap of the alternator with my Turnigy 5058, did not have a noticeable difference.

  • 3. For the efficiency,honestly I don't really know, but only for illustration, the eBuggy was previously powered by 5.5HP briggs&stratton engine, yet a 5KM drives with extreme slopes only discharged a 70Amps car battery by 1 Volt. But please use that only for illustration, I'm not claiming anything. There is one people asking me to proves the power by dyno test :shock: :shock: , I would rather get my self a bag of hall sensor and mosfets :lol:

I think we could think the price should be considered as a blessing as this still a black box, yesterday I've got my self large Sanyo Denki stepper motor for 5 dollar in the flea market, only because he could not run it on battery he says :D

Is this push you gentle enough John ? :D

Additional info:
Just looked at the advertised PM motor, it says "Stop wasting up to 50% of your power energizing the
coil of standard brush type alternators. Get a real PMA!", I believe this is a bit over stated. My field coils resistance is 1.2Ohm, hence with only 4V of supply that will cost me roughly 13.3 watts, and this is only happened during the start up or severe climb up. Used a clamp meter during one climb, and the alternator pulls 50Amps on the fully charged battery, hence simple math 13.8x50=690 Watts. Even on the extreme situation the field coils only consumes 13.3/690= 1.9% of the power :).
 
Widodo,

I have been closely following your posts both here and at RCGroups and am working on building a controller. I have the stator bridge fairly well figured out but am undecided on how to drive the rotor. Thanks for all your info.
 
salty9 said:
Widodo,

I have been closely following your posts both here and at RCGroups and am working on building a controller. I have the stator bridge fairly well figured out but am undecided on how to drive the rotor. Thanks for all your info.

Thanks Salty, means a lot knowing that I'm doing a good thing, anyway may I suggest a PWM driver, so far my rotor saturates within 4-5 Volts, hence that could be starting point. The challenge is that my controller had 4 PWM output , but one of output interact with the interrupt I used for hall input. Hence, thinking on using a TTL decoder (74LS138) and using only 1 PWM output for the high-side driver, with this I will have enough spare PWM generator to drives the rotor.

For some reason, I already decides to just post in this forum, and with so many people trying to builds theirs, I suppose we will have it sooner than I was expected. :D

Would you kindly shares your progress in this thread as well ? :D
 
Something that interests me about those PMAs is the large gaps between the magnets. That seems to me to be plenty of space to add coils around the permanent magnets. I need to first find out how the flux of a permanent magnet behaves with a coil wrapped around it. If with current in the proper direction, it increases the flux, then it seems that we can have the best of both worlds by using the powered field coils to enhance the flux of the PMs only when higher torque is required.

John
 
John in CR said:
Something that interests me about those PMAs is the large gaps between the magnets. That seems to me to be plenty of space to add coils around the permanent magnets. I need to first find out how the flux of a permanent magnet behaves with a coil wrapped around it. If with current in the proper direction, it increases the flux, then it seems that we can have the best of both worlds by using the powered field coils to enhance the flux of the PMs only when higher torque is required.

John

What are you going to do with the current induced in the winding, from passing through the flux created by the stator, when your coil is not energized? Will you not have a troubling voltage to deal with?
 
John in CR said:
Something that interests me about those PMAs is the large gaps between the magnets. That seems to me to be plenty of space to add coils around the permanent magnets. I need to first find out how the flux of a permanent magnet behaves with a coil wrapped around it. If with current in the proper direction, it increases the flux, then it seems that we can have the best of both worlds by using the powered field coils to enhance the flux of the PMs only when higher torque is required.

John

Ah I see, John in CR, this is perhaps where the claim of "no cogging" came from, so instead of having rotor with N-S-N-S.. pole, it removes one of the pole right ?, so the rotor only consist of N --- N --- N ... or perhaps N-s-N-s-N ?, where s denotes a significant smaller flux density compare N due to its bigger air gap between rotor and stator, Did I get this right ? :?

Worked with commutation logic for couple of month, I could confirmed that the commutation sequence AB- , -BC, A-C was actually only pulling and pushing ONLY one of the pole, hence if the "UNUSED" pole was some how weakened or removed, it will provides less burden for the motor. Good Idea, just have to worried about the controlling logic, as the sensorless might not work due to the fact the it requires the BEMF from the unused pole. Sensored controller still might be working but need to have separated hall sensor, as 3 out of 6 commutation steps requires input from those unused pole as well. This is the same situation when I was trying to install the hall sensor to close to one of the pole. I wonder if this is the real reason Bob actually had a separate sensor magnet for his motor :eek:

Still this is very interesting, I think I should put this on my to do list :D , I supposed I could try this using a 8 dollar tower pro motor first to removes the complexity and cost.
 
John500 said:
What are you going to do with the current induced in the winding, from passing through the flux created by the stator, when your coil is not energized? Will you not have a troubling voltage to deal with?

The controller energized the rotor as a function of stator power, high excitation for start up and decreased it linear (in a certain degree) down to 1.5V for top speed, lower than that I will loosed my commutation due to lack of BEMF (Still using the sensorless controller). Hence, never had that situation when stator producing power but rotor was not energized. The problem lies somewhere else though. For PWM rotor driver I'm using walkera brushed ESC, never had a problem, except when doing hard braking going down the slope when its internal protector kicking in, which after that requires a ESC restart. Still couldn't understand the braking commutation logic to understand why this is happened. Anyway for the build in PWM rotor controller, had a free-wheeling diode as well as some over voltage protection as well.

Or perhaps you could just used linear regulator like LM317 or perhaps LM138 and put a potentio to adjust the output, had this implement before, just don't like the fact that the potentio couldn't be controlled by computer :D

hope this usefull
 
Widodo,

I am wondering about the regen possibilities with varying the rotor voltage. It would be logical that increased voltage would increase the generated voltage. I am going to spin the altermotor up with another motor and check the output while varying the rotor voltage. It will give me something to do until the parts I ordered come.

Salty
 
John in CR
Why don't you give me a rough date when you could come up and get your wood ??? I HOPE to be done with this sawing job on Wednesday, then, see about sawing that Balsa Log. I just might Didi off to Florida around May 10 or so, for a couple weeks, and see if I can drag my Son back for a week.

I already converted a 100 Amp Delco, to a permanent magnet, using ½" thick (2), Neo ring magnets of N45 grade. Just got the wrong spacer ring on the inside of the front housing, and pinched the h#ll out of my left index finger. Thought I might have to go to a machine shop, to get the Alt off my finger. Finally screw drivered it off, and bled like a stuck pig. :oops: :shock: :roll: :roll:

Tried to spin it up om my lathe, but, didn't have a sufficient number of fingers working, to take readings. I was wanting a Wind generator unit, using the Alt.

I just need to pull the front off and swap out the spacer rings, using someone else's fingers :lol: :lol:
 
salty9 said:
Widodo,

I am wondering about the regen possibilities with varying the rotor voltage. It would be logical that increased voltage would increase the generated voltage. I am going to spin the altermotor up with another motor and check the output while varying the rotor voltage. It will give me something to do until the parts I ordered come.

Salty

I believe that totally possible, the challenge would be only on how to redirect the stator input (or in this case the generator 3-phase output) into the diode bridge. I used to think about using a band of otomotive relays but still haven't figure out how this large current physical switching will affect the mosfet.

The alternator regulator works by varying the rotor supply to assure that the output always 13.8V under any load condition, so you might experience this in your experiment. Just be careful as unloaded alternator could produce quite high voltage in a given rpm & rotor supply :wink:
 
I removed the brushes and regulator. IF you go Permanent magnet, it will be difficult to keep the slip rings from being destroyed.

I got readings up to 14 V, but, I have the 3 delta windings seperated. Might be why I got low readings, and, the belt kept jumping off. That's my safety overspeed protection on the lathe, so, off balance chunks of wood don't rotate out of control. :shock:
 
salty9 said:
Widodo,

My readings lead me to believe that the mosfet bridge is a two-way bridge and one of the things you have to watch out for is regen into a full battery. Isn't electronics fun. Just look at Live for Physics' video of blowing shit up :mrgreen:
http://www.youtube.com/watch?v=dree0rTr1HM

Ha ha... that was quite a show, a fireball show :D. We're on the RC world really aware of the LiPo precaution, in fact, I have to throw at least 4 3-cell pack to the river already within two year due to cell deformation.

You see every cell in the LiPo battery (i.e. one pouch) should never be allowed to be discharged below 2.7V and never overcharged beyond 4.235V, or the cell would be damaged, noticeably by bulging. If I could recall correctly, in my high school days, there was an experiment of showing up how reactive the alkali was, and for that we used Natrium , one level less reactive than Lithium, and it's already show a very dramatic fireball and smoke when mixed up with water.

Lithium, which was only less superior to hydrogen, was one order higher in the periodic table, even it is reactive to oxygen. Any contact with oxygen will immediately resulting in a fireball. A friend of mine, which unfortunately skips a lot of chemistry class in high school, had learn a lesson by puncturing-in a bulging lipo and produce exactly as filmed by live for physic. Another friend of mine turn his RC helicopter to ashes by over discharging his battery (using old brushed ESC which don't have a low voltage cut off)

Otherwise you are able to have a balance charger (i.e. charging every cell individually) I strongly suggest not have the Regenerative braking, because even though you could guarantee the maximun voltage for a series of LiPo cells, you could not guarantee that all cells are in the same State of Charged (SOC). Hence, it could means that you could easily overcharge one of the cells accidentally even if you are not trans passing the total voltage limitation. Not mentioning another complex behavior of the LiPo, ie, upon reaching 4.2V, lipo was only achieving 70% of charged capacity, charger should maintains the charging voltage of 4.2 until the current drops to 10% of the initial charging current :roll: . Quite complex really, that is why LiPo charger always used microcontroller to control the charging process.

Further, there are also charging limitation imposed on LiPo, the first generation, as I recall two year ago, was only being able to be charge with 1C, which means that for example a 4000mAH lipo battery could only be charge with a maximum current of 4 Amps. Still, new generation LiPo was able to handle 1.5 - 2, and higher in LiPoFe. Hence, otherwise you have a very-very long down slope for you to apply the regen effectively, I suppose it will only provides small contribution due to above limitation.

Anyway, those limitation will not be there if we just use nor Pb car battery. For my eBuggy, I used Delcor Calcium battery which provides significantly higher Charge/Discharge rate while comes in smaller package and weight. But for performance shakes, nothing beats LiPo's family discharge rate.

On the mosfet is a two way half bridge, Honestly I haven't think about it, need to digs out some book first. Maybe perhaps anybody could confirmed this ? :roll:
 
Harold in CR said:
I removed the brushes and regulator. IF you go Permanent magnet, it will be difficult to keep the slip rings from being destroyed.

I got readings up to 14 V, but, I have the 3 delta windings seperated. Might be why I got low readings, and, the belt kept jumping off. That's my safety overspeed protection on the lathe, so, off balance chunks of wood don't rotate out of control. :shock:

You wont be needing the slip ring if you go for PM right Harlod ?, would love to see any picture if you have the opportunity, like John said, we might get the best of both world :D .

Anyway, just curios about the sensor arrangement, have you decided where to put it up ?
 
Back
Top